Completed acquisition by
Hunter Douglas N.V. of
convertible loan notes and
certain rights in 247 Home
Furnishings Ltd. In 2013 and
the completed acquisition by
Hunter Douglas N.V. of a
controlling interest in 247
Home Furnishings Ltd. in
2019
Final report
14 September 2020
© Crown copyright 2020
You may reuse this information (not including logos) free of charge in any format or
medium, under the terms of the Open Government Licence.
To view this licence, visit www.nationalarchives.gov.uk/doc/open-government-
licence/ or write to the Information Policy Team, The National Archives, Kew, London
TW9 4DU, or email: [email protected]si.gov.uk.
Website: www.gov.uk/cma
Members of the Competition and Markets Authority
who conducted this inquiry
Kirstin Baker (Chair of the Group)
Robin Foster
Anne Fletcher
Mark Thatcher
Chief Executive of the Competition and Markets Authority
Andrea Coscelli
The Competition and Markets Authority has excluded from this published version
of the report information which the Inquiry Group considers should be excluded
having regard to the three considerations set out in section 244 of the Enterprise
Act 2002 (specified information: considerations relevant to disclosure). The
omissions are indicated by []. Some numbers have been replaced by a range.
These are shown in square brackets. Non-sensitive wording is also indicated in
square brackets.
1
Contents
Page
Summary .................................................................................................................... 4
Background ........................................................................................................... 5
The Parties ...................................................................................................... 5
The Transactions ............................................................................................. 5
The industry ..................................................................................................... 6
Jurisdiction....................................................................................................... 8
Counterfactual ................................................................................................. 8
Market Definition .............................................................................................. 9
Our approach to assessing the 2019 Transaction ........................................... 9
Countervailing factors .................................................................................... 12
Findings .................................................................................................................... 15
1. The reference ..................................................................................................... 15
2. The Parties ......................................................................................................... 17
Main Parties ........................................................................................................ 17
Hunter Douglas .............................................................................................. 17
247 Home Furnishings .................................................................................. 19
3. The Transactions ................................................................................................ 21
Timeline of events ............................................................................................... 21
Events leading up to the Transactions and valuations ................................... 21
The Parties’ rationale for the Transactions .................................................... 26
4. Industry background ........................................................................................... 29
Window coverings overview ................................................................................ 29
Distribution Channels .......................................................................................... 32
In-store and In-home ..................................................................................... 32
Online ............................................................................................................ 33
Figure 8: 247 Customer Journeys – Devices used March 2013 to April 2020 .......... 34
Competitors and main parties ............................................................................. 36
Trends within the Sector ..................................................................................... 37
Improvements in the customer offering ......................................................... 37
Technology .................................................................................................... 38
5. Relevant merger situation ................................................................................... 39
Relevant Merger Situation – the jurisdictional test .............................................. 39
Enterprises ceasing to be distinct ....................................................................... 40
Enterprises .................................................................................................... 40
Ceasing to be Distinct .................................................................................... 40
The 2013 Transaction .................................................................................... 43
The 2019 Transaction .................................................................................... 48
The Share of Supply Test – Nexus with UK ........................................................ 49
Time at which the relevant thresholds should be assessed ........................... 49
Share of Supply .................................................................................................. 52
Time period for investigating mergers ................................................................. 53
6. Counterfactual..................................................................................................... 55
Introduction ......................................................................................................... 55
The CMA’s counterfactual assessment framework ............................................. 55
Parties’ views on the appropriate counterfactual ................................................ 56
Assessment ........................................................................................................ 58
Introduction .................................................................................................... 58
2
Scenario 1: Continuation of majority ownership by the 247 Founding
Shareholders ................................................................................................ 59
Scenario 2: Alternative purchaser of Founding Shareholders’ stakes in 247 . 60
Scenario 3: Alternative purchaser for 100% of 247 ....................................... 62
Assessment of scenarios ............................................................................... 63
Findings on the most likely counterfactual .......................................................... 65
7. Market definition .................................................................................................. 66
Product market definition .................................................................................... 66
Curtains and shutters .................................................................................... 68
Ready-made blinds ........................................................................................ 71
In-store and in-home ..................................................................................... 80
Geographic market definition .............................................................................. 84
Conclusion on market definition .......................................................................... 84
8. Competitive assessment ..................................................................................... 85
Market shares ..................................................................................................... 85
Approach to market share calculations .......................................................... 86
Market share estimates ................................................................................. 87
How competition works in the market ................................................................. 89
Parties’ submissions on the customer journey and importance of online
advertising .................................................................................................... 90
Traffic and marketing spend .......................................................................... 91
Search behaviour and customer journey ....................................................... 95
Conclusion ................................................................................................... 100
Closeness of competition .................................................................................. 101
The Parties’ service proposition ................................................................... 102
Online presence .......................................................................................... 107
Survey evidence .......................................................................................... 120
Internal documents ...................................................................................... 123
Pricing analysis ............................................................................................ 125
Third party evidence .................................................................................... 125
Competitive significance of 247 ................................................................... 126
Conclusion on closeness of competition ...................................................... 128
Remaining constraints ...................................................................................... 128
Competition from other online M2M blind retailers ...................................... 129
Competition from multi-channel retailers ..................................................... 138
Competition from eBay and Amazon ........................................................... 145
Out-of-market constraints ............................................................................ 149
Conclusion on remaining constraints ........................................................... 152
Impact of the Merger ......................................................................................... 153
Parties submissions on the impact of the merger ........................................ 155
Our assessment of the Parties’ submissions ............................................... 156
Conclusion on competitive assessment ............................................................ 158
9. Countervailing factors ....................................................................................... 161
Entry and Expansion ......................................................................................... 161
CMA framework for assessing entry and expansion .................................... 161
The Parties’ views ....................................................................................... 162
Our assessment of barriers to entry and expansion .................................... 162
Possible sources of entry and expansion .................................................... 171
Conclusions ................................................................................................. 184
10. The decision ..................................................................................................... 186
11. Remedies .......................................................................................................... 187
3
Introduction ....................................................................................................... 187
CMA framework for assessing remedies .......................................................... 188
Overview of remedy options ............................................................................. 188
Full divestiture of 247 ........................................................................................ 190
Description of remedy .................................................................................. 190
Remedy design issues ................................................................................ 191
Assessment of the effectiveness of full divestiture ...................................... 199
Partial divestiture of 247 ................................................................................... 199
Description of remedy .................................................................................. 199
Remedy design issues ................................................................................ 199
Assessment of the effectiveness of partial divestiture ................................. 211
Conclusion on remedy effectiveness ................................................................ 211
Relevant customer benefits (RCBs) .................................................................. 211
Framework for assessing RCBs .................................................................. 212
Parties’ and third-party views on RCBs ....................................................... 213
Assessment of RCBs ................................................................................... 213
The proportionality of effective remedies .......................................................... 213
Framework for assessment of proportionality of remedies .......................... 213
Conclusion on proportionality ...................................................................... 217
Remedy implementation ................................................................................... 217
Divestiture risks ........................................................................................... 218
Composition risk .......................................................................................... 218
Asset risk ..................................................................................................... 218
Purchaser risk .............................................................................................. 219
Decision on remedies ....................................................................................... 220
Appendices
A: Terms of reference and conduct of inquiry
B: Provisional view on the ability of Hunter Douglas to block a sale by the 247
Founding Shareholders
C: CMA analysis of Parties’ survey methodology
D: Price analyses
E: Online presence and the role of Google
F: Generating traffic as a potential barrier to entry and expansion
Glossary
4
Summary
Overview
1. The Competition and Markets Authority (CMA) has found that the completed
acquisition by Hunter Douglas N.V. (Hunter Douglas) of a controlling interest
in 247 Home Furnishings Ltd (247, and together with Hunter Douglas, the
Parties or the Merged Entity) in 2019 (the 2019 Transaction) has resulted, or
may be expected to result, in a substantial lessening of competition (SLC) in
the online retail supply of made-to-measure (M2M) blinds in the UK.
2. The CMA has also found that the completed acquisition by Hunter Douglas of
convertible loan notes and certain rights in 247 Home Furnishings in 2013
(the 2013 Transaction) has not resulted in the creation of a relevant merger
situation (RMS) within the meaning of the Enterprise Act 2002 (the Act).
Our inquiry
3. On 1 April 2020, the CMA referred the 2013 Transaction and the 2019
Transaction (together, the Transactions) for an in-depth phase 2 merger
inquiry.
4. The CMA is required by its terms of reference to decide with respect to each
of the Transactions:
(a) whether the Transaction constitutes an RMS;
(b) if so, whether the Transaction has resulted or may be expected to result in
an SLC within any market or markets in the United Kingdom for goods or
services; and
(c) whether action should be taken for the purposes of remedying, mitigating
or preventing any SLC or resulting adverse effect we have identified.
5. In addressing the questions above, we have considered a range of different
evidence that we received from the Parties, other retailers and suppliers. This
includes evidence received through submissions, responses to information
requests, telephone calls, and hearings. We have also considered a survey of
their customers prepared and submitted by the Parties that we consider is in
accordance with our best practice. Given that competition in the relevant
market primarily occurs online (as discussed below), we also have analysed
how online search is utilised by the Parties and their competitors in the retail
supply of online M2M blinds. Lastly, we have considered the Parties’
response to our Provisional Findings, which we published on 16 July 2020.
5
Background
The Parties
6. The acquirer is Hunter Douglas, a global provider of window coverings,
including blinds, shutters and curtains. In the UK, Hunter Douglas operates
through different companies at manufacturing, wholesale and retail level,
using several different brands. With respect to online M2M blinds, Hunter
Douglas is active in the UK through its subsidiary Blinds2Go Limited
(Blinds2Go). Blinds2Go is the UK’s largest online M2M retailer for blinds. In
2019 Hunter Douglas had global revenues of approximately £3 billion.
7. The target, 247, is a UK-based online supplier of window coverings including
blinds, shutters and curtains. In 2019, 247’s global turnover was £22.2 million.
8. The Parties overlap in the supply of window coverings in the UK (including the
online retail supply of blinds, shutters and curtains). However, the principal
area of overlap between the Parties is between Hunter Douglas’ subsidiary
Blinds2Go and 247 in relation to the online retail supply of M2M blinds in the
UK. Accordingly, this competitive overlap has been the focus of our inquiry.
The Transactions
9. Hunter Douglas acquired its interests in 247 through two separate
transactions in 2013 and 2019, respectively. Notwithstanding these separate
transactions, the Parties submit that they entered into the 2013 Transaction
on the understanding that this was a single acquisition by Hunter Douglas of
247 that would ultimately complete in 2019. The Parties accordingly view the
2019 Transaction as a formality that gave effect to their previous agreement in
2013.
The 2013 Transaction
10. Pursuant to the 2013 Transaction, Hunter Douglas invested in 247 via the
acquisition of convertible loan notes which had been issued by 247 to 247’s
founding shareholders (the 247 Founding Shareholders).
11. Attached to these loan notes were certain rights in 247 granted to Hunter
Douglas, including: (i) 49% of the voting rights and a 49% share of the profits
in 247; (ii) the right to convert the loan notes at any time to ordinary shares;
(iii) the right to nominate a non-executive Director to the 247 Board; and (iv)
certain veto rights in respect of the 247 business.
12. At the same time, reciprocal put and call options were granted to both Hunter
Douglas and each of the 247 Founding Shareholders. Under the put and call
6
options, the 247 Founding Shareholders could each require the purchase of
their shares by Hunter Douglas and Hunter Douglas could require the sale of
the shares held by each of the 247 Founding Shareholders by written notice in
the period 1 March to 1 June 2019.
13. The terms of the 2013 Transaction prevented either Party from publicising the
transaction. The Parties submitted that the 2013 Transaction was kept
confidential in order to avoid the potential for ‘channel conflicts’ between
Hunter Douglas, as a wholesale supplier, and its customers as retail
suppliers. We understand that Hunter Douglas did not have a retail presence
in the supply of online M2M blinds in the UK prior to the 2013 Transaction.
The 2019 Transaction
14. Pursuant to the 2019 Transaction, Hunter Douglas acquired 100% of the
shares in 247. This followed an indication from the 247 Founding
Shareholders to Hunter Douglas that they both intended to exercise their put
options granted in 2013. The 2019 Transaction completed on 28 February
2019.
Other relevant transactions in the period between the 2013 and 2019 transactions
15. Hunter Douglas acquired two additional businesses active in the retail supply
of online M2M blinds in the UK in the intervening period between the
Transactions.
16. On 21 June 2016 Hunter Douglas acquired a 60% stake in Blinds2Go (the
2016 Transaction). Hunter Douglas subsequently acquired a further 5%
interest in Blinds2Go in 2019.
17. On 21 July 2017 Hunter Douglas acquired Hillarys (the 2017 Transaction),
which at the time had a presence in the supply of online M2M blinds through
Web Blinds. Web Blinds has subsequently been incorporated into Blinds2Go.
This acquisition was reviewed and cleared by the CMA at phase 1.
The industry
18. As noted above, the primary area of overlap between the Parties is the online
retail supply of M2M blinds. These products are part of the broader window
coverings sector, which also includes curtains and shutters.
19. Window coverings (including blinds) typically are supplied in either a ready-
made or M2M format. Ready-made products are largely finished and sold in
one of many available sizes, whereas M2M products are tailored to the
7
specifications of the customer. The main channels through which window
coverings are sold in the UK are the in-home, in-store and online retail
channels.
20. In-store and in-home are traditional retail channels in which customers have
some degree of interaction with the product or a salesperson prior to
purchase. We also note that some of these retailers also have an online
presence, although not all in-store or in-home retailers sell online. Those
retailers who sell in-store/in-home and online are referred to as multi-channel
retailers.
21. With respect to online M2M blinds, these products are purchased through
websites that enable customers to customise blinds in accordance with their
desired measurements and design preferences. This differentiates M2M
blinds from ready-made products. In contrast to the in-home and in-store
channels, the leading retailers of online M2M blinds provide limited sales
advice prior to purchase and typically require customers to fit the blind
themselves once they have received their order. Competition between
retailers primarily occurs online and so retailers’ generation of website traffic
through online search (primarily through Google), their position in search
rankings and the use of online advertisements are of particular competitive
importance in the supply of online M2M blinds. We have therefore considered
these parameters of competition as part of our competitive assessment.
22. In addition to retailers’ websites, online marketplaces (namely Amazon and
eBay) also allow retailers to sell blinds. We understand that the majority of
sales through these channels are for ready-made blinds. This may be
reflective of the fact that these platforms do not offer the same functionality
and customer service options as online M2M blinds retailers and therefore are
not directly comparable.
23. The competitive landscape of the window coverings sector differs by product
type and channel. With respect to the broader window coverings sector as a
whole, multi-channel retailers are the leading suppliers with Dunelm, Hillarys,
John Lewis, and Next being largest competitors.
24. We note however that the competitive landscape is different for the online
retail supply of M2M blinds. In particular, the leading retail suppliers of online
M2M blinds in the UK are focussed primarily on supplying M2M blinds online
(although they may supply other window coverings to a lesser extent).
8
Findings
Jurisdiction
25. We have assessed whether each of the 2013 Transaction and the 2019
Transactions created a RMS.
26. We conclude that the 2013 Transaction did not create a RMS. The rights
attached to the convertible loan notes acquired by Hunter Douglas through
the 2013 Transaction were sufficient to give it material influence over 247’s
policy. However, we were not satisfied that the share of supply test is met in
relation to the 2013 Transaction, taking account of the particular and unusual
circumstances of this case, in particular, the very lengthy period which had
elapsed since the 2013 Transaction occurred and the lack of overlap between
the Parties at the time of the 2013 Transaction.
27. In contrast, we have found that the 2019 Transaction created a RMS. We find
that Hunter Douglas’ acquisition of 100% of the shares in 247 amounts to the
acquisition of a controlling interest in 247. In particular, as a consequence of
owning 100% of 247, Hunter Douglas acquired the ability to unilaterally
determine 247’s strategic policy and increased its share of the company’s
profits. Moreover, we find that the share of supply test is met as a result of the
Parties having a combined share in excess of 25% in the online retail supply
of M2M blinds in the UK.
28. In light of our findings on jurisdiction, our substantive assessment considers
whether the 2019 Transaction has resulted, or may be expected to result, in
an SLC in the UK.
Counterfactual
29. The counterfactual is an analytical tool used to help answer the question of
whether a merger may be expected to result in an SLC. It does this by
providing the basis for a comparison of the competitive situation in the market
with the merger against the most likely future competitive situation in the
market absent the merger.
30. We may examine several possible scenarios to determine the appropriate
counterfactual. We have found no evidence to suggest that Blinds2Go would
have done anything other than continue to compete in line with the conditions
prevailing at the time of the 2019 Transaction. For 247, we have considered
three scenarios:
9
(a) Scenario 1: Continuation of majority ownership by 247 Founding
Shareholders
(b) Scenario 2: Alternative purchaser of the 247 Founding Shareholders’
stake in 247
(c) Scenario 3: Alternative purchaser for 100% of 247
31. We find that, absent the 2019 Transaction, the most likely scenario is that the
247 Founding Shareholders would have sought to sell their shares in 247 to a
third-party buyer (as per Scenario 2). In our view, it was the intention of the
247 Founding Shareholders to sell their shares in 247 and that, at the point of
the 247 Founding Shareholders selling their shares, Hunter Douglas would no
longer have been able to exercise the veto and other rights it previously held
in 247. This would have resulted in 247 having more independence than it
had prior to the 2019 Transaction.
Market Definition
32. Our finding is that the relevant market for the assessment of the 2019
Transaction is the online retail supply of M2M blinds in the UK. This position is
supported by our assessment of the Parties’ own customer survey, their
monitoring activities, as well as the views received from third parties.
33. We have not included other window covering products, ready-made blinds or
the in-store and in-home channels in the relevant market. However, we note
that market definition does not determine the outcome of our competitive
assessment and we take into account the constraint of these alternative
products where relevant. With respect to ready-made blinds in particular, we
acknowledge that these products do act as a distant competitor to online M2M
blinds.
Our approach to assessing the 2019 Transaction
34. We have assessed the competitive effects of the 2019 Transaction by
reference to a horizontal unilateral effects theory of harm, that is where one
firm merges with a competitor that previously provided a competitive
constraint, allowing the merged firm profitably to raise prices on its own and
without needing to coordinate with its rivals. In particular, we have assessed
whether Hunter Douglas acquiring 100% control over 247 and increasing its
share of the company’s profits as a result of the 2019 Transaction would likely
result in Hunter Douglas increasing prices and/or lowering the quality of its
products or customer service, and/or reducing the range of its
products/services across the brands it controls.
10
Competitive assessment
35. We have found that the 2019 Transaction may be expected to result in an
SLC in relation to the retail supply of online M2M blinds in the UK.
36. In reaching this view, we found that the Parties are two of only three retailers
with a market share above 5%. Blinds2Go is the largest supplier of online
M2M blinds in the UK and several times larger than the second largest
supplier, while 247 is of meaningful scale in this market as the third largest
supplier and is approximately three times larger than the fourth largest
supplier. Outside of the top three suppliers, we have identified few retailers
with a market share above 1%. In light of these findings, we find that the
combined share of the Parties is very high, at 60-70%, and that the increment
from 247 is significant in the context of an already concentrated market.
37. We find that the Parties’ offerings in terms of price, quality and range are
similar and we have identified only two other retailers that have a broadly
similar offering to the Parties.
38. Our assessment of the Parties’ online presence shows that the Parties are
two of only three retailers that consistently rank highly in Google paid search
results, indicating that the Parties are both highly effective at competing for
the top positions in paid search results. Our assessment also shows that the
Parties, together with two other retailers, feature frequently in the top three
organic search positions.
39. Evidence on the Parties’ monitoring of competitors’ prices, as recorded in their
own documents, shows that the Parties monitor each other and that the set of
other retailers monitored is relatively small. Survey evidence on the reported
diversion of Blinds2Go’s and 247’s customers is also consistent with the
Parties being close competitors, with the constraint from Blinds2Go on 247
appearing to be stronger than the constraint from 247 on Blinds2Go.
40. Overall, we find the Parties to be close competitors that pose a significant
competitive constraint on each other.
41. We have also assessed the post-merger constraints on the Parties. With
respect to other suppliers of online M2M blinds, we consider that Interior
Goods Direct, which is only slightly larger than 247, is the only other
significant constraint on the Parties. Whilst we have identified a number of
smaller online M2M blinds retailers, we do not view them as an effective
competitive constraint on the Parties, individually or in aggregate. This is
reflected in their limited share of the market, the fact that they do not appear
to be closely monitored by the Parties, the Parties’ own survey evidence, and
also the limited visibility of smaller suppliers in search results.
11
42. Further, we find that multi-channel retailers currently exert only a limited
constraint on the Parties and are not an effective alternative for most of the
Parties’ customers. This is reflected in the limited share of multi-channel
retailers in the supply of online M2M blinds, their potentially differentiated
product range (with respect to price and quality), limited online range and lack
of prominence in online search. We also note the lack of consistent monitoring
of multi-channel retailers by the Parties and other online suppliers. We find
that the Parties’ survey evidence relating to multi-channel retailers potentially
overstates the strength of their constraint. In particular, we consider that the
reported diversion to large multi-channel retailers is likely subject to an
upward bias, due to customers being more familiar with these brands but
potentially unaware of the true nature of their offerings. Notwithstanding this
finding, we have assessed whether the constraint they exert may increase
going forward in our assessment of the potential entry and expansion of
retailers.
43. We have found that online marketplaces are a weak constraint on online M2M
blinds retailers. In particular, these platforms are not comparable to online
retailers’ websites in terms of functionality, and the majority of their sales are
of ready-made rather than M2M blinds. We also note that marketplaces do not
constitute a standalone constraint given that their position is attributable to
collections of individual retailers.
44. With respect to out-of-market constraints, we find that other window covering
products and M2M blinds sold through the in-store and in-home channel do
not pose a material competitive constraint on the Parties, while ready-made
blinds pose a weak competitive constraint. Whilst we recognise that the out-
of-market constraints, in aggregate, impose some degree of constraint on the
Parties’ ability to raise prices due to the aggregate diversion to these
alternatives, we find that this is likely to only exert a weak competitive
constraint on the Parties, and further note that the Parties’ survey evidence
and monitoring activities indicate other retailers’ online M2M blinds as being
the main competitive constraint on the Parties.
45. As part of our competitive assessment we have found that the 2019
Transaction results in Hunter Douglas having the ability and the incentive to
raise both 247 and Blinds2Go’s prices (or otherwise worsen the offering of
247 and Blinds2Go). This conclusion is informed by our findings that the
Parties are close competitors, with evidence of diversion between them.
Hunter Douglas has acquired the ability to increase 247’s prices as a direct
consequence of the 2019 Transaction. Additionally, we find that Hunter
Douglas will have an incentive to increase 247’s prices, as Hunter Douglas
will benefit from a significant share of sales that would likely be diverted to
Blinds2Go. At the same time, we also find that Hunter Douglas has the ability
12
and the incentive to increase Blinds2Go’s prices, with Hunter Douglas now
benefitting from a 100% interest in 247, rather than only 49% pre-merger.
Countervailing factors
Entry and Expansion
46. We have considered factors that may mitigate or prevent the effect of the
merger on competition and in particular whether entry or expansion by third
parties might prevent the SLC identified. We have concluded that no such
entry or expansion would be timely, likely, and sufficient as regards any
individual current or potential competitor, or when considered in aggregate. In
reaching this view, we have considered both whether any barriers to
entry/expansion in the relevant market exist, and whether there is evidence of
actual or planned entry/expansion by rivals.
47. We have found that there is some evidence of barriers to entry and expansion
in the retail supply of online M2M blinds. These barriers relate to generating
website traffic, and to a lesser extent to website costs, brand awareness and
customer loyalty. However, we also note that the impact of any such barriers
may vary depending on the nature of the firm seeking to enter or expand.
48. Whilst it may be the case that individual barriers may in some circumstances
be overcome, we note that a new entrant to the market will likely find
themselves faced with a series of barriers to entry which might have a
significant cumulative effect on entry. With respect to existing rivals, we find
that barriers to further expansion may not be as high as for new entrants,
however the Parties’ existing strengths in the market for online M2M blinds
(as discussed in the competitive assessment section) mean that it is likely to
be difficult for rivals to achieve sufficient expansion to replace the loss of 247
as an independent competitor.
49. In any event, evidence obtained from third parties in relation to actual and/or
planned entry or expansion in this market does not show that entry or
expansion will be timely, likely and sufficient.
50. Of the leading online M2M blinds suppliers contacted in our inquiry, we
understand that certain other competitors may have plans to grow, however,
the evidence available to us does not reliably indicate how these growth plans
would be achieved. In particular, we note that Decora, a manufacturer of M2M
blinds, recently entered the market through its acquisition of Swift Direct
Blinds. However, this acquisition reflects the expansion of an existing
competitor, rather than entry by a new competitor. Moreover, the growth plans
of smaller online M2M blinds retailers would have to considerably outperform
13
an already fast-growing market in order to provide a sufficient constraint to
mitigate the effects of the Merger between the first and third largest retailers.
In this regard, we have found insufficient evidence to demonstrate how any
stated growth plans would be achieved so as to result in these competitors
providing a significantly increased individual or aggregate constraint on the
Parties post-Merger. We also observe limited growth from smaller existing
retailers in recent years. Indeed, the fact that there has been little change in
the identity of the leading suppliers in the market over several years suggests
that there is a degree of incumbency advantage in the market that may
constrain further expansion.
51. The evidence received from multi-channel retailers suggests a variety of
different plans regarding entry or expansion, however the evidence does not
demonstrate that any expansion or re-entry into the market will be timely,
likely, and sufficient. In addition, whilst they may have expressed a previous
interest in developing a presence in this market, all of these retailers have
indicated to us that their plans have been significantly impeded by the current
COVID-19 pandemic. Moreover, current plans for future growth through the
online channel for multi-channel retailers encompass their entire online
product offering, of which M2M blinds comprise a small part. Therefore, even
if entry or expansion from these retailers was timely and likely (which we do
not consider to be the case), based on the evidence provided to us (including
pursuant to our legal information gathering powers), we consider that any
entry or expansion would not be sufficient (either individually or in aggregate)
to constrain the Merged Entity.
52. Whilst we have considered different potential sources of entry and expansion
in the online M2M blinds market, the evidence available to us indicates that
even if they were to be considered on an aggregated basis, they would not be
timely, likely and sufficient.
Conclusion on the substantial lessening of competition test
53. We find that the 2013 Transaction did not create an RMS and that the 2019
Transaction did create an RMS.
54. For the reasons discussed above, we have found that the 2019 Transaction
has resulted in, or may be expected to result in, an SLC as a result of
horizontal unilateral effects in the online retail supply of M2M blinds in the UK.
In particular, we find that the 2019 Transaction removes a direct competitor
from this market, resulting in an ability and incentive for the Merged Entity to
increase retail prices, lower the quality of its products or customer service,
and/or reduce the range of its products/services.
14
Remedies
55. Having concluded that the 2019 Transaction has resulted in, or may be
expected to result in, an SLC, we are required by the Act to decide what, if
any, action should be taken to remedy, mitigate or prevent that SLC or any
adverse effect resulting from the SLC.
56. In deciding on the appropriate remedy, the CMA will seek remedies that are
effective in addressing the SLC and its resulting adverse effects and will then
select the least costly and intrusive remedy that it considers to be effective,
having regard to the need to achieve as comprehensive a solution as is
reasonable and practicable to the substantial lessening of competition and
any adverse effects resulting from it. The CMA will also seek to ensure that no
remedy is disproportionate in relation to the SLC and its adverse effects.
57. We considered the following remedy options:
(a) Requiring the full divestiture of 100% of the ordinary share capital of 247;
and
(b) Requiring a partial divestiture of 51% of the ordinary share capital of 247.
58. We have found that both of these options would, in principle, be an effective
and proportionate remedy to address the SLC and its resulting adverse
effects we have found, provided a suitable purchaser could be found.
59. We find that a partial divestiture of 51% of the ordinary shares of 247 to be the
least onerous effective remedy. While full divestiture has a lower cost
associated with it than the partial divestiture option as it would not require any
ongoing monitoring by the CMA, this option is also significantly more intrusive
as it would leave Hunter Douglas with no shareholding in 247. This compares
to the conditions of competition in the counterfactual that we found, where it
would still hold a 49% stake. In addition, we have identified a number of
conditions in order to ensure that a suitable remedy is achieved and that the
CMA has sufficient oversight over the remedies process. In particular, any
suitable purchaser will be required to operate the business in a manner that
ensures effective competition between Blinds2Go and 247.
15
Findings
1. The reference
On 1 April 2020, in exercise of its duty under section 33(1) of the Enterprise
Act 2002 (the Act) the Competition and Markets Authority (CMA) referred:
(a) the completed acquisition by Hunter Douglas N.V. (Hunter Douglas)
of convertible loan notes and certain rights in 247 Home Furnishings
Ltd (247) in 2013 (the 2013 Transaction); and
(b) the completed acquisition by Hunter Douglas of a controlling interest
in 247 in 2019 (the 2019 Transaction) (together with the 2013
Transaction, ‘the Transactions’).
for further investigation and report by a group of independent panel
members (the Inquiry Group).
The terms of reference, along with information on the conduct of the
inquiry, are set out in Appendix A. The Inquiry Group is required to publish
its final report by 15 September 2020.
Section 35 of the Act sets out the statutory questions that the CMA needs
to decide in a Phase 2 inquiry pursuant to, as in this case, a reference
under section 22 of the Act. In this investigation, the Inquiry Group must
decide:
(a) For each of the 2013 Transaction and the 2019 Transaction, whether
a relevant merger situation (RMS) has been created (the First
Statutory Question); and
(b) If so, in each case, whether the creation of that situation has resulted
or may be expected to result in a substantial lessening of
competition (SLC) within any market or markets in the United
Kingdom for goods or services (the Second Statutory Question).
If the Inquiry Group answers both statutory questions in the affirmative in
relation to any of the Transactions, then the Group must also decide what
action the CMA should take for the purposes of remedying, mitigating or
preventing any SLC or resulting adverse effect we have identified. This is
the subject of the notice of possible remedies we have published alongside
these provisional findings, in which we have discussed what measures
could effectively remedy the SLC we have provisionally found.
16
This document, together with its appendices, constitutes the Inquiry
Group’s findings, published in line with the CMA’s rules of procedure.
Further information relevant to this Inquiry, including non-confidential
versions of the submissions received from Hunter Douglas and 247, can be
found on the CMA’s website.
Throughout this document we refer to Hunter Douglas and 247 collectively
as the Parties and as applicable, the Merged Entity.
17
2. The Parties
Main Parties
Hunter Douglas
Hunter Douglas is a global provider of window coverings such as blinds,
shutters and curtains, and is headquartered in The Netherlands. The
company was founded by H. Sonnenberg in 1946, although the origins of
the company go back to 1919. The Hunter Douglas Group went public in
1969; the majority of its shares are owned by the Sonnenberg family.
1
The
Hunter Douglas Group is comprised of 134 companies with 47
manufacturing and 87 assembly operations and marketing organisations
across more than 100 countries.
2
In the UK, Hunter Douglas operates through different companies at the
manufacture, wholesale and retail level, using several different brands.
3
At
the retail level, Hunter Douglas is active in the UK through Thomas
Sanderson Limited, Hillarys Blinds Limited, Blinds2Go Limited (Blinds2Go),
Tuiss LLP and 247 Home Furnishings Limited. This is illustrated in Figure 1
below.
1
The Sonnenberg Family own 82.68% common shares and 99.4% preferred shares. Hunter Douglas website,
accessed 30 June 2020.
2
See Hunter Douglas Annual Report 2019, page 1
3
At wholesale level, Hunter Douglas is active in the UK through: Stevens (Scotland) Limited, Arena Blinds
Limited, Custom West Trading Limited, Holis Industries Limited, Orgon Windows Fashion Limited and Orgon
Limited Sunflex, Luxaflex, and HD Direct. Hunter Douglas used the following brands at wholesale level in the UK:
Sunflex, Luxaflex, and HD Direct.
18
Figure 1: Blinds, shutters and curtains supply chain and the Parties’ business units active at
each level
Source: CMA’s summary based on Parties’ submissions
In 2019 Hunter Douglas had global revenues of just over £3.0 billion ($3.7
billion
4
) and UK revenues of £[] in 2019. In the UK, Hunter Douglas
supplies blinds online through the 247 blinds, Blinds2Go and Web Blinds
brands and websites.
5
Blinds2Go
Established in 2000, Blinds2Go is the UK’s largest online made-to-measure
(M2M) retailer for blinds. Web Blinds was formerly part of the Hillarys
Group, but was subsequently integrated into the Blinds2Go Group by
Hunter Douglas after its acquisition of Hillarys in 2017.
Blinds2Go’s total UK revenue for the financial year (FY) 2019 was £[].
Based on the data in table 1 below, Blinds2Go’s total revenue increased on
average by 45% per year between FY 2013 to FY 2019. The average rate
of growth between FY 2018 and FY 2019 has been considerably lower at
18% per year, compared to an average of 58% from FY 2013 to FY 2017.
Blinds2Go’s average gross profit has been 39% over the last 7 years.
4
See Hunter Douglas Annual Report 2019. See also Hunter Douglas’s response to CMA’s request for
information dated 21 February 2020.
5
Web Blinds is owned by Blinds2Go and Hunter Douglas has a 65% controlling interest in Blinds2Go.
19
Table 1: Blinds2Go Total Revenue and Profit margin FY 2013 to FY 2019
£ million
2013
2014
2015
2016
2017
2018
2019
Total Revenue
[
]
[
]
[
]
[
]
[
]
[
]
[
]
Gross Profit
[]
[]
[]
[]
[]
[]
[]
Gross Profit %
[
]
[
]
[
]
[
]
[
]
[
]
[
]
Figure 2 below shows Blinds2Go’s performance from FY 2008 to FY 2019.
Consistent with Table 1 above, this shows a stable gross profit margin of
between 35% and 40% over this period.
Figure 2: Blinds2Go UK sales of M2M blinds and gross and net profit margins FY 2008 – FY
2019)
[]
Source: Blinds2Go Management accounts (2013-2019) and financial reports (FY 2008 – FY 2012)
247 Home Furnishings
247 is a UK-based and online supplier of window coverings to retail
customers, including blinds, shutters and curtains. The company was
founded in 1997 by Jason Peterkin and David Maher (247 Founding
Shareholders). The company was a mail-order business for the first 8 years
while the 247 Founding Shareholders built up the web design and e-
commerce aspects before launching their in-house e-commerce project in
2005. This move also included the offer of M2M blinds online. Following the
success of their first two years’ sales of M2M blinds, with second-year
sales exceeding £3 million, in 2007 the company moved to focus
exclusively on selling M2M blinds online.
In 2019, 247’s total turnover for all products for the year ended 19 February
2019 was £22.2 million, of which £[] was in the UK.
The data in Table 2, and illustrated by Figure 3, set out 247’s total revenue
from selling M2M blinds in the UK. 247’s total revenue increased on
average by 10% from 2008 to 2019. 247’s average gross profit margin has
been around 34% over the same timeframe.
20
Table 2: 247 UK M2M Blind income and UK M2M blind gross profit margin 2008 to 2019
£ million
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
Total Revenue
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
Gross Profit
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
Gross Profit %
[
]
[
]
[
]
[
]
[
]
[
]
[
]
[
]
[
]
[
]
[
]
[
]
Source: 247 management accounts 2008 to 2019
Figure 3 : 247 UK M2M Blind income and UK M2M blind gross profit margin 2008 to 2019
[]
Source: 247 management accounts 2008 to 2019
Note: 247’s Financial year ends after February, therefore the FY year represents the figures for the year ended 28 or 29
February the year after (e.g. FY 2009 is the year ending 28 February 2010).
21
3. The Transactions
Timeline of events
Hunter Douglas acquired its interests in 247 through two separate
transactions in 2013 and 2019, respectively. In the intervening period,
Hunter Douglas made changes to its voting rights in 247, the details of
which are outlined below. Also during this period, Hunter Douglas
expanded its presence in the UK window coverings sector through
separate acquisitions of a majority stake in Blinds2Go and of Hillarys.
These are indicated in the timeline below, which shows developments in
relation to 247 below the timeline, and other Hunter Douglas developments
above the timeline:
Figure 4: A timeline of the Parties’ transactions
Source: CMA analysis of Parties’ submissions
Events leading up to the Transactions and valuations
2013 Transaction
The Parties submitted that the timeline of the events leading up to the
decision to enter into the 2013 Transaction is as follows:
6
(a) In September 2012 David Sonnenberg (then and currently co-CEO
of Hunter Douglas) initiated discussions with the 247 Founding
Shareholders concerning a possible investment by Hunter Douglas.
This was followed up by another meeting the following month
6
[].
22
between David Maher and Michiel de Heer (of Hunter Douglas) to
discuss further details of the investment.
(b) On 18 December 2012, there was a further meeting between David
Sonnenberg, David Maher and Jason Peterkin in Rotterdam to
discuss details of the investment.
(c) In early January 2013, negotiations took place leading to the
agreement of Heads of Terms for the 2013 Transaction on 10
January 2013.
(d) On 30 April 2013, the 2013 Transaction was then implemented by
way of a series of agreements; a [].
Hunter Douglas completed the 2013 Transaction by [].
7
This translates to
an Enterprise Value of £[] for 247 at the time of the 2013 Transaction.
Attached to these loan notes were certain rights in 247, which Hunter
Douglas also acquired as a result. These included:
(a) 49% of the voting rights at shareholder level and a 49% share in the
profits in 247;
(b) A right to convert the loan notes at any time to ordinary shares;
(c) A right to nominate a non-executive Director to the 247 Board; and
(d) Veto rights, notably over the following matters:
(i) Appointment of additional directors (beyond the 247 Founding
Shareholders);
(ii) Approval of the annual budget;
(iii) Acquisitions;
(iv) Entering into new lines of business other than (a) M2M window
coverings, (b) curtain-in-a-box in the UK; standard Velux roof-window
blinds, accessories associated with the above and any other items
sold by 247 on its UK website at the date of the agreement, all of
which are to be sold principally through the internet without
specifically targeting the large scale B2B market (interior designers,
property management companies and letting agents);
7
[].
23
(v) Geographic expansion;
(vi) Any backward integration into assembly or production of any of the
products sold by 247;
(vii) Long term agreements (exceeding one year in duration);
(viii) Financing arrangements with banks or other parties;
(ix) Dividends in excess of 35% of profit after tax;
(x) Offers on the website at less than 15% gross profit; and
(xi) Termination of the existing supply agreement with Hunter Douglas.
8
Notwithstanding the contractual provisions between the Parties, the Parties
have submitted that 247 has been operated as an independent business
since the time of the 2013 Transaction. The Parties stated in their main
submission that Blinds2Go and 247 have continued to compete as
independent rivals from 2016 to the present date.
9
Hunter Douglas has
said this is a matter of policy and principle; co-CEO David Sonnenberg
stated at the hearing with the CMA ‘we have a very long history, over 60
years, of conducting ourselves in the industry in this way, leaving our
companies alone, very much like the way Berkshire Hathaway manages its
subsidiaries. In practice, the only way I am able to oversee 130 companies
is by running companies independently.’
Reciprocal put and call options were granted to Hunter Douglas and the
247 Founding Shareholders (the Put and Call Options) under separate Put
and Call Option Agreements between Hunter Douglas and the 247
Founding Shareholders. Under the Put and Call Options, the 247 Founding
Shareholders could require the purchase of 100% of their shares by Hunter
Douglas and/or Hunter Douglas could require the sale of the shares held by
the 247 Founding Shareholders, under normal circumstances,
10
by written
notice, in the period 1 March to 1 June 2019.
A Stakeholders Agreement was entered into between Hunter Douglas
(through Buismetaal), 247 and the 247 Founding Shareholders. Key
provisions of the Stakeholders Agreement included granting the rights set
out at paragraph 3.4 above to Hunter Douglas.
Additionally, two identical ‘Bonus Agreements’ were set up between Hunter
Douglas and each of the 247 Founding Shareholders which paid a bonus of
8
[].
9
Main Submission, 20 May 2020, paragraph 3.23.
10
Under the bad leaver, death or disability provisions the option can be exercised outside this period.
24
[]% to the 247 Founding Shareholders for annual purchases of finished
blinds up to £[] within the Hunter Douglas Group, and []% for any
purchases exceeding £[]. This agreement was valid for the financial year
ending 28 February 2014 to the financial year ending 28 February 2017. An
‘additional bonus payment’ of []% was made on all finished blinds from
within the hunter Douglas Group. in the financial years ending 28 February
2014 and 28 February 2015.
Confidentiality provisions
The terms of the 2013 Transaction prevented either Party from publicising
the Transaction. This allowed Hunter Douglas to keep its investment in 247
confidential.
The Parties told us that the 2013 Transaction was kept confidential in order
to avoid the potential for ‘channel conflicts’ between Hunter Douglas, as a
wholesale supplier, and its customers as retail suppliers. Hunter Douglas
submits that it was concerned about the reaction of (and thus ‘conflict’ with)
its retail level customers to news that Hunter Douglas was itself entering
that market through its investment in 247. 247 submitted that it wished to
keep the 2013 Transaction confidential in order to avoid disrupting its
relationships with suppliers other than Hunter Douglas.
Other relevant Transactions in the intervening period
After the 2013 Transaction, but before the 2019 Transaction, Hunter
Douglas acquired two other businesses who were active in the supply of
window coverings.
On 21 June 2016 Hunter Douglas acquired a 60% equity stake in
Blinds2Go for £[] million
(the 2016 Transaction). Hunter Douglas
subsequently acquired a further 5% interest in Blinds2Go in 2019.
11
Hunter
Douglas has future options to acquire the remainder of the shares in
Blinds2Go; two put and call options have been built into the structure of the
deal for Hunter Douglas to acquire a further 15% equity stake in 2021 and
the remaining 20% in 2026.
12
On 17 July 2017 Hunter Douglas acquired Hillarys (the 2017 Transaction).
This acquisition was reviewed and unconditionally cleared by the CMA at
phase 1. The Merger Notice submitted by Hunter Douglas in relation to the
2017 Transaction did not disclose the existence of the 2013 Transaction.
11
Main submission, footnote 26.
12
Main submission, paragraph 6.10.
25
The only reference in the Merger Notice to Hunter Douglas having any
relationship with 247 was in a footnote, which stated that: []
13
[].’
Hunter Douglas reduction of voting rights in 247
On 6 December 2016, Hunter Douglas reduced its shareholder level voting
rights in 247 from 49% to 24.9%. Hunter Douglas submits that this arose as
a result of changes to UK corporate governance rules as part of the
implementation of the EU Fourth Money Laundering Directive, specifically
the requirement for UK companies to maintain a register of persons with
significant control (PSC). The CMA understands that reducing its voting
rights in 247 to under 25% allowed Hunter Douglas to maintain
confidentiality over its interest in 247 by avoiding having to be identified on
247’s register of PSCs. The CMA understands that this desire to continue
maintaining the confidentiality of Hunter Douglas’ interest in 247 was
motivated by the perceived risk of ‘channel conflicts’ noted above. Hunter
Douglas’ other rights set out at paragraph
3.4 above remained unchanged.
Subsequently, on 11 May 2017, just prior to the 2017 Transaction, Hunter
Douglas further reduced its shareholder level voting rights in 247 from
24.9% to 4.9%. Again, Hunter Douglas’ other rights, as per paragraph 3.4
above in 247 remained unchanged.
Hunter Douglas submits that its reason for reducing its voting rights was to
avoid having to disclose its interest to the CMA in any investigation of the
2017 Transaction. Hunter Douglas have stated that they were concerned
that the CMA might disclose the 2013 Transaction in the course of any
such investigation, leading to the transaction no longer being confidential:
The reason for this was that Hunter Douglas was aware that the
CMA might investigate its proposed acquisition of Hillarys, a
much larger business with overlaps upstream in components and
assemblers, as well as retail operations including a small online
web brand (Web Blinds) and its main in-home service. Hunter
Douglas was concerned that any such review might lead to the
disclosure in the decision of its interest in 247. This would have
undermined the position so recently taken to reduce its voting
rights from 49% to 24.9% to preserve confidentiality.
13
[].
26
2019 Transaction
The Parties submit that the timeline of the events leading up to the decision
to enter into the 2019 Transaction is as follows:
(a) In December 2017 the 247 Founding Shareholders met with Hunter
Douglas to discuss continuation plans if the put or call options were
exercised.
(b) In 2018 the 247 Founding Shareholders made it known to Hunter
Douglas that they would be exercising their respective put options.
(c) In 2018/2019 a number of calls concerning the optimum deal
structure occurred between the 247 Founding Shareholders and
Hunter Douglas.
(d) On 19 February 2019 the 2019 Transaction was implemented. The
2019 Transaction took place slightly earlier than anticipated in 2019
(as stated above, it was originally planned to take place between
March and June 2019) to ensure it came before the end of 247’s
financial year.
(e) The 2019 Transaction completed on 28 February 2019, at the end of
247’s 2018/19 financial year. The terms of the 2019 Transaction
gave effect to the Put and Call Options, albeit these were not
formally exercised.
The total consideration paid for all the issued share capital of 247 was
£[]. This is consistent with the purchase price that would have been paid
had the Put or Call Option been exercised, the calculation of which was to
be based upon 247’s profits in the two years prior to the exercise of the Put
or Call Option.
The Parties’ rationale for the Transactions
The Parties submit that the 247 Founding Shareholders ‘entered the 2013
Transaction in the belief that they were selling the business to Hunter
Douglas in a staggered transaction that would finally complete in 2019 and
which would ensure the most tax efficient outcome (through entrepreneur
relief) for the individuals.’
As a result, the Parties were not of the view that the 2013 and 2019
transactions were independent. At the main party hearing with 247, Jason
Peterkin stated of his and David Maher’s understanding:
27
David never considered the 2013 and 2019 transactions to be
distinct from each other. We looked at the 2013 as a single
transaction which ended in 2019... David and myself to a lesser
extent were both very keen to give ourselves an exit plan. The
deal that Hunter Douglas offered us gave us exactly that. It also
gave us access, certainly perceived access, to their resource.
247 has submitted that ‘the structure provided a guaranteed exit route.
David Maher, in particular, was looking to retire and the put option allowed
both of the 247 Founding Shareholders to exit the business whilst
maximising the value which could be achieved.’
Notwithstanding the stated exit plan for the 247 Founding Shareholders,
the Parties’ submitted the following further considerations motivating the
Transactions.
2013 Transaction
Hunter Douglas submitted the following in relation to the 2013 Transaction:
In 2013, we were aware of the emerging importance of e-com to
our business. Home Depot had bought Blinds.com and we were
supplying them, so we could see the growth. We began to worry
that over the long run we may need to transform into more of a
retailer ourselves, as difficult as that might be given the efficiency
of the direct-to-consumer businesses in e-com, in particular,
versus traditional retail. UK e-com seemed to be emerging and a
good place to learn.
Hunter Douglas [], describing, in a Board report at the time of the
Blinds2go acquisition, that ‘[]’.
The Parties have submitted that another motivation for the 2013
Transaction was to support 247’s expansion overseas,
14
more specifically
across mainland Europe.
15
Hunter Douglas stated, in internal documents at
the time of the transaction, that:
‘[247] requires a reliable strategic supplier with wide geographic
reach and product breadth as well as additional capital for
geographic expansion on the European Continent.
14
247 submitted that the 247 Founding shareholder were considering expansion plans at the time of the
2013.transaction; an extension into flooring products, or international expansion.
15
[].
28
HD is willing to acquire a 49% stake in [247] to support this
expansion.
The structure of the existing cooperation as well as the expansion
to Continental Europe should be optimized to reflect the fiscal
position of the parties, e.g. to allow for Entrepreneurial Relief for
David and Jason, where possible earning income in a permanent
establishment in a jurisdiction with lower tax rate than the U.K.,
licensing structures for existing and future IP rights, gearing,
etc.
16
’
The Strategic Business Plan prepared as part of the 2013 Transaction
states that [].
17
As a result, a Bonus Agreement was entered into with
247’s Founding Shareholders.
2019 Transaction
As described above, the 2019 Transaction was at the instigation of the 247
Founding Shareholders, and in particular the retirement of David Maher.
The 247 Founding Shareholders discussed potential continuation plans
after the call and put options with David Sonnenberg as early as 2017
before discussing their intention to exercise their put option in 2018 and exit
from the business.
[].
18
[].
16
[].
17
[].
18
[].
29
4. Industry background
In this section we provide an overview of the industry in which the Parties
are active. In particular, we will discuss the wider window coverings sector
in the UK, with a focus on the retail market for online M2M blinds, being the
principal area of overlap between the Parties.
Window coverings overview
The window coverings sector includes blinds, curtains, shutters, and
suspension systems. Customers typically purchase products in the sector
in ready-made format (ie a product that is largely finished and available in
one of many available sizes) or M2M (ie tailored to the exact specifications
of the customer). As discussed below, customers in the UK primarily
purchase window coverings through in-home, in-store, and online retail
channels.
Market research reports received estimate that in 2018 the size of the UK
window coverings sector was approximately £1.5 billion. AMA Research,
Domestic window coverings market report UK 2020-2024 (the AMA report),
estimated the sector to be approximately £1.5 billion in size and
GlobalData, Window dressings November 2018 (the GlobalData Report),
estimates the sector size at £1.6 billion.
19
This includes sales for all
channels; online, in-store and in-home. Global Data estimate the total
growth of the sector as a whole between 2018 and 2023 is expected to be
9 to 11%, an annual average growth of approximately 2%.
20
Window coverings fall in the broader category of home accessories, and
account for the largest proportion of consumer spending within the
category. According to a Mintel report, ‘Accessorising the home – March
2020’, almost half of consumer spending on home accessories in the UK is
on window coverings. Figure 1 below shows that M2M window coverings
(curtains and blinds) is the single largest sub-category within home
accessories, followed by ready-made window coverings (curtains and
blinds).
21
19
We note these reports are both now two years old and so their estimates of market size and growth should be
treated with some caution. We note that we do not hold industry reports that contain market data for online M2M
blinds specifically. The industry reports quoted in this paper draw their findings from consumer research,
including consumer surveys, to inform their forecasts data for the wider sector.
20
Global Data estimates the window coverings sector will grow 9.2% from 2018-2023, Global Data Report, Page
8 and AMA research estimate the market to grow 6% from 2020-2042, AMA report, page 10.
21
Mintel, Accessorising the home – March 2020, p24.
30
Figure 5: Consumer spending on accessorising the home, % of total 2019, UK
Source: Mintel Report
22
As a category of home accessories, the performance of the window
coverings sector is driven to some extent by the performance of the
housing market. AMA Report described growth of the sector as ‘relatively
consistent’ in the period 2012 to 2017 with year on year growth. See Figure
2 below taken from the AMA Report.
22
Mintel, Accessorising the home – March 2020, p24.
31
Figure 6: UK Window Coverings Market and Forecasts 2015 to 2024 by value (£m RSP)
Source: AMA Research Ltd/Trade Estimates
23
Prior to the COVID-19 pandemic, the window coverings sector had seen
year-on-year growth for the last several years. A previous AMA Report
stated that ‘better performance in the housing market and wider economy
has fuelled demand for both replacement and new window coverings.’
24
However, the GlobalData Report states that ‘2018 has seen the window
dressings category grow by just 0.8%, with a slow housing market and
cautiousness about spending large amounts on homes…with consumers
nervously following Brexit negotiations.’ As a result, ‘the category has been
reliant on sales to “improve not movers.”’
The current COVID-19 pandemic has affected retailers and suppliers in the
sector. The AMA report says ‘whilst indications at the beginning of the year
tended towards underlying but modest growth rates into the medium-term,
the events of March and April have turned this forecast on its head. The
revised forecast for 2020 shows decline of around 9% in 2020 followed by
annual growth rates of 3-5%.’
25
However these trends are for the window
coverings sector as a whole and not specific to the in-home, in-store, or
online retail channels.
[].
26
[].
27
However, there has been some knock-on effect for customers. [].
28
23
AMA Research, Domestic window coverings market report UK 2020-2024, page 10.
24
AMA Research, Domestic window coverings market report UK 2018-2022, page 8.
25
AMA Research, Domestic window coverings market report UK 2020-2024, page 12.
26
[].
27
[].
28
[].
32
In Figure 7 below, the GlobalData Report charts the forecast relative
growth and size of each window covering type in the UK across distribution
channels.
29
This indicates that blinds (ready-made and M2M) are estimated
to be the highest growth products, with total forecast growth rates of 10.7%
and 9.7% respectively over the period 2018 to 2020 in the UK. M2M blinds
is the largest category at £518.6m, roughly one third of the total window
coverings sector and almost 30% larger than the market for M2M
curtains.
30
Figure 7: Forecast category growth and size: 2018-20, UK
Source: GlobalData, Window Dressings November 2018, p16
Distribution Channels
Distribution within the window coverings sector is highly diverse, with
options available in-store, in-home and online for the relevant products.
These are described as ‘channels’ within the industry. Each channel
presents customers with a different range of potential purchasing options,
as well as service and installation propositions. We discuss the main
features of each below.
In-store and In-home
The in-store and in-home channels are traditional retail channels in which
customers have a degree of interaction with a salesperson and/or the
products before purchase. These interactions take place either inside the
29
GlobalData, Window dressings November 2018, p16.
30
GlobalData estimates the M2M curtains market to be £371.5m in 2018.
33
retailer’s physical store (in-store) or inside the customer’s home (in-home).
Both ready-made and M2M blinds are sold through these channels.
As set out in paragraph 7.57, in-store and in-home retailers appear to
position themselves as full-service providers, offering a more personal
experience and the option of a measuring and installation service. The
evidence we have seen also indicates that in-store and in-home retailers
tend to be more expensive than online retailers.
31
Retailers in the in-store channel are able to display a variety of window
coverings as well as their range of fabric options, although they are limited
by the available space. As a result, customers can review the colour, style
and materials for their window coverings, whether ready-made or M2M, in
person. Often, there will be a window covering specialist to assist a
customer to design their desired window covering.
As part of a retailer’s in-home offering, a consultant visits a customer’s
home with a limited range of samples in order to review colour, style and
materials for their window coverings, and may offer style advice as part of
the consultation. The consultant may also take measurements for the
customer, and fit the final product once complete.
Some in-store and in-home retailers also have an online presence,
although not all such retailers sell blinds online. Retailers who sell blinds
online and in-store are referred to as ‘multi-channel’ retailers. We note that
no supplier active in the in-store and in-home channel has a significant
position within the online M2M blinds market, and, for some, their online
presence is part of a general online store (ie covering products and
services unrelated to blinds or window coverings).
Online
Customers are able to purchase both ready-made and M2M window
coverings online. However, measuring and fitting typically are not provided
by retailers with an exclusively online presence. Customers select colour,
style and materials for their window covering online using retailer-supplied
photographs. Many online retailers, including the Parties, offer a sample
service where fabric swabs can be requested and posted out to the
customer ahead of final product selection.
Online comparison of prices is a key feature of the online channel for
window coverings, including blinds. The GlobalData report states that
64.2% of shoppers start their research online ‘with customers consulting
31
See paragraph 7.57.
34
reviews, comparing prices and scrutinising pictures and videos of the
product they are interested in’.
32
The large online offering of products
allows customers to easily compare product offers compared to in-store.
33
[].
34
The AMA Report states that ‘window coverings are… widely available
online and consumers are… much more confident in ordering these
products themselves.’
35
Hunter Douglas submitted that a key recent
market development is the growing expansion of online sales of blinds, in
particular of M2M blinds.
36
We note, as discussed at 4.7, that the COVID-
19 pandemic has indirectly resulted in a significant increase in online sales
volume.
The growth of online spend has also seen a rise in mobile purchasing, with
smartphones predicted to become even more central to the purchasing
journey within the next two years.
37
This is in line with the evidence
provided by the Parties, that increased sales through mobile sites has led
to emphasis placed on developing the online mobile sites. [].
Figure 8: 247 Customer Journeys – Devices used March 2013 to April 2020
[
]
The COVID-19 pandemic led to closure of physical retail stores and
suspension of in-home sales for a period of time, therefore demand has
increasingly been redirected online. Mintel reports, this ‘should further
catalyse the channel, and favour well-equipped multi-channel and online
only retailers. With current government guidance to self-isolate within the
home... this could actually boost the market over this period, as people
spend more time inside.’
38
Online M2M blinds
As noted above, the main area of overlap between the Parties is the supply
of online M2M blinds. Online M2M blinds are tailored blinds purchased by
customers online through a retailer’s website. Customers can browse the
website for various styles of blind (eg roman, venetian) in a variety of
materials and colours, and usually are able to order a free sample of a
given blind. Having selected their desired blind, customers provide
32
GlobalData, Window dressings November 2018, p6.
33
Hilary’s brand report.
34
[].
35
AMA Research, Domestic window coverings market report UK 2020-2024, p13.
36
[].
37
Mintel Report, page 15.
38
Mintel Report, Accessorising the home – March 2020, page 25.
35
dimensions, measured in width and drop.
39
Customers may also be able to
make other customisations, for example by selecting the width of the slats
on a venetian or wooden blind.
On receipt of an order, online M2M retailers source the blind to fulfil the
order and arrange delivery. Some retailers, including both of the Parties,
arrange for delivery directly from manufacturers to customers and do not
hold their own inventory.
40
Once the customer has received the order, the customer fits the blinds
themselves by either following instructions provided by the retailer, or by
arranging for someone else to fit the blinds independently.
The largest online suppliers of M2M blinds sell exclusively online and do
not have an in-store or in-home presence. As mentioned above, some
multi-channel retailers do supply M2M blinds online, however, their product
offering is to-date more limited and their share of the online market is
modest.
Online platforms
In addition to retailers’ websites, Amazon and eBay are online marketplace
platforms that include ready-made window coverings from multiple sellers
in their listings.
41
However, third party retailers have not previously had an
effective way of selling M2M blinds on these platforms.
Customers can either search listings on the online marketplace or browse
by type of blind (eg venetian, or roman).
42
There are various ways listings
allow tailoring – either by requiring that the customer contacts the seller
separately, or (on Amazon) by using the recently-added ‘customise now’
button. Even with these functions, they are not configured particularly
effectively to enable the sale of M2M blinds.
43
Amazon submits that the
‘customise now’ button is a very new feature on Amazon’s website.As a
result, we are aware that currently use of this function is limited.
While some listings on online marketplaces do offer free sample swatches,
we understand that [].
44
As with standard online M2M blinds, once an
39
Instructions on how to take these measurements are usually available on online M2M retailers’ websites.
40
[].
41
We note that in addition to hosting third-party sellers, Amazon sells its own ready-made blinds directly on its
platform, but not M2M blinds.
42
Although the filtering options on online marketplaces are much narrower than on most online M2M blinds
websites.
43
There is no way of filtering to only show listings which use the ‘customise now’ feature.
44
[].
36
order is made online M2M retailers source the blind to fulfil the order and
arrange delivery.
45
Both the Parties
46
have highlighted Amazon and eBay as growing
platforms, [].
47
Swift Direct Blinds submits that ‘a small (currently approx.
[]%) and growing portion of its sales are generated through eBay and
Amazon’.
48
Competitors and main parties
In the window coverings sector more generally (including in-store and in-
home supply of ready-made and M2M blinds, as well as other types of
window coverings) multi-channel retailers are the leading suppliers, with
Dunelm, Hillarys, John Lewis and Next being the largest retailers.
Describing the retailers who supply window coverings, the 2018 GlobalData
Report states that ‘Dunelm remains the clear leader in the category, with its
market share (17.4%) continuing to dwarf that of second largest player
Hillarys (7.0%)’. These retailers offer both ready-made and M2M window
coverings, although their presence varies by distribution channel.
However, the competitive landscape is different for the online M2M blinds
market, being the principal area of overlap between the Parties.
The main competitors for the online retail supply of M2M blinds are online-
only retailers with none of the in-store or in-home retailers occupying a
significant position in the market. The Parties are two of the largest retail
suppliers M2M blinds online. Hunter Douglas, through its majority owned
subsidiary Blinds2Go, is the largest supplier in the market with 247 being
the third largest. In addition to the Parties, there are a handful of retailers
with revenues over £5 million: these are Interior Goods Direct (which now
includes the Wilsons Online Retail business following a recent acquisition),
Swift Direct Blinds and Bloc Blinds. In addition to these retailers, there is a
long tail of smaller retailers, none of whom has revenues above £5 million.
The Parties’ share of supply in the market for online M2M blinds is further
analysed in the Competitive Assessment section below.
Some multi-channel retailers do have a presence in the online M2M blinds
market (whether click-to-order or otherwise), for example Next and John
45
Once the customer has received the order, the customer fits the blinds themselves by either following
instructions provided by the retailer, or by arranging for someone else to fit the blinds independently.
46
[].
47
Main submission, 24 May 2020, Paragraph 1.11.
48
[].
37
Lewis. At the same time, some multi-channel retailers do not have a full-
function online M2M offering.
The multi-channel retailer Dunelm submitted that its website does not
currently offer click-to order functionality for M2M blinds, however, it has
introduced an online consultation service.
49
Whilst a response to the
current restrictions on in-story shopping, it provides a new level of customer
service to their current online offering and enables the firm to take orders
online without the specialised website functionality. Dunelm describes the
process as follows,
A customer makes an appointment with a Dunelm consultant, and this
is carried out on Microsoft Teams. The consultant will then discuss the
customer’s needs and suggest a solution to them. The consultant will
then assist the customer to place an order. The customer cannot place
an order directly online as yet without the assistance of the consultant,
this requires some Tech development.
50
Further, in-home retailers Hillarys’ and Thomas Sanderson’s websites only
provide lead generation and enable customers to provide contact details
and request order samples.
51
The degree of competitive constraint exerted
on the Parties by offline retailers is considered in the Competitive
Assessment section below.
Trends within the Sector
Improvements in the customer offering
The Parties have commented that M2M blinds are increasingly
competitively priced against ready-made blinds online as well as M2M
blinds in-store.
52
The Parties have said that effective supply-chain
management has enabled these lower prices to be offered and this has
been a key factor in the growth of the online M2M blinds market.
53
Additionally, online M2M blind specialist websites such as Blinds2Go and
247 blinds offer a large range of products, without the restrictions of
physical space occupied in-store, at various price points with short delivery
times. For example, 247 submitted that their blinds are delivered in 5-7
49
[].
50
[].
51
Hillarys and Thomas Sanderson websites, accessed on 23 June 2020.
52
Main submission, 20 May 2020, paragraph 6.72 and Figures 6.15 and 6.16.
53
[].
38
working days,
54
as compared to four weeks for Dunelm’s M2M service.
55
It
is noted that there are some instances where delivery times could be
shortened by in-store retailers offering a click and collect service.
Nonetheless, the wide product offering online, combined with shorter
timescales attracts those customers who may previously have shopped in-
store for high quality blinds, as well as customers who might have opted for
cheaper ready-made blinds.
Furthermore, retailers of online M2M blinds are increasingly improving the
visual appearance of their websites. While CGI is already common across
many websites, Decora believe that visualisation software will become a
key part of the online offering by retailers. Third parties have indicated that
website quality and appearance matter to customers, we note that this
could affect a retailer's position on Google search results. This is discussed
further in the Competitive Assessment chapter.
Technology
Retailers are increasingly seeing the incorporation of smart technology
within their retail offering. AMA Research observes that, ‘The market has
always been innovative, and suppliers have been quick to adapt to
changing demand patterns for more advanced products. In addition, the
ability to continue to add value – eg more automated ranges – is likely to
underpin future demand levels’.
56
[].
57
However, no third party who we
spoke to identified motorisation as an important trend.
54
[].
55
Dunelm website.
56
AMA report, page 16.
57
[].
39
5. Relevant merger situation
This section sets out our conclusion on the First Statutory Question (see
paragraph 1.3 above), namely whether each of the 2013 Transaction and
the 2019 Transaction constitutes the creation of an RMS. This is known as
the jurisdictional test.
Relevant Merger Situation – the jurisdictional test
Under section 23 of the Act, an RMS arises when the following conditions
are met:
(a) two or more enterprises have ceased to be distinct enterprises at a time
or in circumstances falling within section 24
58
of the Act (enterprises have
ceased to be distinct); and
(b) one (or both) of the following conditions is (or are) satisfied:
(i) as a result, a share of supply of 25 per cent or more is created or
enhanced in respect of goods or services of any description which are
supplied in the UK, or a substantial part of the UK (the share of supply
test);
(ii) the value of the turnover in the UK of the enterprise being taken over
exceeds £70 million (the turnover test);
This second element establishes sufficient connection with the UK on a
turnover or share of supply basis.
On the basis of the turnover figures set out at paragraphs 2.5 and 2.8
above, we conclude that the turnover test is not met in this case. In this
section we first consider whether enterprises have ceased to be distinct
and, second, whether the share of supply test is met.
58
Section 24(1) of the Act notes that ‘For the purposes of section 23 two or more enterprises have ceased to be
distinct enterprises at a time or in circumstances falling within this section if…(a) the two or more enterprises
ceased to be distinct enterprises before the day on which the reference relating to them is to be made and did so
not more than four months before that day; or… (b) notice of material facts about the arrangements or
transactions under or in consequence of which the enterprises have ceased to be distinct enterprises has not
been given…’
40
Enterprises ceasing to be distinct
Enterprises
The first element of the jurisdictional test considers whether two or more
enterprises have ceased to be distinct as a result of each of the
Transactions.
The Act defines an ‘enterprise’ as ‘the activities or part of the activities of a
business’. A ‘business’ is defined as including ‘a professional practice and
includes any other undertaking which is carried on for gain or reward or
which is an undertaking in the course of which goods or services are
supplied otherwise than free of charge’.
59
The Parties are both active in the online retail supply of different types of
blinds, shutters and curtains. Hunter Douglas is also active at the
manufacturing and wholesale levels of the supply chain for different types
of window furnishings, including assembled blinds, raw materials and
components for blinds.
60
247 is only present at the retail level of the supply
chain. As noted above, in 2019 Hunter Douglas had global revenues of just
over £3 billion and UK revenues of £[] and in 2019, 247’s total turnover
for the period ended 19 February 2019 was £22.2 million, of which £[]
was in the UK.
We are therefore satisfied that Hunter Douglas and 247 are each a
‘business’ within the meaning of the Act and that, accordingly, the activities
of Hunter Douglas and 247 are ‘enterprises’ for the purposes of the Act.
Ceasing to be Distinct
Section 26 of the Act explains the concept of ‘ceasing to be distinct’. Two
enterprises cease to be distinct once they are brought under common
ownership or common control. Control includes situations falling short of
outright voting control, including the ability directly or indirectly to control or
materially to influence the policy of an enterprise, pursuant to section 26(3)
of the Act. Three levels of interest are therefore recognised as being
sufficient to amount to an RMS: a controlling interest; the ability to control
policy (de facto control); and the ability materially to influence policy
59
Sections 129(1) and (3) of the Act.
60
At a wholesale level, some wholesalers and retailers purchase or import fully-assembled blinds, while others
purchase components and materials and fabricate the assembled blinds, using machinery or by hand. Most
wholesalers of blinds are able to provide a range of different blind types, although there are some that focus on
specific products.
41
(material influence). The ability to exercise material influence is the lowest
level of control that may give rise to an RMS.
61
This interpretation is confirmed by the explanatory Notes to section 26 of
the Act, which state that (emphasis added):
Subsections (3) …envisage three levels of control of an
enterprise. These are: material influence over policy; control of
policy (often called de facto control); and a controlling interest in
the enterprise (often called de jure control). What constitutes
material influence or control will be considered on a case-by-
case basis by the competition authorities according to the
particular circumstances of the case. Under the FTA
62
the
authorities have treated the acquisition of the ability to appoint a
director or having a 15% shareholding as sufficient to give
material influence for these purposes. De facto and de jure
control will arise at higher levels of shareholding, with de jure
normally requiring more than 50% of the voting rights
The CMA’s view, informed by previous decisional practice, of the distinction
between these levels of control is set out in CMA2 at paragraphs 4.12 to
4.30.
Section 26(4) of the Act allows for a new RMS to be created if the acquiring
firm, which is already able to exert material influence over the policy of a
target firm, acquires ‘de facto’ control or a controlling interest in the target
firm. The same applies to a move from ‘de facto’ control to a controlling
interest.
A key point of difference between the CMA and the Parties concerns the
level of interest acquired by Hunter Douglas in 247 as a result of the 2013
Transaction. The Parties do not dispute that Hunter Douglas and 247
ceased to be distinct as a result of the 2013 Transaction. However, they
claim that the 2013 Transaction should be considered as conferring a
controlling interest in 247 to Hunter Douglas or, at least, ‘de facto’ control,
which the Parties claim should be treated as a controlling interest.
63
On that
basis, the Parties claim that there could have been no new RMS as a result
of the 2019 Transaction under section 26(4)
64
of the Act as the level of
interest did not change.
61
Mergers: Guidance on the CMA’s jurisdiction and procedure (CMA2), January 2014, para. 4.14.
62
This refers to the Fair Trading Act 1973, the predecessor to the Act.
63
Response to Issues Letter, paras 5.19-5.20 and Supplementary Submission, paras 2.5-2.12.
64
This provision states that ‘For the purposes of subsection (1), in so far as it relates to bringing two or more
enterprises under common control, a person or group of persons may be treated as bringing an enterprise under
his or their control if…(a) being already able to control or materially to influence the policy of the person carrying
42
However, in the SLC Decision the CMA found that:
(a) the CMA has no ability to treat one level of interest as another for the
purposes of its jurisdictional assessment;
65
(b) the 2013 Transaction could only have given Hunter Douglas the ability to
materially influence the policy of 247
66
; and
(c) therefore, the 2019 Transaction, by increasing the level of interest Hunter
Douglas held in 247 from material influence to a controlling interest, did
amount to a new RMS.
67
The Parties’ submissions and our conclusions on these points are set out
below.
The CMA’s ability to treat one level of interest as another
The Parties submit that the CMA may treat one level of interest as another
for the purposes of the CMA’s jurisdictional assessment. Specifically, they
submit that:
(a) the CMA may treat ‘de facto’ control as a controlling interest under section
26(3) of the Act; and
(b) paragraph 4.29 of the CMA Guidance makes clear that the CMA’s
practice is to treat de facto control as a controlling interest whenever it
considers that the test for reference is met.
The CMA disagrees with the Parties’ interpretation of this section of the Act
and the relevant CMA Guidance.
As a preliminary matter, we note that the purpose of section 26 of the Act
more generally is to specify the circumstances in which enterprises may
cease to be distinct. Sections 26(1) and 26(2) state that enterprises cease
to be distinct ‘if they are brought under common ownership or common
control’ as well as the circumstances in which they may be treated as being
under common control.
68
In this context, the purpose of section 26(3) is to
specify the circumstances in which the CMA may find that sufficient control
had been acquired to conclude that enterprises have ceased to be distinct
on the enterprise, that person or group of persons acquires a controlling interest in the enterprise or, in the case
of an enterprise carried on by a body corporate, acquires a controlling interest in that body corporate; or (b) being
already able materially to influence the policy of the person carrying on the enterprise, that person or group of
persons becomes able to control that policy.’
65
SLC Decision, paragraph 45.
66
Paras 46-47.
67
SLC Decision, paragraph 49-50.
68
Sections 26(1) and 26(2) of the Act, respectively.
43
for the purposes of Sections 26(1) and 26(2).
69
This is very clearly set out
in the text of that section which notes that (emphasis added):
A person or group of persons able, directly or indirectly, to control
or materially to influence the policy of a body corporate, or the
policy of any person in carrying on an enterprise but without
having a controlling interest in that body corporate or in that
enterprise, may, for the purposes of subsections (1) and (2),
be treated as having control of it.
Thus section 26(3) of the Act makes clear that each of material influence,
‘de facto’ control or a controlling interest will be sufficient to find that
enterprises have ceased to be distinct in the circumstances specified in the
section. In light of this, we find that it does not, as suggested by the Parties,
give the CMA the discretion to treat one level of interest or control as
another for the purposes of its jurisdictional assessment.
Further, we note that paragraph 4.29 of the CMA Guidance discusses the
circumstances in which de facto control may be treated as a controlling
interest. Crucially, however, this paragraph of the guidance cites section
26(3) of the Act as its basis and must therefore be interpreted consistently
with that provision. It states (emphasis added):
The CMA has the ability under section 26(3) of the Act to decide
whether or not to treat ‘de facto’ control as a controlling interest
for the purposes of the Act; but, as explained in the context of
material influence…, its practice in the context of Phase 1
decisions is to do so whenever it considers that the test for
reference would be met in the case in question.
To be consistent with section 26(3) of the Act, which is determinative as to
the law on this point, the reference in this paragraph of the CMA Guidance
to a ‘controlling interest’ must be read as a reference to the level of interest
that is sufficient to find that enterprises have ceased to be distinct, not to
any discretion on the part of the CMA that otherwise does not exist for this
section of the Act.
The 2013 Transaction
As noted above, through the 2013 Transaction, Hunter Douglas acquired
convertible loan notes in 247. Attached to these loan notes were certain
69
Reference 26(3).
44
rights in 247, which Hunter Douglas also acquired as a result. A summary
of the main rights is set out at paragraph 3.4 above.
As noted above, the Parties have submitted that Hunter Douglas acquired
a controlling interest or, at least, ‘de facto’ control over 247 through the
2013 Transaction. By definition, if a party does not have ‘de facto’ control it
will not have a controlling interest.
CMA2 states that ‘arrangements may give rise to a position of ‘de facto’
control when an entity controls a company’s policy, notwithstanding that it
holds less than the majority of voting rights in the target company (that is, it
does not have a controlling interest).’
70
Hunter Douglas did not acquire a controlling interest
We understand that Hunter Douglas did not acquire the majority of voting
rights, either at shareholder or board level, in 247 through the 2013
Transaction. It only acquired 49% of voting rights at shareholder level and
the right to appoint a single director, which we understand it never
exercised. Even if it had, this would not have given Hunter Douglas a
majority at board level as the 247 Founding Shareholders, who each held
half the remaining 51% of voting rights, both had seats on the board.
Therefore, we conclude that Hunter Douglas did not acquire a controlling
interest over 247 through the 2013 Transaction.
In their submissions the Parties note that paragraph 4.30 of CMA2, which
states that 'A ‘controlling interest’ generally means a shareholding
conferring more than 50% of the voting rights in a company’, means that
‘there is no requirement to have more than 50% of the voting rights in a
company for a controlling interest.’ The caveat in this section of the
guidance is intended to allow for exceptional circumstances where the legal
rights acquired by an entity grant an equivalent level of control to having
the majority of the voting rights in a company. This is why that paragraph of
CMA2 also states that ‘Only one shareholder can have a controlling
interest.’ This is not the case here. While Hunter Douglas did acquire
various rights, including those set out at paragraph 3.4 above, in 247
through the 2013 Transaction, our conclusion is that none of these
individually or collectively gave Hunter Douglas the same legal level of
control it would have had from having the majority of the voting rights in
247 as they did not give Hunter Douglas the ability to exercise more than
50% of the voting rights at board or shareholder level.
70
Para 4.28.
45
Hunter Douglas did not acquire ‘de facto’ control
Indeed, as per the finding in the SLC Decision,
71
in our view the rights
acquired by Hunter Douglas through the 2013 Transaction were not even
sufficient to give it ‘de facto’ control over 247. CMA2 notes that the exercise
of ‘de facto’ control requires the ability to control i.e. unilaterally determine
(as opposed to just materially influence) the acquired company’s policy.
72
In this context, ‘policy’ means ‘the management of …[the acquired]
business, and thus includes the strategic direction of a company and its
ability to define and achieve its commercial objectives.’
73
As noted in the
SLC Decision, examples where this might occur include where the
shareholder has, in practice, control over more than half of the votes cast at
a shareholder meeting or where an investor’s industry expertise leads to its
advice being followed to a greater extent than its shareholding would seem
to warrant.
74
In submitting that Hunter Douglas acquired at least ‘de facto’
control through the 2013 Transaction, the Parties highlighted the following
factors:
(a) The rights set out at paragraph 3.4 above, including the veto rights and
the fact that Hunter Douglas had the largest share of the voting rights
75
;
(b) the relative imbalance in size and influence between the position of
Hunter Douglas and that of the 247 Founding Shareholders;
(c) the disparity in experience and resources between Hunter Douglas and
the 247 business;
(d) the fact that the 247 Founding Shareholders sought guidance from senior
executives at Hunter Douglas on occasion;
(e) the existence of the Put and Call Options;
(f) the bonus arrangements for the 247 Founding Shareholders set out in the
Bonus Agreement; and
(g) the fact that 247 was required, from 2013, to attend the same
management review meetings as Hunter Douglas’ wholly owned
subsidiaries
71
SLC Decision, paragraph 46.
72
Para 4.28.
73
CMA2, para 4.14.
74
SLC Decision, paragraph 42 and CMA2, para 4.28.
75
The founding shareholders of 247 Home Furnishings each held 25.5% voting rights after the 2013.
Transaction.
46
However, in our view none of these factors, alone or together, led to Hunter
Douglas having the ability to unilaterally determine 247’s policy. Taking
each factor in turn:
(a) While Hunter Douglas did acquire extensive rights (including notably
those set out at paragraph 3.4 above) in 247 through the 2013
Transaction, the 247 Founding Shareholders benefitted from the same
veto rights set out at paragraph 3.4 as Hunter Douglas.
76
These
equivalent veto rights mean that Hunter Douglas did not have the ability to
unilaterally determine, and thus control, 247’s policy on any issue as the
247 Founding Shareholders possessed at least as much ability to
determine such policy. In fact, as noted above, it was the 247 Founding
Shareholders, not Hunter Douglas, who held the majority of voting rights
at both board and shareholder level;
77
(b) The above position does not seem to have been affected by any
imbalance in size and influence between the position of Hunter Douglas
and the 247 Founding Shareholders, any disparity in experience and
resources between Hunter Douglas and the 247 Founding Shareholders
or by the fact that the 247 Founding Shareholders occasionally sought
guidance from senior executives at Hunter Douglas. There does not
appear to be any evidence that any of these factors would have enabled
Hunter Douglas to direct the behaviour of the 247 Founding Shareholders
and given, as noted above, that the 247 Founding Shareholders held the
same veto rights set out at paragraph 3.4 as Hunter Douglas and greater
voting rights at board and shareholder level, these factors would not have
given Hunter Douglas the ability to determine the policy of 247 without the
cooperation of the 247 Founding Shareholders.
78
At most, in our view,
these factors could have assisted Hunter Douglas in influencing 247’s
policy;
(c) In relation to the Put and Call Options, as far as we are aware, they could
be exercised unilaterally by the 247 Founding Shareholders and Hunter
Douglas alike and thus seem unlikely to have provided much leverage
over the behaviour of the 247 Founding Shareholders. The anticipated
acquisition of the remaining shares in 247 in 2019 seems likely to have
76
See Stakeholders’ Agreement, 30 April 2013, Section 5, para 2: ‘All decisions in respect of the following
matters [consisting of the matters set out in para 26 above] must be made by unanimous decision of the
Shareholders (for clarification including HD, if HD is at the time a noteholder)’.
77
We have seen no evidence to the contrary. Indeed, we understand that Hunter Douglas never actually
exercised its voting rights at shareholder level or appointed a director at board level (see Hunter Douglas’s
response to the s109 Notice dated 14 May 2020, para 3.1).
78
We have seen no evidence to the contrary. Indeed, we have seen no evidence that Hunter Douglas ever
sought to use its size, influence, experience, resources and/or guidance to actually try and control the policy of
247.
47
only incentivised the 247 Founding Shareholders to maximise the profits
of 247, rather than necessarily following the direction of Hunter Douglas
79
;
(d) As noted by the Parties, the effect of the Bonus Agreements, described at
paragraph 3.8 above, was that the 247 Founding Shareholders had [].
While the Bonus Agreements may have been capable of influencing the
behaviour of the 247 Founding Shareholders in relation to the purchase of
finished blinds by 247, we do not consider that these arrangements would
have resulted in the 247 Founding Shareholders necessarily following the
direction of Hunter Douglas in relation to their policy.
80
In any event, as
noted at paragraph 3.8 above, the Bonus Agreements were only valid
between the financial years ending 28 February 2014 and 28 February
2017 and so any incentive effect ceased as of that latter date. As noted at
paragraph 5.27(b) above, Hunter Douglas did not have the ability to
determine the policy of 247 without the cooperation of the 247 Founding
Shareholders; and
(e) Finally, the fact that 247 was required to attend the same management
review meetings as Hunter Douglas’ 100% subsidiaries does not imply
that Hunter Douglas must have had the same level of control over 247.
While these meetings, again, may have influenced the policy of 247 we
are not convinced that their attendance evidences that Hunter Douglas
was, in fact, determining their policy.
Hunter Douglas did acquire material influence
We conclude that the rights set out at paragraph 3.4 and the other factors
at paragraph 5.27 , whilst falling short of giving Hunter Douglas the ability
to unilaterally determine 247’s policy, are sufficient to show that Hunter
Douglas acquired the ability to exercise material influence over 247’s policy
as a result of the 2013 Transaction, for the reasons set out in the SLC
Decision. This is not now disputed by the Parties. As noted in the SLC
Decision, the 49% of the voting (and economic) rights in 247 acquired as a
result of the 2013 Transaction is well above the 25% threshold for
presuming the existence of material influence.
81
In addition, as noted at
paragraph 3.4 above, Hunter Douglas acquired the right to nominate a non-
executive director and veto rights covering many aspects of 247’s strategic
decisions (including the appointment of additional senior management,
annual budgets, any financing, expansions into new lines of business and
pricing offers below 15% gross profit). Even though the 247 Founding
79
We have seen no evidence to the contrary. Indeed, we have seen no evidence that Hunter Douglas ever
sought to use the existence of these options to influence, let alone control, the policy of 247.
80
And there does not appear to be any evidence to the contrary.
81
CMA2, para 4.20.
48
Shareholders benefitted from the same veto rights, these rights (alongside
the other factors set out at paragraph 5.27 above) would likely have given
Hunter Douglas the ability to restrict 247’s autonomy to carry out its
business activities, e.g. by being able to veto key decisions. This is
sufficient to find that Hunter Douglas acquired the ability to materially
influence 247’s strategic direction and commercial objectives as a result of
the 2013 Transaction.
82
The fact that some of these rights were not
exercised does not affect this finding as establishing that the ‘material
influence' threshold is met does not require the exercise of material
influence, merely the ability to do so.
83
The 2019 Transaction
As a result of the 2019 Transaction, Hunter Douglas acquired 100% of the
shareholding in 247 and, as such, increased its share in the profits of 247.
We conclude that this clearly amounts to the acquisition of a controlling
interest. As the sole shareholder, Hunter Douglas would have the ability to
unilaterally pass any shareholder resolutions allowing it to unilaterally
determine 247’s policy, including all aspects of 247’s competitive strategy
(including the ability to set 247’s prices) . In addition, Hunter Douglas was
no longer subject to any voting or other rights previously held by the 247
Founding Shareholders, including their previous veto rights. This is not
disputed by the Parties.
As noted above, pursuant to section 26(4) of the Act, an increase in the
level of influence from material influence or ‘de facto’ control to a controlling
interest is sufficient to constitute a new RMS. Therefore, even if the 2013
Transaction conferred ‘de facto’ control over 247, which we do not agree is
the case for the reasons set out above, the acquisition of a controlling
interest in 247 in 2019 is sufficient to find a new RMS. The CMA has
exercised its discretion to assert jurisdiction over such changes in the level
of interest in a number of other cases
84
and we see no basis to not
exercise that discretion in this case. Indeed, on the contrary, in our view,
the fact that there has been a substantive change in the rights held by
Hunter Douglas as a result of the 2019 Transaction, e.g. moving from being
subject to the veto and voting rights of the 247 Founding Shareholders after
the 2013 Transaction, to having all the voting rights and no longer being
82
SLC Decision, paragraph 47.
83
CMA2, para 4.14.
84
See for example the Anticipated acquisition by Cavendish Square Partners (General Partner) Limited of a
controlling interest in each of Lakeside 1 Limited (Keepmoat) and Apollo Group Holdings Limited (Apollo)
(ME/5213/11 and ME/5291/11), OFT decision of 24 November 2011 and the Anticipated acquisition by Guardian
Media Group of Trader Media Group, OFT Decision of 29 September 2003.
49
subject to any vetoes from the 247 Founding Shareholders after the 2019
Transaction, supports the exercise of our discretion in this case.
While Hunter Douglas’ voting rights in 247 decreased following the 2016
Transaction (from 49% to 24.9%) on 6 December 2016
85
and just prior to
the 2017 Transaction (from 24.9% to 4.9%) on 11 May 2017
86
, in our view
these changes did not remove Hunter Douglas’ ability to exercise material
influence over 247’s policy. Indeed, at their hearing, the representative from
Hunter Douglas noted that these reductions of Hunter Douglas’ voting
rights were a purely ‘technical matter’ designed to keep the 2013
Transaction ‘confidential.’
87
All the other rights set out at paragraph 3.4
above remained unchanged. As noted at paragraph 5.29 above,
notwithstanding the voting rights acquired by Hunter Douglas, the other
rights acquired through the 2013 Transaction would likely have given
Hunter Douglas the ability to continue restricting 247’s autonomy to carry
out its business activities and enabled Hunter Douglas to materially
influence 247’s strategic direction and commercial objectives. Even if these
changing voting rights had reduced the level of influence Hunter Douglas
had over 247’s policy, this would not have affected the analysis of the 2019
Transaction. The move from a lower level of influence, to a controlling
interest would still be sufficient to constitute a new RMS pursuant to section
26(4) of the Act.
The Share of Supply Test – Nexus with UK
The second element of the jurisdictional test seeks to establish a sufficient
connection with the UK on a turnover and/or share of supply basis. As
noted at paragraph 5.4 above, the turnover test is not met in this case. Our
conclusions on whether the share of supply test is met in relation to the
Transactions is set out below.
Time at which the relevant thresholds should be assessed
As noted above, in this case, we are considering the two Transactions, one
of which completed on 19 February 2019 (the 2019 Transaction) just over a
year before the CMA’s decision on 1 April 2020 to refer the Transactions to
a Phase 2 inquiry (the Reference Decision)
88
and one of which completed
85
We understand that this was for the reasons set out at paragraph 3.14 above.
86
The Parties have confirmed that this was for the reasons set out at paragraph 3.16 above.
87
Despite this, the CMA notes that Hunter Douglas made no mention of either the 2013 Transaction or having
material influence over 247 in their Merger Notice or in any submissions filed as part of the CMA’s investigation
into the 2017 Transaction (https://www.gov.uk/cma-cases/hunter-douglas-bellotto-merger-inquiry).
88
The CMA’s decision to refer case ME/6867/19 under section 22 of the Act
(https://assets.publishing.service.gov.uk/media/5e8b2c2086650c18cf162793/Decision_to_refer.pdf)
50
almost seven years prior to the Reference Decision on 30 April 2013 (the
2013 Transaction).
In the SLC Decision, the CMA considered that, for the purposes of
assessing whether the 2013 Transaction amounts to an RMS, it is
necessary, pursuant to section 23(9) of the Act, to assess whether the
share of supply test was met immediately before the CMA’s reference
decision (i.e. in this case, as at the time of the Reference Decision).
The Parties made a number of representations on this issue prior to the
SLC Decision. These were contained in a standalone submission dated 12
February 2020 and the Parties’ response to the CMA’s Issues Letter dated
4 March 2020. These representations were addressed in the SLC Decision.
Following the SLC Decision, the Parties submitted a white paper on 14
April 2020, and a response to our working paper on jurisdiction, on 16 June
2020, containing further representations on this issue as well as repeating
some previous representations.
89
These representations submitted that, contrary to the CMA’s interpretation,
the share of supply threshold should be assessed at the time the 2013
Transaction completed.
We note that section 23(9) of the Act states that, for completed mergers,
‘the question whether a relevant merger situation has been created shall be
determined as at… immediately before the time when the reference [to a
Phase 2 inquiry – see Section 22] has been, or is to be, made.’ (emphasis
added). In our view this language means that the question of whether an
RMS has been created, which as per sections 23(1) and (2) of the Act
includes the finding that the jurisdictional thresholds are met, is to be
determined ‘as at’ the time immediately before the Reference Decision, not
when the relevant Transaction has completed. This interpretation is
consistent with the body of CMA decisional practice
90
and the CMA’s
89
The Parties made no further substantive representations on this issue or on our jurisdictional findings more
broadly in their Response to the Provisional Findings.
90
See for example, Completed acquisition by Tesco Stores Limited of Brian Ford Discount Store Limited,
(ME/3827/08) OFT decision of 22 December 2008; Completed acquisition by Ryanair Holdings plc of a minority
interest in Aer Lingus Group plc (ME/4694/10), OFT decision of 15 June 2012; A report on the completed
acquisition by Intercontinental Exchange, Inc. of Trayport, Final Report of the Competition Commission dated 17
October 2016 (not appealed on this issue); A report on the completed acquisition by Sonoco Products Company
of Weidenhammer Packaging Group GmbH, Final Report of the Competition Commission dated 3 July 2015; A
report on the completed acquisition by Xchanging plc of certain companies comprising all of the European
operations of Agencyport Software Group, Final Report of the Competition Commission dated 29 April 2015; A
report on the completed acquisition by Alliance Medical Group Limited of the assets of IBA Molecular UK Limited
used to manufacture 18F-Fluorodeoxyglucose, Final Report of the Competition Commission dated 15 August
2014; A report on the completed acquisition by Groupe Eurotunnel S.A. of certain assets of former SeaFrance
S.A., Final Report by the Competition Commission dated 6 June 2013; Completed acquisition by MWUK Holding
Company Limited of Dimensions Clothing Limited and Completed acquisition by MWUK Holding Company
Limited of certain assets of Alexandra plc (in administration) ME/4664/10 (MWUK Holding), OFT Decision of 19
November 2010.
51
guidance, which notes at footnote 73
91
that ‘in accordance with section
23(9) of the Act, the CMA assesses whether the share of supply test is met
at the time of its decision on reference.’
The Parties have disputed the CMA’s reading of section 23(9) of the Act
submitting that the provision is merely one ‘which imposes an obligation on
the CMA to make a determination and is not a provision which determines
the facts that are to be taken into account in making that determination.’
However, the Parties’ reading of section 23(9) is inconsistent with the
natural interpretation of the words ‘as at’ in section 23(9) of the Act, which
is ‘based on the situation at the time’. This supports our view that the share
of supply test should be calculated based on the situation of the Parties at
the time of the Reference Decision. In their WP Response, the Parties
contested the CMA’s interpretation of these words, submitting that they
‘simply mean that the CMA must satisfy itself ‘as at’ the point of reference
that any RMS which arose at the time enterprises ceased to be distinct
persists at the time of the reference.’ However, we find this submission to
be unconvincing as the CMA’s interpretation more closely tracks the natural
interpretation of the words ‘as at’.
The Parties claimed that the CMA’s interpretation of the Act has not
accounted for and is inconsistent with the effect of section 23(2A) of the
Act. This provision states that: ‘The share of supply test is met if… as a
result of the enterprises ceasing to be distinct enterprises, one or both
of the conditions mentioned in subsections (3) and (4) below prevails or
prevails to a greater extent [which set out the share of supply thresholds]’
(emphasis added). Contrary to the Parties’ interpretation, our view is that
this provision merely requires the CMA to ensure that the share of supply
test is passed as a result of an increment to the shares of supply of the
enterprises that have ceased to be distinct. The provision must also be
read consistently with section 23(9) of the Act, which, for the reasons
stated above, we find specifies the point in time at which the shares of
supply and thus any increment should be assessed.
In their representations the Parties also claimed that the CMA’s
interpretation is contrary to the intention of Parliament. In support, the
representations quote Melanie Johnson MP, then-Minister for Competition
and Consumers, who said, in relation to section 23 of the Act, during its
passage through Parliament: ‘The purpose of the test is to take out of the
scope of merger control a large number of transactions that are of no
economic concern and to give business regulatory certainty that they will
not fall within merger control.’ The Parties also noted that the CAT referred
91
CMA2.
52
to business certainty at paragraph 82 of its judgment in the Lebedev
92
case: ‘Altogether, the merger control regime in the Act is replete with time-
limits for the various subsidiary stages, and very specific prescriptive
provisions regarding the circumstances in which those limits can be
extended and for how long. That approach clearly supports business
certainty regarding potentially major transactions.’ The CMA does not
dispute the importance of business certainty. However, both the quote from
Melanie Johnson MP and the paragraph of the Lebedev case cited by the
Parties, set out above, are concerned with the existence and compliance
with statutory time limits. In this case it is not disputed that the CMA has
complied with the time limits in the Act for making a reference decision.
Indeed, the Parties do not dispute that the CMA was not out of time in
choosing to call in the 2013 Transaction once the CMA became aware of
the material facts about the 2013 Transaction.
Nevertheless, as explained below, in the particular and unusual
circumstances of this case, we consider that the very lengthy period of
almost seven years which elapsed between the 2013 Transaction and the
Reference Decision, combined with the fact that it was clear that any
overlap between the Parties at the time of the 2013 Transaction could not
have satisfied the share of supply test, are matters which we can and
should take into account in interpreting the evidence as to whether the
share of supply test was satisfied ‘as at’ the time of the Reference
Decision.
Share of Supply
The share of supply test will be met if, at the time it is determined, the
merging parties supply or acquire 25% or more of particular goods or
services, in the UK.
At the time of the 2013 Transaction the Parties did not overlap in the online
retail supply of M2M blinds in the UK and it was clear that any overlap
between the Parties at that time could not have satisfied the share of
supply test. As noted by the Parties, the only retail operation of Hunter
Douglas in the United Kingdom prior to the 2013 Transaction was Thomas
Sanderson, who, at that time, did not and continues not to offer its services
online.
However, at the time of the Reference Decision, both Parties were active in
the online retail supply of M2M blinds and had a combined share (by
92
Lebedev Holdings Limited and Independent Digital News and Media Limited v DCMS [2019] CAT 21.
53
revenue) of more than 25% in the online retail supply of M2M blinds in the
UK in 2019.
For the reasons set out above, we consider that it is appropriate to assess
whether the share of supply test is met based on the situation as at the
time of the Reference Decision. Therefore, were we to focus solely on the
Parties’ shares of supply as at the date of the Reference Decision, we
would conclude that the share of supply test is met in relation to both
Transactions.
However, we also note that the facts of this case are unusual. In particular,
there was an exceptionally lengthy period of almost seven years between
the 2013 Transaction and the Reference Decision. In addition, in this case,
it was clear that any overlap between the Parties at the time of the 2013
Transaction could not have satisfied the share of supply test. Thus,
although we disagree with the Parties’ interpretation of the Act, we
recognise the need to take account of this combination of unusual features
in interpreting the evidence in this particular case.
Considering the evidence in the round, we are not satisfied that the share
of supply test is met in relation to the 2013 Transaction. In contrast, we
conclude that the share of supply test is met in relation to the 2019
Transaction, on the basis of the Parties’ combined shares of supply by
revenue as at the date of the Reference Decision, as noted above.
While the Parties’ combined share of supply did not increase as a result of
the 2019 Transaction, where, as here, the RMS is the result of an increase
in the level of interest held by a party, we consider section 26(4) of the Act
allows for the acquirer to be ‘treated’ as bringing the target under its control
(notwithstanding that it already had the ability to exercise material influence
or ‘de facto‘ control over the target’s policy) such that there would therefore
(under such ‘treatment’) be an increment in the share of supply.
93
Time period for investigating mergers
To meet the criteria for an RMS, the enterprises must have ceased to be
distinct either not more than four months before the date on which the
reference is made or where the Transaction took place without having been
made public and without the CMA being informed of it, or four months from
the earlier of the time that material facts are made public or the time the
93
CMA2, footnote 44.
54
CMA is told of material facts.
94
The four-month period may be extended
under section 25 of the Act.
In this case, as noted above, the Stakeholders Agreement entered into as
part of the 2013 Transaction contained confidentiality provisions requiring
both Parties to keep Hunter Douglas’ participation in 247 strictly
confidential. The Parties do not dispute that the 2013 Transaction was, as a
result, not subsequently made public and that material facts were not
provided to the CMA at the time the 2013 Transaction completed. We do
not consider that notice of material facts regarding the 2013 Transaction
and the 2019 Transaction was given by Hunter Douglas to the CMA prior to
22 November 2019 and 28 October 2019, respectively. The Parties do not
dispute this.
We therefore conclude that the applicable statutory time limits in relation to
this reference have been complied with.
94
Section 24 of the Act.
55
6. Counterfactual
Introduction
The counterfactual is an analytical tool used to help answer the question of
whether a merger has resulted, or may be expected to result, in an SLC.
95
It does this by providing the basis for a comparison of the competitive
situation in the market with the merger against the likely future competitive
situation in the market absent the merger.
96
The latter is called the
counterfactual.
97
As we have found (at paragraph 5.48 above) that we have no jurisdiction
over the 2013 Transaction, this section only considers the appropriate
counterfactual for the 2019 Transaction.
The CMA’s counterfactual assessment framework
As part of its counterfactual assessment, the CMA may examine several
likely future scenarios, one of which may be the continuation of the pre-
merger situation. The CMA will select the most likely of these, based on the
facts of the case, as the counterfactual scenario.
98
It will incorporate into
the counterfactual only those aspects of scenarios that appear likely, based
on the facts available to it and the extent of its ability to foresee future
developments.
99
The foreseeable period can sometimes be relatively
short.
100
However, even if an event or its consequences are not sufficiently
certain to include in the counterfactual they may be considered in the
context of the competitive assessment.
101
The application of the SLC test involves a comparison of the prospects for
competition with the merger against the competitive situation without the
merger.
102
Consequently, the counterfactual cannot include a relevant
merger situation that has the same effect as the merger we are assessing
(in this case, the 2019 Transaction), given that such alternative relevant
merger situation would be subject to the provisions of Part 3 of the Act, and
95
MAGs, paragraph 4.3.1.
96
MAGs, paragraphs 4.3.1 and 4.3.6.
97
MAGs, paragraph 4.3.1.
98
MAGs, paragraph 4.3.6. In contrast, at Phase 1, the effect of the merger is compared with what is considered
to be the ‘most competitive’ counterfactual (provided that this situation is considered to be a realistic prospect).
99
MAGs, paragraph 4.3.6.
100
MAGs, paragraph 4.3.6.
101
MAGs, paragraph 4.3.2.
102
MAGs, paragraph 4.3.1.
56
would not therefore carry sufficient certainty to mean it was the most likely
future scenario.
The CMA seeks to avoid importing into the assessment of the appropriate
counterfactual any spurious claims to accurate prediction or foresight.
Given that the counterfactual incorporates only those elements of scenarios
that are foreseeable, it will not in general be necessary to make finely
balanced judgements about what is and what is not included in the
counterfactual.
103
However, where we consider that the choice between
two or more counterfactual scenarios will make a material difference to the
competitive assessment, the CMA will carry out additional detailed
investigation before reaching a conclusion on the appropriate
counterfactual.
104
Depending on the evidence, the choice of the counterfactual could be a
situation either more or less competitive than the competitive conditions
prevailing at the time the merger occurred. Therefore, the selection of the
appropriate counterfactual may increase or reduce the prospects of an SLC
finding.
105
In reaching its view on the appropriate counterfactual, the CMA determines
what future developments it foresees arising absent the merger based on
the facts available. Insofar as future events or circumstances are not certain
or foreseeable enough to include in the counterfactual, the analysis of such
events can take place in the assessment of competitive effects.
Owing to the
inherent uncertainty of predicting future events, the CMA benefits from a
‘margin of appreciation’ or evaluative discretion, in relation to its conclusions
and is likely to be deemed to have acted rationally provided it has taken
account of all relevant information.
106
Parties’ views on the appropriate counterfactual
The Parties noted that the pre-merger conditions of competition were those
in which Hunter Douglas had acquired a controlling interest in Blinds2Go
107
and, at least, material influence over 247 (through the 2013
Transaction)
108
.
103
MAGs, paragraphs 4.3.2 and 4.3.6.
104
MAGs, paragraph 4.3.6.
105
MAGs, paragraph 4.3.4.
106
See BAA Ltd v Competition Commission [2012] CAT 3 at [20] (in the context of a market investigation);
Stagecoach Group Plc v Competition Commission [2010] CAT 14, paragraph 45.
107
On 21 June 2016 Hunter Douglas acquired 60% equity stake of Blinds2Go [].
108
As noted at paragraph 5.23 above, the Parties have submitted that Hunter Douglas acquired a controlling
interest or, at least, ‘de facto’ control over 247 through the 2013 Transaction. However, for the reasons set out at
paragraph 5.29 above, we find that Hunter acquired only material influence over 247’s policy through the 2013
Transaction.
57
The Parties therefore submitted that the correct counterfactual was one
where 247 would have continued to exercise the same degree of
competitive constraint that it would have done absent the 2019
Transaction.
109
Their submissions claimed that 247 exercised only a
‘limited constraint’ on Blinds2Go before the 2019 Transaction, and that the
acquisition by Hunter Douglas of a controlling interest in 247 through the
2019 Transaction had no impact on the competitive structure of the
market.
110
Hunter Douglas also submitted that ‘[i]n respect of the 2019 Transaction,
the only plausible counterfactual is the prevailing conditions of competition,
a situation where Hunter Douglas holds, at the very least, the ability to
exercise material influence over the activities of 247.’
Furthermore, Hunter Douglas submitted:
(a) Hunter Douglas would have been able to prevent the sale of either
or both of the 247 Founding Shareholders’ stakes;
(b) As a result, it would have been able to impose conditions on any
purchaser, including the retention of the rights which it held under
the Stakeholder Agreement;
(c) the CMA has not considered the likelihood that a third-party
purchaser of a 51% stake in the business would proceed without
entering into an agreement with Hunter Douglas similar in nature to
the Stakeholder Agreement.
In addition, Hunter Douglas submitted that the CMA’s assessment, in its
Provisional Findings, of Hunter Douglas’s ability to block a sale of shares
by the Founding Shareholders was wrong as a matter of law.
Hunter Douglas also submitted that the assumption that the Founding
Shareholders would have declined to exercise their put options does not
withstand scrutiny, and that there was no reason to believe that both
Founding Shareholders would have disposed of their shares in the event
the options were not exercised.
111
Hunter Douglas proposed an alternative counterfactual scenario, where it
would have bought the shares of one of the 247 Founding Shareholders,
109
Main submission, 24 May 2020, paragraph 3.36.
110
Main submission, 24 May 2020, paragraph 3.32.
111
Response to the Provisional Findings, paragraph 3.7
58
giving it a 74.5% stake in 247. It said that this scenario was ‘more plausible
(or at the very least, just as likely)’
112
as the CMA’s preferred scenario.
Assessment
Introduction
As noted at paragraph 5.29, we have found that Hunter Douglas had the
ability to exercise material influence over 247’s policy
113
prior to the 2019
Transaction.
We note the Parties’ submission at paragraph 6.9 above that the correct
counterfactual is one where 247 would have continued to exercise the
same degree of competitive constraint that it would have done absent the
2019 Transaction. This submission, in effect, asserts that the most likely
counterfactual for the 2019 Transaction is the continuation of the pre-
merger situation. While this is often the CMA’s starting point, our guidance
is clear that, in a phase 2 inquiry, the CMA ‘may examine several possible
scenarios, one of which may be the continuation of the pre-merger
situation’ and of these ‘only the most likely scenario will be selected as the
counterfactual.’
114
As regards Hunter Douglas’ proposed alternative counterfactual scenario
summarised at paragraph 6.14, the CMA notes this contradicts the principle
noted at paragraph 6.4 above that the counterfactual cannot incorporate a
scenario with the same effect as the merger. In this case the relevant
merger situation we have found as a result of the 2019 Transaction is the
increase in the level of interest Hunter Douglas held in 247 from material
influence to a controlling interest. The scenario proposed by Hunter
Douglas would have the same effect. Even though a shareholding of 74.5%
is less than the 100% shareholding Hunter Douglas acquired through the
2019 Transaction, it would still amount to Hunter Douglas acquiring a
controlling interest
115
in 247. As a result, for the reasons set out in
paragraph 6.4 above we do not consider this scenario to be the most likely
counterfactual.
112
Response to the Provisional Findings, paragraph 3.36
113
The policy of the target in this context means the management of its business, and thus includes the strategic
direction of a company and its ability to define and achieve its commercial objectives (See, paragraph 4.14),
114
MAGs, paragraph 4.3.6.
115
It would be more than de facto control as defined at paragraph 4.28 of Mergers: Guidance on the CMA’s
jurisdiction and procedure, CMA2: ‘Merger arrangements may give rise to a position of ‘de facto’ control when an
entity controls a company’s policy, notwithstanding that it holds less than the majority of voting rights in the target
company (that is, it does not have a controlling interest).’
59
We consider Hunter Douglas’ submissions set out in paragraph 6.12 above
in Appendix B. For the reasons set out in that appendix, we are
unconvinced by Hunter Douglas’ arguments that it had the legal ability to
block any sale by the Founding Shareholders.
Hunter Douglas’ other submissions in their Response to our Provisional
Findings, as set out in paragraphs 6.11 and 6.13 above, are considered in
our assessment of Scenarios 1 and 2 below.
We have found no evidence to suggest that Blinds2Go would have done
anything other than continue to compete in line with the conditions
prevailing at the time of the 2019 Transaction, with Hunter Douglas
potentially increasing its shareholding but maintaining a controlling interest.
For 247, we have considered three scenarios relating to its ownership,
which are set out below.
Scenario 1: Continuation of majority ownership by the 247 Founding
Shareholders
The situation prior to the 2019 Transaction was that Hunter Douglas held
certain rights in 247, which were attached to convertible loan notes it had
acquired in 2013 and set out in a stakeholder agreement (the Stakeholder
Agreement) between Hunter Douglas and the 247 Founding Shareholders.
These convertible loan notes were subject to the terms of the LNI between
the Parties. This instrument provided that ‘Notes shall be repaid in full at
par on 30 June 2020 or (if later) on the tenth business day following
finalization of the audited Accounts of the Company for the fiscal year
ending 28 February 2019.’ The audited accounts of 247 have been filed for
that fiscal year.
This means that after 30 June 2020, if no other action was taken, the loan
notes would have matured and been redeemed. If this had occurred Hunter
Douglas would no longer have had any interest in 247 and any rights it held
in 247 prior to the 2019 Transaction, attached to the convertible loan notes
it acquired through the 2013 Transaction, would have fallen away. In our
view, 247 would have then competed with Blinds2Go, and any other Hunter
Douglas group companies, as a fully independent rival.
However, the maturing and redemption of the loan notes would have led to
Hunter Douglas losing substantially all of the value of their 2013 investment
in 247. Hunter Douglas told us that ‘[g]iven that the value of 49% of the
share capital of 247 would have run to many millions of pounds at whatever
point the conversion right was exercised, it would have made no sense to
have redeemed the loan notes in 2020.’ As a result, in our view, Hunter
60
Douglas would have had a strong incentive to exercise its right under the
LNI to convert its loan notes into 49% of the equity shares of 247.
Therefore, under this scenario:
(a) Hunter Douglas would have initially continued to have the same level of
influence over 247’s policy that it had prior to the 2019 Transaction.
(b) At some point before 30 June 2020, Hunter Douglas would have
exercised its right to convert its loan notes into equity shares, providing it
with a 49% stake in 247. Hunter Douglas would also have continued to
enjoy the voting, veto and other rights that are set out in the Stakeholder
Agreement, summarised at paragraph 3.4 above, which does not have a
termination date and has effect regardless of whether Hunter Douglas’
interest in 247 is held as shares or loan notes. Therefore, on conversion
of the loan notes to equity, Hunter Douglas would have retained a similar
ability to influence 247’s policy
116
that it did prior to the 2019 Transaction.
As a result, we find that, under this scenario, the conditions of competition
throughout the counterfactual period may be similar to those that prevailed
prior to the 2019 Transaction. Even after Hunter Douglas exercised its
option to convert the loan notes to equity, which, in our view, would be
likely to happen at some point before 30 June 2020, it would continue to
benefit from the additional rights set out in the Stakeholder Agreement.
This would mean that conditions similar to the prevailing conditions of
competition that existed prior to the 2019 Transaction would continue for
the foreseeable future.
Scenario 2: Alternative purchaser of Founding Shareholders’ stakes in 247
We also considered whether, in the absence of the 2019 Transaction, the
247 Founding Shareholders’ stakes would have been acquired by an
alternative purchaser. Given the ability and incentive for Hunter Douglas to
convert its loan notes into equity before 30 June 2020, we find that it would
also do so under this scenario.
In considering this scenario, in our view, the following evidence clearly
indicates that it was the continuing intention of the 247 Founding
Shareholders to ultimately sell their shares in 247. Hunter Douglas told us
that the 2013 Transaction was structured to provide for the eventual exit of
247’s Founding Shareholders from the business. It said ‘From the 247
perspective, the structure provided a guaranteed exit route. David Maher,
in particular, was looking to retire and the put option allowed both of the
116
As noted at paragraph 5.29 above, we find this to be material influence.
61
247 founding shareholders to exit the business whilst maximising the value
which could be achieved.’ Moreover, as discussed at paragraph 3.17(a),
the 247 Founding Shareholders also discussed their intention to exercise
the put option, which they acquired in 2013, with Hunter Douglas in 2017,
indicating their continued intention to sell their shares in 247. This would
have required Hunter Douglas to acquire 100% of the shares in 247 in
2019.
[].
117
[],
118
[].
The evidence above shows that several potential purchasers were
interested in acquiring 247 around the time of the 2019 Transaction. Given
that, for the reasons set out in paragraph 6.27 above, it was the continuing
intention of the 247 Founding Shareholders to sell their shares, we find that
247 Founding Shareholders are likely to have been supportive of any such
sale to a third party. The competitive outcome under this scenario would be
similar to Scenario 1 until the sale to an alternative purchaser. At this point,
Hunter Douglas would own 49% of the equity in 247 with a third party now
holding a majority 51% of the shares in 247.
One of the rights held by Hunter Douglas under the LNI was a veto over
any transfer of the 247 Founding Shareholders’ shares. We find that while
Hunter Douglas had the benefit of this right it would have both the ability
and the incentive to use it as leverage to impose conditions or obligations,
including imposing the additional rights it held under the Stakeholder
Agreement (summarised at paragraph 3.4 above), against any third-party
buyer.
However, as noted in paragraph 6.24(b) above, in our view it is likely that
Hunter Douglas would have converted its loan notes into equity before they
matured and redeemed on 30 June 2020. At this point, Hunter Douglas
would no longer benefit from the veto right in the LNI over the transfer over
the 247 Founding Shareholders shares for the reasons set out in of
Appendix B.
We also considered whether there were any other terms in the Stakeholder
Agreement, or each of the Call Option Agreements or Put Option
Agreements that would have given Hunter Douglas the ability to prevent a
sale of the 247 Founding Shareholders’ shares in 247, and thus provided
Hunter Douglas with similar leverage to impose any conditions or
obligations against any third-party buyer of the 247 Founding Shareholders’
shares. For the reasons set out in Appendix B once the loan notes either
117
[].
118
[].
62
matured, and were redeemed, or were converted to equity, in our view,
Hunter Douglas no longer had such an ability and, thus, no such leverage.
In its response to the Provisional Findings, Hunter Douglas disputed that it
would not have leverage in this scenario. It said that it had the ‘ability to
bring considerable commercial pressure to bear’ on 247’s new owner as
part of negotiations around joint venture arrangements. It also said that the
247 Founding Shareholders and any new owner would have ‘strong
incentives’ to avoid disputes and maintain a working relationship with it,
given its continuing 49% stake.
119
Because any third-party buyer would not be a party to the Stakeholder
Agreement, we find that, if the 247 Founding Shareholders were to have
sold their shares to a third party after Hunter Douglas converted their loan
notes to equity, Hunter Douglas would not have had a legal veto in respect
of such a sale and would not have had the leverage from such a veto to
impose additional rights against that third party beyond those it would have
as a result of holding 49% of the shares in 247.
Notwithstanding the lack of a veto, it is possible that Hunter Douglas would
be able to negotiate some limited rights with the new majority owner.
However, in our view, this would, at most, be limited to having rights that
protected the value in Hunter Douglas’ remaining equity stake in 247 and
would not extend to being able to negotiate rights that would have a
material impact on the ability of the purchaser to be able to pursue an
independent competitive strategy with 247 from Hunter Douglas’ other
business interests, including Blinds2Go. In our view it is unlikely that Hunter
Douglas would have been able to negotiate operational veto rights similar
to the ones it enjoyed before the 2019 Transaction from a new third-party
purchaser. As a result, this would leave Hunter Douglas with less ability to
influence the policy of 247, including its strategic direction and commercial
objectives, than it had prior to such a sale.
Scenario 3: Alternative purchaser for 100% of 247
Under this scenario, an alternative purchaser would have acquired 100% of
247, as a result of the 247 Founding Shareholders and Hunter Douglas
both selling their stakes in 247.
Given the incentive and stated objective of Hunter Douglas to remain
invested in 247, and given the stated objective of the 247 Founding
Shareholders to sell their shares and for David Maher to exit the business
119
Parties’ Response to Provisional Findings, paragraph 3.29
63
at some point, our initial view is that this scenario might occur under the
following conditions:
(c) No alternative purchaser could be found for just the 247 Founding
Shareholders’ 51% stake; and
(d) The differing objectives between Hunter Douglas and the 247 Founding
Shareholders were such that they led to a decline in the performance of
247 to the extent that Hunter Douglas concluded that the best way to
realise the value of its investment would be through a sale.
Under this scenario, the conditions of competition would be similar to those
prevailing at the time of the 2019 Transaction until 247 was sold, when it
would compete under its new ownership as an independent rival to Hunter
Douglas.
Hunter Douglas submitted that it had never considered the possibility of
selling its stake in 247. At the main party hearing, the representatives of
Hunter Douglas stated that Hunter Douglas had never sold a core window
covering business and that it saw itself as a ‘permanent home for
entrepreneurs who want to sell their business’. They went on to state that ‘it
would cause real reputational damage if we were to be perceived as people
who buy and sell companies.’
Hunter Douglas also pointed out that 247 had not suffered a decline in
performance and that it was continuing to perform well at the time of the
2019 Transaction.
Assessment of scenarios
For 247, we have considered three counterfactual scenarios:
(a) Scenario 1: Continuation of majority ownership by the 247 Founding
Shareholders;
(b) Scenario 2: Alternative purchaser of the 247 Founding Shareholders’
stakes in 247;
(c) Scenario 3: Alternative purchaser for 100% of 247.
We note the Parties’ arguments with respect to Scenario 3. While the
circumstances required for a 100% sale of 247 appear possible, in our view
this scenario is not the most likely counterfactual.
We then considered whether Scenario 1 or Scenario 2 was a more likely
counterfactual. Under both scenarios Hunter Douglas would continue to
64
own 49% of 247. However, only under scenario 1 would Hunter Douglas
have substantial additional rights in 247, namely those it held prior to the
2019 Transaction under the Stakeholder Agreement.
We considered Hunter Douglas’ submission, set out in paragraph 6.13
above, that our assumption that the 247 Founding Shareholders would
have ‘declined’ to take up their put options ‘does not withstand scrutiny’.
We would note that, as explained at paragraph 6.17 above, we cannot
consider the exercise of one of the call or put options as the most likely
counterfactual as this would result in Hunter Douglas acquiring a controlling
interest in 247, and thus have the same effect as the relevant merger
situation we found in relation to the 2019 Transaction.
We also note Hunter Douglas’ submission that there was no reason to
believe that both 247 Founding Shareholders would wish to sell their
stakes.
120
As evidence that this was not the case, Hunter Douglas
submitted an email exchange from June 2018 between Mr Peterkin, one of
the 247 Founding Shareholders, and Mr Sonnenberg. Hunter Douglas
submitted that this email exchange demonstrated Mr Peterkin’s desire to
remain with the 247 business. In the email exchange, [].
We recognise that the 247 Founding Shareholders had individual
agreements with Hunter Douglas. It is therefore legally possible that only
one of the 247 Founding Shareholders could have sold their interest to a
third party who would then be a minority shareholder holding 25.5%
alongside the remaining 247 Founding Shareholder. This would lead to an
ownership structure where Hunter Douglas held 49% of 247, with the other
247 Founding Shareholder and the third party each holding 25.5%. We also
recognise that, following the 2019 Transaction, while one shareholder
retired, the fact that the other, namely Mr Peterkin, continued to be
employed by Hunter Douglas to manage 247 indicates that Mr Peterkin had
an interest in remaining involved in the 247 business. In our view, [].
We consider that the scenario outlined above is less likely than a sale of
both 247 Founding Shareholders’ stakes together to a single third-party
purchaser for the following reasons:
(a) Given the previous joint ownership and management of 247 by the
247 Founding Shareholders, we consider they would be unlikely to
make such a significant decision independently of each other. We do
not find that the email exchange Hunter Douglas submitted to us,
summarised at paragraph 6.45 above, is contrary to this view;
120
Parties’ response to Provisional Findings, paragraph 3.33
65
(b) In addition, in our view, a minority stake in 247 would be less to be
attractive to a potential purchaser than a controlling stake. The sale
of both of the 247 Founding Shareholders’ stakes together would
attract a control premium from any buyer and so be worth more to
each of them than selling their stakes individually to separate
buyers; and
(c) Mr Peterkin selling his stake would not prevent him remaining
involved with 247. Under the 2019 Transaction itself, Mr Peterkin
sold his stake in 247 but was able and willing to remain involved in
managing 247. A similar arrangement could be put in place by a
single third-party purchaser of both 247 Founding Shareholders’
stakes.
We note that 247 is a profitable business with a track record of growth. The
247 Founding Shareholders had already received one formal offer to
purchase the business (from []). We also note the expressed desire of
the 247 Founding Shareholders to sell their shares and for David Maher to
exit the business (see paragraph
6.27 above). We consider that these
factors combined provide reasonable evidence that the 247 Founding
Shareholders would, in the absence of the 2019 Transaction, have been
most likely to have sold their shares to a third party. We accept, for the
reasons set out at paragraph 6.35, that this third party might agree to grant
some rights to Hunter Douglas but that these would be limited to protecting
the value of Hunter Douglas’ investment asset out in paragraphs 22 and 23
of Appendix B and at paragraph 6.35 above.
Findings on the most likely counterfactual
We find it most likely that, in the absence of the 2019 Transaction, the
shares held by the 247 Founding Shareholders would at some point have
been bought by a third party after Hunter Douglas had converted their loan
notes to equity. For the reasons set out at paragraph
6.27 above, in our
view, it was the continuing intention of the 247 Founding Shareholders to
sell their shares in 247. We therefore find that, absent the 2019
Transaction, the most likely scenario is that both the 247 Founding
Shareholders would have sought to sell their shares in 247 to a single third-
party buyer, as per scenario 2. At this point, for the reasons set out at
paragraphs 6.32 to 6.35 above, Hunter Douglas would no longer be able to
exercise the veto and other rights it held under the Stakeholder Agreement,
leading to 247 having more independence that it had prior to the 2019
Transaction.
66
7. Market definition
In this section we present our assessment of the relevant product and
geographic market. Market definition provides the appropriate framework
for assessing the competitive effects of the Merger and involves an element
of judgement. The boundaries of the market do not determine the outcome
of the analysis of the competitive effects of the 2019 Transaction, as it is
recognised that there can be constraints on merging parties from outside
the relevant market, segmentation within the relevant market, or other ways
in which some constraints are more important than others. We will take
these factors into account in our competitive assessment.
121
Product market definition
The relevant product market is a set of products that customers consider to
be close substitutes, for example in terms of utility, brand or quality.
122
In
identifying the relevant product market the CMA will pay particular regard to
demand side factors (the behaviour of customers and its effects). However,
the CMA may also consider supply-side factors (the capabilities and
reactions of suppliers in the short term) and other market characteristics.
123
The Parties overlap in:
(a) the online retail supply of M2M blinds;
(b) the online retail supply of M2M curtains;
(a) the online retail supply of ready-made curtains; and
(b) the online retail supply of shutters.
We have restricted our inquiry to consider the overlap between the Parties
in the online retail supply of M2M blinds, this being the principal area of
overlap between the Parties.
As discussed in the Industry Background section, online M2M blinds are
tailored blinds purchased by customers online through a retailer’s website.
Customers can browse the website for various styles of blinds (eg roman,
venetian) in a variety of materials and colours, and usually are able to order
a free sample of a given blind.
124
Having selected their desired blind,
121
MAGs, paragraph 5.2.2.
122
MAGs, paragraph 5.2.5(a).
123
MAGs, paragraph 5.2.6.
124
Although some retailers charge a fee for samples.
67
customers provide dimensions, measured in width and drop.
125
Customers
may also be able to make other customisations, for example by dictating
the width of the slats on a venetian or wooden blind.
On receipt of an order, online M2M retailers source the blind to fulfil the
order and arrange delivery. Some retailers, including both of the Parties,
arrange for delivery directly from manufacturers to customers and do not
hold their own inventory.
Once the customer has received their order, they fit the blinds themselves
by either following instructions provided by the retailer, or by arranging for
someone else to fit the blinds independently.
The Parties submitted that the relevant market to assess the Merger should
be wider than M2M blinds, and also wider than blinds more generally. The
Parties specifically noted the constraint from different window coverings
including curtains and shutters and from ready-made blinds. Additionally,
the Parties noted the constraint from in-store and in-home channels on
M2M blinds supplied through the online channel.
126
The Parties submitted
that even if these are not considered as part of the relevant market, it is
highly relevant to consider how such out-of-market factors exercise a
constraint on their activities.
127
The Parties further submitted a customer survey (the ‘BDRC Survey’),
which was commissioned to understand the purchase journey and choices
of the Parties’ customers.
128
We take into account the survey evidence as
part of our market definition and also assess the general robustness of the
survey results in Appendix C.
In this section, using the online retail supply of M2M blinds as our starting
point, we consider whether the market should be widened to include:
(a) curtains and shutters;
(b) ready-made blinds; and
(c) other sales channels, eg in-store and in-home.
In addition to our assessment of the relevant market, we will take into
account the strength of the competitive constraint of alternative products
125
Instructions on how to take these measurements are usually available on online M2M retailers’ websites.
126
Main submission, 20 May 2020, paragraph 6.76 and 6.77.
127
Main submission, 20 May 2020, paragraph 6.65.
128
Main submission, 20 May 2020, paragraph 1.8; BVA-BDRC Blinds Survey Final Report_20 May 2020; Main
submission, 20 May 2020, Annex 0093.
68
and retail channels (as noted in paragraph 7.9 above) as part of the
Competitive Assessment (see paragraphs 8.241 to 8.252).
Curtains and shutters
The Parties’ views
The Parties submitted that any competitive analysis must consider the
extent to which retailers of other window covering products exercise a
competitive constraint on the activities of the Parties.
129
The Parties’ argued that there has been a shift in demand from alternative
window coverings to online M2M blinds, and that customers of online M2M
blinds consider other window coverings before purchasing online M2M
blinds:
(a) The Parties submitted that a report by AMA Research (the ‘AMA
report’) highlights shifts in demand between blinds, curtains and
shutters, with blinds and shutter purchases increasing at the
expense of curtains.
130
Additionally, the Parties submitted that
Blinds2Go’s growth has not been based on winning share from other
online M2M blinds retailers, but persuading customers to choose
online M2M blinds instead of alternative window coverings (or
instead of ready-made blinds and in-store M2M blinds).
131
(b) The Parties submitted that customers start their purchase journey by
considering a range of window covering alternatives they find
appealing across blinds, curtains and shutters.
132
In support of this,
the Parties noted that the BDRC Survey shows that 8-11% of their
customers consider curtains and 16-17% consider shutters before
purchasing M2M blinds online.
133
Our assessment
In Hunter Douglas/Hillarys the CMA concluded that there was a product
frame of reference for the retail supply of blinds, separate from curtains and
shutters.
134
129
Main submission, 20 May 2020, paragraph 6.77 and 6.78.
130
Main submission, 20 May 2020, paragraph 6.78.
131
Main submission, 20 May 2020, paragraph 1.8.
132
Main submission, 20 May 2020, paragraph 6.20.
133
Main submission, 20 May 2020, paragraph 6.78.
134
Hunter Douglas/Hillarys, paragraphs 53 to 59. With respect to curtains, the vast majority of retailers
responding to the CMA’s market testing in that case indicated that curtains are not a demand-side substitute for
blinds. The CMA found a similar lack of demand-side substitution between blinds and shutters and also found
69
For the reasons set out below, our view is that, consistent with the finding
in Hunter Douglas/Hillarys the curtains and shutters should not form part of
the same product market as M2M blinds.
With respect to demand-side substitution, the BDRC Survey suggests that
the constraint from curtains and shutters on the Parties’ online M2M blinds
offering is weak. When asked what they would do in a hypothetical situation
where the respective Party had stopped selling blinds, only 5% of
respondents for Blinds2Go and only 3% of respondents for 247 said that
they would divert to other window coverings.
While we recognise that, as the Parties have highlighted, the BDRC Survey
shows that a higher (albeit still relatively low) proportion of respondents
considered purchasing curtains or shutters before ultimately purchasing
M2M blinds, the fact that customers ‘considered’ other products does not
allow for strong conclusions regarding their willingness to switch to such
products. In particular, ‘consider’ does not reveal whether a customer
would switch to the product considered in case of a price rise or in case its
preferred option was not available. We therefore find that the extent to
which customers considered other window coverings is not particularly
informative as to the competitive constraint these products exert on M2M
blinds.
Responses from third parties were in line with the customer evidence from
the BDRC Survey: of the thirteen third party suppliers of M2M blinds and/or
other window coverings which commented on whether customers would
switch from blinds to curtains or shutters if the price of all blinds increased
by 5%, ten responded that customers would not switch (with the remaining
three suggesting that customers would potentially switch). Additionally,
John Lewis told us that customers typically choose a specific type of
interior window covering early on based on aesthetic and functional
considerations and would postpone the purchase or shop around within a
product category in case of a price rise.
We also assessed the extent to which blinds and other window coverings
are taken into account jointly or presented jointly by the Parties as part of
their competitive offering. We find that blinds are typically positioned
separately from other window coverings, which is consistent with blinds and
other window coverings forming separate relevant markets:
limited evidence of supply-side substitution between blinds and shutters (contrary to Hunter Douglas’
submissions at the time).
70
(a) The Parties do not tend to monitor alternative window coverings vis-
à-vis their M2M blinds.
135
This lack of monitoring by the Parties
supports our view that the Parties’ do not consider alternative
window coverings as a material competitive constraint.
(b) The contemporaneous record of the board-level consideration of
Hunter Douglas’ proposed acquisition of Blinds2Go in 2015
discusses the rationale for that transaction by reference to blinds
only (with there being no mention of curtains or shutters). Whilst
acknowledging that this document was prepared in the context of
that acquisition at the time, we have not been presented with any
further internal documents that indicate that, since 2015, curtains or
shutters pose a material constraint on blinds.
In response to the Parties’ argument on growth in blinds being driven by
migration from other window coverings, we do not consider that this
indicates that other window coverings pose a competitive constraint on
blinds:
(a) First, even if customers would move from other window coverings to
blinds in response to a small but significant non-transitory increase
in price (‘SSNIP’), this would only indicate that blinds exercise a
competitive constraint on other window coverings, but not
necessarily that other window coverings constrain blinds.
(b) Second, a migration to blinds (to the extent it is actually occurring)
could reflect a shift in customer preferences, rather than a result of
substitution, and therefore does not necessarily support the position
that customers would switch from other window coverings to blinds
in response to a SSNIP.
(c) Third, it is unclear to what extent such migration actually took place,
given the Parties submitted that the growth of online M2M blinds has
come primarily at the expense of sales of ready-made blinds as well
as M2M blinds through other distribution channels.
136
Finally, we have considered the possibility of supply-side substitution
between different window coverings. In particular, we have considered the
extent to which suppliers of other types of window coverings may be readily
able to supply online M2M blinds in response to a change of prices. We
note the following in this regard:
135
[].
136
Main submission, 20 May 2020, paragraph 4.16.
71
(a) At the manufacturing/wholesale level, neither the Parties nor third
parties have provided us with evidence that manufacturers of
different window coverings have the ability or incentive to shift
production between different window coverings, or that they actually
do so currently. Further, we also note that window covering retailers
purchase the different types of window coverings from different
manufacturers and that not all of these suppliers supply blinds,
shutters and curtains. This therefore indicates that it may not be
straightforward to use an existing supply chain to be able to supply
different types of product.
137
(b) At the retail level, whilst we acknowledge that certain retailers may
supply more than one type of window coverings, we also note that
suppliers’ market positions (as evidenced by shares of supply) in
blinds, curtains and shutters appear to be materially different.
138
This
material difference between product category shares indicates that
the conditions of competition for different product categories are
different, which is consistent with separate markets for different
products.
Overall, we find that curtains and shutters do not form part of the same
product market as M2M blinds. We take into account any out-of-market
constraint exerted by curtains and shutters in the Competitive Assessment
section, paragraphs 8.241 to 8.252.
Ready-made blinds
The Parties’ view
The Parties also submitted that M2M blinds compete directly with ready-
made blinds.
139
In this regard, the Parties submitted that the BDRC Survey shows that 26%
of Blinds2Go’s customers and 30% of 247’s customers consider ready-
made blinds before purchasing M2M blinds online.
140
The Parties also
submitted that, in addition to multi-channel retailers that sell both M2M and
137
For example, we understand (i) that Rectella and J Rosenthal & Sons, two key suppliers of ready-made
curtains, do not supply blinds; (ii) that fabric suppliers play an important role for made-to-measure curtains but not
for M2M blinds and (iii) that TCMM Shutter Group and Plantation Shutters, two key suppliers of shutters, do not
supply blinds.
138
For example, the main retailers of shutters are diyshutters.co.uk, Californiashutters.co.uk,
shutterlyfabulous.com, shuttercraft.co.uk and Plantation Shutters, neither of which has any significant position in
M2M blinds. Diyshutters.co.uk, Californiashutters.co.uk, shutterlyfabulous.com, shuttercraft.co.uk and Plantation
Shutters.
139
Main submission, 20 May 2020, paragraphs 6.76 and 6.77.
140
Main submission, 20 May 2020, paragraph 6.66.
72
ready-made blinds, the BDRC Survey identifies B&Q, Ikea, Homebase and
Argos as retailers that the Parties’ customers are considering during their
purchase journey and would divert to in the event the Parties were
unavailable. Additionally, the Parties noted that Google Trends data for
searches related to ‘blinds’ also identifies Ikea and Argos.
The Parties also submitted that M2M blinds purchases are increasing
primarily at the expense of ready-made blinds (as well as M2M blinds
through other distribution channels) and that Blinds2Go’s growth has been
based on persuading customers to choose online M2M blinds instead of
ready-made blinds.
141
The Parties submitted that a high proportion of the
Parties’ products have a ready-made equivalent and noted that the leading
ready-made blinds retailers offer, in total, 352 different size combinations,
with an additional 93 combinations available from eBay.
142
The Parties also
noted that if customers were only interested in blind dimensions, 48% of
Blinds2Go’s and up to 54% of 247’s products would be available in a
ready-made equivalent, while almost 40% of Blinds2Go’s and 45% of 247’s
would have a ready-made equivalent if customers wanted a particular type
of blind (e.g. roller or roman).
143
The Parties noted that these percentages
are likely to significantly understate the competitive constraint imposed by
ready-made blind retailers due to the analysis only including approximately
70% of the market, the ability for customers to modify ready-made blinds to
suit their needs, and the ability for blinds to be fitted outside the recess,
where tolerances for width and drop dimensions are much wider.
144
Further, the Parties submitted that, with respect to range, even if customers
viewed colour and/or pattern as an important feature, the majority of M2M
blind sales are in ‘neutral colours’ that are readily available in ready-made
alternatives.
Finally, the Parties submitted that ready-made blind alternatives are priced
very competitively compared with M2M and provided a comparison of
Blinds2Go’s average retail prices with ready-made average retail prices,
both by type of blind and by type of blind and retailer.
145
The Parties noted
that while there are differences in quality between M2M blinds and some
ready-made products, even for products with comparable quality, ready-
made blinds are still priced to undercut Blinds2Go’s M2M products.
146
141
Main submission, 20 May 2020, paragraphs 1.8 and 4.16.
142
Main submission, 20 May 2020, paragraph 6.68.
143
Main submission, 20 May 2020, paragraph 6.70.
144
Main submission, 20 May 2020, paragraph 6.71.
145
Main submission, 20 May 2020, paragraph 6.72 and Figures 6.15 and 6.16.
146
Main submission, 20 May 2020, paragraph 6.74.
73
Our assessment
In Hunter Douglas/Hillarys, the CMA adopted a product frame of reference
for the retail supply of M2M blinds, separate from ready-made blinds, on
the basis that the majority of third parties did not consider ready-made
blinds to be a substitute for M2M blinds and evidence indicating that the
conditions of competition for M2M and ready-made blinds are different.
147
For the reasons set out below, our view is that, consistent with the finding
in Hunter Douglas/Hillarys ready-made blinds should not form part of the
same product market as M2M blinds.
With respect to demand-side substitution, the BDRC Survey submitted by
the Parties shows that only 13% of respondents for Blinds2Go and 12% of
respondents for 247 would divert to ready-made blinds when asked what
they would do in a hypothetical situation where the respective Party had
stopped selling blinds.
148
These percentages are higher than the proportion
of respondents indicating they would divert to other window coverings,
suggesting that ready-made blinds are a more important constraint on the
Parties than other window coverings. The Parties submitted that these
percentages are very similar to the diversion (as implied by the BDRC
Survey) from Blinds2Go to 247, which the CMA called ‘significant’.
149
In
making this comparison, the Parties are comparing diversion to a single
competitor (ie 247) with the aggregate diversion to ready-made blinds sold
by a number of different retailers. However, in contrast, comparing the
proportion of respondents indicating they would divert to ready-made blinds
with the proportion of respondents indicating they would divert to other
retailers of online M2M blinds, namely 66% for Blinds2Go and 75% for 247,
we find that diversion to ready-made blinds is comparatively small. This
suggests that ready-made blinds are a significantly weaker constraint on
the Parties than online M2M blinds offered by other retailers.
Additionally, we note that the proportion of respondents that said that they
would divert to ready-made blinds is small when compared to the relative
size of sales of ready-made blinds. In particular, our market share
estimates show that sales of online M2M blinds by Blinds2Go’s competitors
amounted to £84 million in 2019 (see paragraph 8.10), while the Parties
estimated that sales of ready-made blinds amounted to £125 million in
2019. Despite ready-made sales being significantly larger, Blinds2Go’s
customers were substantially more likely to switch to other retailers selling
147
Hunter Douglas/Hillarys, paragraphs 64 to 65.
148
Main submission, 20 May 2020, paragraph 6.59.
149
Parties' response to Provisional Findings, 7 August 2020, paragraphs 4.24 and 4.94.
74
online M2M blinds than to ready-made blinds. This suggests that ready-
made blinds would at best be a distant competitive constraint.
150
The Parties submitted that this assessment fails to account for the
significance of a customer’s journey, noting that the BDRC Survey is a
survey only of those customers who ultimately ended up choosing an
online M2M blinds retailer and that these customers are more likely to
consider another M2M blinds retailer.
151
However, responses from
customers that ultimately decide to purchase online M2M blinds are highly
relevant, as it is the behaviour of those customers (namely their willingness
to switch to ready-made blinds) that determines the extent to which ready-
made blinds constrain M2M blinds. We therefore find that it is appropriate
to focus on the responses of those customers.
While we recognise that, as the Parties have highlighted, the BDRC Survey
shows that a material proportion of respondents considered purchasing
ready-made blinds instead of M2M blinds, the fact that customers
‘considered’ ready-made blinds does not allow for strong conclusions
regarding their willingness to switch and hence the competitive constraint
ready-made blinds exert on M2M blinds.
152
This is consistent with the view
that customers may consider the option of ready-made blinds but ultimately
decide at an early stage that they are not suitable for their requirements.
With respect to the Parties’ submission that Google Trends data for
searches related to ‘blinds’ also identifies Ikea and Argos, we have not
received any evidence that this data is driven by customers considering
both ready-made and M2M blinds as part of their purchase journey. In fact,
we note that the BDRC Survey shows that only 1% or less of respondents
visited the website of Ikea or Argos prior to their purchase of online M2M
blinds and that only a very small proportion of respondents would switch to
Ikea and Argos. We therefore consider that Ikea and Argos appearing in
the Google Trends data is likely to be driven, at least to an extent, by
customers looking exclusively for ready-made blinds. Additionally, Ikea and
Argos are not formally monitored by any online M2M blind retailer.
Third party views on the propensity of customers to switch to ready-made
blinds tended to suggest that ready-made blinds should not form part of the
same relevant product market as M2M blinds. In particular, of the retailers
150
In particular, if ready-made blinds were as close a competitor as online M2M blinds, we would expect
diversion rates to be in line with the magnitude of sales, ie we would expect higher diversion to ready-made
blinds than to online M2M blinds.
151
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.22.
152
We do no dispute that the evidence on what customers ‘considered’ is objective. However, we find that the
word ‘consider’ does not reveal whether a customer would switch to the product considered in case of a price rise
or in case its preferred option was not available.
75
that expressed a view and sell both categories of blinds, [].
153
[]
154
Notably, 247 also mentioned that customers that ended up buying M2M
blinds from 247 may well need M2M, which suggests that these customers
would not switch in response to a SSNIP.
We further assessed the propensity of customers to switch by considering
to what extent ready-made blinds offer the same dimensions as M2M
blinds. In this regard, we note that evidence from customers and third
parties suggests that the need for specific dimensions is a key reason for
customers to choose a M2M blind. For example, the BDRC Survey shows
that 37% of Blinds2Go’ respondents and 36% of 247’s respondents
selected ‘specific dimensions needed made-to-measure’ as a reason for
purchasing M2M blinds rather than alternative window coverings.
155
Similarly, two M2M blind retailers – both of which sell both M2M and ready-
made blinds – []. No third party provided a view to contradict this.
When counting those ready-made blinds that have exactly the same
dimensions, we find that only around 2% of the Parties’ sales of M2M blind
have a ready-made equivalent. If we allow for a tolerance of +/-1cm for
each of width and length, around 5% of the Parties’ sales of M2M blinds
have a ready-made equivalent.
156
In contrast, the Parties submitted that
40% of Blinds2Go’s and 45% of 247’s sales of M2M blinds have a ready-
made equivalent. The analysis on which the Parties base these results
applies different tolerances in terms of which sizes would still constitute an
‘equivalent’, with tolerances depending on the type of blind and whether the
blind is fit inside or outsides the recess. The Parties submitted that these
tolerances are very conservative.
157
However, at least for some blinds, the
applied tolerances appear to be very generous and we have not received
153
[].
154
Interior Goods Direct, [], Swift Direct Blinds and MakeMyBlinds all submitted that customers would not
Switch. However, Concept Systems noted that customers would possibly switch, while Wilsons Blinds submitted
that customers would switch.
155
Similarly, market research submitted by Swift Direct Blinds shows that ‘made to measure’ was selected as the
single most important factor in the respondent’s decision making by the highest proportion of respondents
(namely 25.4%) and further scored highest in terms of importance to the respondents’ buying decision.
156
Our analysis is based on the sales data submitted by the Parties in response to the CMA Questionnaire of the
14 May 2020, Annex 0114 and Annex 0130 and on the measures of ready-made blinds available submitted by
the Parties as Annex 0102 to the main submission. We note that the Parties submitted that the measures of
ready-made blinds available only consider the 11 top ready-made blinds retailers (representing the 80% of the
ready-made market in 2018). With respect to the Parties’ argument that only the 10 top ready-made blind
retailers (representing approximately 70% of the market in 2018) were included in the analysis, we do not think
that this is likely to substantially bias the results, as we would reasonably expect the 10 top ready-made blind
retailers to most of the available sizes.
157
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.100.
76
any evidence that customers would be willing to accept such tolerances.
158
159
We also note that the claimed availability of ready-made equivalents calls
into question why customers purchase M2M blinds when ready-made
blinds are considerably cheaper even for comparable quality (see
paragraph 7.44). The Parties submitted that a narrow focus on price
ignores the other factors that some customers value (including easy to use
website, product quality, and product range). However, and as described in
paragraph 7.44, the price difference persists even for products with
reportedly comparable quality. We also note that the Parties themselves
argued that product range is of limited importance,
160
and that the Parties
have not provided any evidence to show that the listed factors are indeed
causing the observed price differences. Overall, we find that the Parties
have not convincingly explained what warrants the significant price
differences we observe.
We note the Parties’ argument that M2M blinds are just cut-down ready-
made blinds and that customers are able to modify ready-made blinds to
suit their needs. However, it is irrelevant for a customer if the retailer from
which the customer purchases the product obtains the product by
modifying an existing ready-made product or by sourcing an entirely new
M2M product. With respect to the prospect of the customer modifying the
ready-made product themselves, the Parties submitted that many tutorials
with step-by-step guidance are readily available online, and that these
videos/tutorials would not be so popular (and retailers would not go to the
effort of producing them) if customers were not willing to modify the ready-
made blinds themselves.
161
The Parties further submitted that the latest
AMA report highlights how many customers are willing to alter ready-made
blinds.
162
We acknowledge that cutting ready-made blinds to size may be an option
for some customers – and this is consistent with the BDRC Survey showing
that 13% of respondents for Blinds2Go and 12% of respondents for 247
would divert to ready-made blinds. However, the Parties’ submission on
tutorial availability and number of views does not show how widespread the
willingness to cut ready-made blinds is. Moreover, the Parties have not
158
Depending on the type of blind, the positive width or drop tolerance amounted to up to 50cm.
159
Additionally, we note that, as explained by the Parties, the analysis for 247 relies on applying the uplift factor
from Blinds2Go between inside recess vs outside recess. Without this modification, only 26% of 247’s sales of
M2M blinds have a ready-made equivalent (if otherwise following the Parties’ methodology).
160
The Parties submitted that sales are concentrated among a relatively small number of popular blinds and that
range as a competitive constraint should not be overstated; Parties’ response to Provisional Findings, 7 August
2020, paragraphs 4.62 and 4.85.
161
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.102.
162
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.102.
77
provided any evidence that any such willingness poses a competitive
constraint on their online M2M blind businesses. The only third-party
retailer of M2M blinds to provide a view on customers altering ready-made
blinds told us that, although some ready-made blinds can be cut to size,
there is a question about whether a customer is likely to do this effectively.
Additionally, we note that a widespread willingness to cut ready-made
blinds to size would call into question why customers purchase M2M blinds
when ready-made blinds are considerably cheaper even for comparable
quality.
In addition to the evidence on relative dimensions and the survey evidence,
the []
163
- although the Parties submitted that the majority of customers
purchase ‘commodity products’ (by which the Parties appear to mean
standard products).
164
A comparison of the number of stock keeping units (SKUs) for each of
M2M blinds and ready-made blinds that are offered by different online
retailers suggests that [].
165
166
We consider that this may further limit the
competitive constraint from ready-made blinds on M2M blinds.
The Parties submitted in this regard that []% of M2M blinds sold in March
2020 (by value) were blinds in neutral colours which are readily available
as ready-made blinds. However, we consider that the Parties’ assessment
of ‘neutral colours’ does not take into account nuances in colour, patterns,
fabrics and other options (such as DuoShade
TM
or DuoLuxe) available for
M2M blinds. These variables allow for a degree of customisation in the
product that otherwise would not be available in ready-made options. As
such, we are not convinced that this submission demonstrates that there is
not significantly less choice with respect to ready-made blinds.
We also considered how the prices of M2M blinds and ready-made blinds
compare. The Parties submitted a comparison of Blinds2Go’s prices for
M2M blinds with other retailers’ prices for ready-made blinds for a number
of different types of blinds (including the key types, ie venetian, roller,
vertical and roman blinds). This comparison indicates that M2M blinds are
163
The BDRC Survey highlighted that a more suitable design was the key purchase reason for the majority of
both Parties’ customers. Moreover, ‘had what specifically wanted’ and ‘good/wide product range’ was selected by
a significant proportion of respondents as a reason influencing the choice of retailer. BVA-BDRC Blinds Survey
Final Report 20 May 2020, Main submission, 20 May 2020, Annex 0093. Market research submitted by Swift
Direct Blinds further shows that ‘good selection of colours’ was the product unique selling proposition that
obtained the second highest score with respect to the impact it would have on buying decisions, Swift Direct
Blinds Brand Bible V3, page 7. [].
164
Main submission, 20 May 2020, paragraph 6.17.
165
We note that while e.g. different colours and fabrics constitute different SKUs, different sizes do not constitute
different SKUs. Our comparison takes into account the Parties for M2M blinds and Dunelm, John Lewis, Next and
Argos for ready-made blinds (i.e. the key retailers that offer online ready-made blinds excl. Amazon).
166
[].
78
significantly more expensive than ready-made blinds. We note that there
are a number of examples where Blinds2Go’s average price for a specific
type of blind is more than twice as high as the average ready-made price of
a specific retailer for the same type of blind.
167
Additionally, even for
products with reportedly comparable quality (the Parties indicate that this
would be products sold by John Lewis and Dunelm),
168
Blinds2Go’s M2M
blinds are still significantly more expensive than the equivalent ready-made
blinds: the lowest price difference amounts to £9 (or 19%) for John Lewis’
venetian blinds, with price differences amounting to more than £50 (or
more than 100%) for John Lewis’s and Dunelm’s roman blinds. While price
differences in and of themselves do not necessarily imply that products are
not part of the same relevant market, we consider that such large-scale
price differences, especially for products of similar quality, do indicate that
ready-made blinds do not compete directly with M2M blinds.
We also considered the extent to which the Parties and other retailers of
M2M blinds monitor the prices of ready-made blinds. Such monitoring
appears to be limited, suggesting that M2M blinds retailers do not consider
ready-made blinds to be an important competitive constraint to M2M blinds.
(a) While the Parties [].
169
(b) [].
170
(c) All other online M2M blind retailers that provided evidence on the
subject told us that they do not monitor the prices of ready-made
blinds and that they do not consider pure ready-made blind retailers
as competitors.
With respect to supply-side substitution, we note that suppliers of online
M2M blinds appear to be distinct from suppliers of ready-made blinds:
(a) None of the four largest online retailers of M2M blinds (ie Blinds2Go,
247, Interior Goods Direct and Swift Direct Blinds) sell ready-made
blinds, and more generally, none of the online M2M retailers we
contacted sell ready-made blinds. We also note that there are
differences further up the supply chain, with Decora, a key
manufacturer of M2M blinds, not manufacturing any ready-made
blinds.
167
We note that the price comparisons submitted by the Parties in phase 1 show a similar picture of M2M blinds
tending to be substantially more expensive than ready-made blinds, although the Parties argued ‘this price
differential has materially fallen’. Parties’ substantive submission, paragraph 4.18, 20 December 2019.
168
Main submission, 20 May 2020, paragraph 6.74.
169
[].
170
[].
79
(b) While Hunter Douglas told us that there is a very large business in
the US of providing cut-down services in the stores, the fact that the
largest ready-made retailers have a relatively weak position with
respect to online M2M blinds suggests that this is not a common
phenomenon in the UK, nor has Hunter Douglas provided any
evidence that it will be in the near-future.
Indeed, Hunter Douglas
told us that Dunelm is not currently doing this in the UK.
Finally, in response to the Parties’ argument that M2M blinds purchases
are increasing primarily at the expense of ready-made blinds and that
Blinds2Go’s growth has been based on persuading customers to choose
online M2M blinds instead of ready-made blinds,
171
in our view this does
not imply that ready-made blinds pose a competitive constraint on M2M
blinds such that they should be included in the same relevant market:
(a) First, even if customers would move from ready-made blinds to M2M
blinds in response to a SSNIP, this would only indicate that M2M
blinds exercise a competitive constraint on ready-made blinds, but
not vice versa. By the same token, we do not consider that the
Parties’ submission that [] shows that ready-made blinds constrain
M2M blinds.
172
173
(b) Second, a migration to M2M blinds does not support the position that
customers would switch to M2M blinds in response to a SSNIP.
Overall, we find that ready-made blinds should not form part of the same
product market as online M2M blinds. Notwithstanding this view, we
acknowledge that ready-made blinds do act as a distant competitor to
online M2M blinds and we have taken any out-of-market constraint exerted
by ready-made blinds into account in the competitive assessment section,
see paragraphs
8.241 to 8.252.
171
We note, however, that we have received mixed evidence on this. See for example paragraph 4.10, which
indicates higher rates of growth ready-made products as compared to M2M.
172
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.97.
173
[], at most shows that M2M blinds constrain ready-made blinds, but it does not show conclusively whether
customers of M2M blinds would switch to ready-made blinds and whether ready-made blinds exercise a
competitive constraint on M2M blinds.
80
In-store and in-home
The Parties’ view
The Parties submitted that they currently compete directly with offline
retailers of M2M blinds, both in-store and in-home.
174
In this regard, the
Parties made the following points:
(a) First, the Parties submitted that there has been some substitution
away from the offline channel to the online channel, arguing that this
implies that customers are increasingly seeing these channels as
readily interchangeable.
175
(b) Second, the Parties submitted that Blinds2Go offers a range of
products that compete with in-store and in-home products both in
terms of price and quality.
176
Additionally, the Parties submitted that customers consider and interact
with different channels as part of their purchase journey.
177
They submitted
that the BDRC Survey shows that the Parties’ customers also considered
purchasing in-store and that 19% of Blinds2Go’s customers and 24% of
247’s customers visited physical stores before placing their online orders.
178
At least with respect to in-store, the Parties argue that this suggests that
online retailers are constrained by the window covering offers available
through the offline channel.
179
Our assessment
In Hunter Douglas/Hillarys the CMA adopted, on a cautious basis, separate
product frames of reference for the retail supply of M2M blinds by sales
channel (ie separate frames of reference for online, in-store and in-home).
The CMA’s conclusion in that case was based on evidence from the
merging parties’ internal documents as well as third parties, which indicated
that there was a distinction between online, in-store and in-home sales. In
particular, third-party evidence in that case indicated that constraints for
online retailers come from other online retailers rather than from physical
stores (ie in-store) and in-home.
180
174
Main submission, 20 May 2020, paragraph 6.76.
175
Main submission, 20 May 2020, paragraph 6.76.
176
Main submission, 20 May 2020, paragraph 5.12.
177
Main submission, 20 May 2020, paragraphs 4.7 and 6.44.
178
Main submission, 20 May 2020, paragraphs 1.8 and 6.43.
179
Main submission, 20 May 2020, paragraph 6.44.
180
Hunter Douglas/Hillarys, paragraphs 72 to 79 and 80 to 84.
81
For the reasons set out below, our view is that, consistent with the finding
in Hunter Douglas/Hillarys , M2M blind sales through in-store and in-home
channels should not form part of the same product market as M2M blinds
sold through the online channel.
In relation to demand-side substitutability, the BDRC Survey submitted by
the Parties indicates that the constraint from the in-store and in-home
channels on the Parties’ online M2M blinds offering is weak. When asked
what they would do in a hypothetical situation where the respective Party
had stopped selling blinds, only 7% of respondents for Blinds2Go and only
4% of respondents for 247 would divert to the in-store M2M channel. The
equivalent figures for the in-home channel are 6% for Blinds2Go and 3%
for 247.
The Parties submitted that store closures in the context of COVID-19 may
have affected these answers and that the survey responses are therefore
likely to understate in-store diversion (ie understate the number of
customers that would switch to in-store if their current option was no longer
available).
181
The Parties further submitted that the CMA fails to give due
weight to this.
182
We acknowledge that store closures at the time at which
the survey was conducted may have affected survey results. However, the
survey was carefully drafted to explore consumers’ recent purchase
journeys pre-COVID-19 (as the Parties point out),
183
and therefore
designed to minimize any bias resulting from store-closures.
We recognise that, as the Parties have highlighted, the BDRC Survey
indicates that a material proportion of online customers visited physical
stores before placing their online orders (looking for a similar product or
comparing the price of a similar product).
184
However, we consider that
customers visiting stores does not necessarily imply that customers would
switch to the in-store channel in response to a price increase and that the
in-store channel constrains online retailers. Indeed, and as discussed in
paragraph 7.52 above, the survey shows that the Parties’ customers are
unlikely to switch to the in-store channel.
Evidence collected from online M2M blind retailers during the course of our
investigation shows relatively low levels of engagement with retailers in
other channels, which indicates that in-store and in-home retailers are
181
Main submission, 20 May 2020, paragraph 6.62.
182
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.80.
183
Main submission, 20 May 2020, paragraph 6.62.
184
With respect to the Parties’ submission that that the Parties’ customers consider purchasing in-store, we note
that the BDRC Survey does not appear to explore whether customers considered other channels. In any case,
even if customers did ‘consider’ purchasing in-store, this would not allow for strong conclusions regarding their
willingness to switch to purchasing in-store and hence the competitive constraint thein store-channel exerts on
M2M blinds.
82
unlikely to pose a significant competitive constraint on online M2M blind
retailers. [].While the Parties submitted that Blinds2Go also frequently
visits physical stores that sell M2M blinds, [].
185
Second, the Parties
submitted that they are not able to estimate market shares for the in-store
M2M segment because they are not active in this channel.
[].
186
As part of our investigation, we also compared the retailers active in the
different (ie, online, in-home, in-store) channels. This comparison revealed
significant differences in the business model and service proposition of
online retailers compared to in-store and in-home retailers. Both in-store
and in-home retailers appear to position themselves as full-service
providers, offering a more personal experience (including in-person advice)
and the option of an at-home measuring and installation service. In
contrast, online retailers appear to be more focused on price, with the
evidence we have seen showing that online retailers tend to be
substantially cheaper than in-store retailers:
(a) Blinds2Go states on its website that it compared its prices with the
five main M2M blind retailers on the high street and that ‘in almost
25% of cases we were cheaper by 60% or more’. It further
advertises on its website that ‘you can also save up to 60% off High
St. prices’.
187
(b) John Lewis noted that in relation to price, its products tend to fall
within the mid-range to higher end of the market, due to its use of
higher quality materials relative to some of its competitors and its
higher overheads compared to online competitors.
However, Next told us that it charges the same prices for its M2M blinds
online and in-store, with the price comparison submitted by the Parties in
phase 1 suggesting that Next prices are towards the higher end compared
to other online M2M retailers.
We do not consider that differences in business model and service
proposition necessarily imply a lack of competitive constraint, given that
customers may, in principle, be willing to substitute between paying more
for more service and less for less service. A similar point is made by the
Parties themselves, as they argue that customers are likely to be making a
quality-price trade-off. However, we note that the Parties also submitted
185
[].
186
[].
187
Blinds2Go website, accessed 23 June 2020.
83
that ‘in-home and in-store products are considered over-priced when faced
with a comparable product at significantly lower price which can be
measured and fitted with little effort.’
188
With respect to supply-side substitution, we note that suppliers’ market
positions in the different channels appear to be materially different. In
particular, whilst some in-store retailers have established an online
presence in the sale of blinds (both ready-made and, to a lesser extent,
M2M), we observe that none of the in-store or in-home retailers have a
significant position within the online M2M blinds market. In particular, we
note that in-store retailers selling M2M blinds online typically have a much
smaller product range than exclusively online retailers. Moreover, we also
note that some in-store and in-home retailers do not have any online sales.
For example, for the in-home retailers Hillarys and Thomas Sanderson,
their websites only provide lead generation and enable customers to
provide contact details and request order samples.
189
Additionally, the
multi-channel retailer [].
190
Finally, we are unconvinced by the Parties’ argument that M2M blinds
purchases are increasing due to substitution away from the in-store and in-
home channels to the online channel. We do not consider that the Parties
submission implies that offline channels pose a competitive constraint on
the online channel:
(a) First, a migration to online M2M blinds could reflect a shift in
customer preferences, rather than a result of substitution, and
therefore does not prove that customers would switch to online M2M
blinds in response to a SSNIP. The Parties have not provided any
evidence that supports that the observed migration is a result of
substitution.
(b) Second, even if customers would move from the in-store and in-
home channels to the online channel in response to a SSNIP, this
would only show that the online channel exercised a competitive
constraint on the offline channel, but not vice versa.
Overall, we find that the in-store and in-home channels should not form part
of the same product market as online M2M blinds. We take into account
any out-of-market constraint exerted by M2M blinds sold through the in-
188
Main submission, 20 May 2020, paragraph 5.16.
189
Hillarys and Thomas Sanderson websites, accessed on 23 June 2020.
190
[].
84
store and in-home channels in the competitive assessment section, see
paragraphs 8.241 to 8.252 below.
Geographic market definition
The Parties submitted that any retail product market identified should be
considered national in scope.
In the SLC Decision, the CMA assessed the online retail supply of M2M
blinds on a national basis.
191
We have not received any evidence to suggest that an alternative
geographic market definition would be more appropriate.
We therefore find that a national market definition is appropriate.
Conclusion on market definition
For the reasons discussed above, we find that the appropriate market
definition in this case is the online retail supply of M2M blinds in the UK.
191
Hunter Douglas/Hillarys, paragraph 99.
85
8. Competitive assessment
In this section, we assess the competitive effects of the 2019 Transaction
as they relate to the online retail supply of M2M blinds in the UK. To inform
this assessment, we have assessed whether, as a result of Hunter
Douglas’ ability to unilaterally determine all aspects of 247’s competitive
strategy (including the ability to set 247’s prices), as well as its increased
interest in the profits of 247, the 2019 Transaction would be likely to
substantially lessen competition by resulting in the Merged Entity
increasing prices, lowering the quality of its products or customer service,
or reducing the range of its products/services.
192
This is a horizontal
unilateral effects theory of harm.
This section is structured as follows:
(a) We first discuss our market share estimates for the Parties and their
competitors in the online retail supply of M2M blinds in the UK.
(b) We then describe the key characteristics of competition with respect to the
retail supply of online M2M blinds.
(c) We assess the closeness of competition between the Parties.
(d) We assess the extent of the competitive constraint on the Parties from
other competitors.
(e) Finally, we set out our assessment of the impact of the Merger on
competition.
Market shares
Market shares can give an indication of the potential extent of a firm’s
market power. The combined market shares of the merger firms, when
compared with their respective pre-merger market shares, can provide an
indication of the change in market power resulting from a merger.
193
While
market shares do not provide a full picture, they can give an indication of
the presence and size of the key competitors within the market as defined.
In this section we consider the market shares of the Parties and other
suppliers in the market for the online retail supply of M2M blinds in the UK.
192
MAGs, paragraphs 4.2.3 and 5.4.1.c.
193
MAGs, paragraph 5.3.4.
86
Approach to market share calculations
The Parties submitted market share estimates for the online retail supply of
M2M blinds in the UK in 2019 based on their own sales data and their
estimates for the sales of third parties. With respect to third-party sales
estimates, the Parties initially estimated the sales of third parties for 2018
and then applied the growth rate of Blinds2Go in 2018 to these estimates to
obtain 2019 values.
194
We were able to develop this methodology further by using third parties’
actual sales data for 2019, which is not available to the Parties.
195
While we
contacted all suppliers of online M2M blinds for which the Parties provided
contact details, there are, however, retailers listed in the Parties’ market
share estimates from whom we did not receive responses and hence do
not hold actual sales data.
Despite this lack of actual sales data for some retailers, we still consider it
most appropriate to calculate market shares solely based on the actual
sales data we hold:
(a) While we reviewed information contained in market reports, we have not
received any market reports that contain a market size estimate for online
M2M blinds. We also find that it is not possible to obtain a reliable
estimate for online M2M blinds in the UK by triangulating different market
reports.
(b) While the Parties submitted estimates for the retailers from whom we did
not receive responses, we consider that the Parties’ estimates are likely to
be highly inaccurate and therefore not reliable. In particular, a comparison
of the Parties’ estimates with actual sales data for suppliers from whom
we did received responses revealed that, in some cases, the Parties’
estimates were overstating the actual sales by []. For example, the
Parties estimated that Swift Direct Blinds’ sales of online M2M blinds in
2019 amounted to £10.7 million, while Swift Direct Blinds submitted that
its total sales of online M2M blinds in 2019 were lower than the Parties’
estimates at £[] million. Similarly, the Parties estimated that eBay’s
sales of online M2M blinds in 2019 amounted to £34.8 million[].
196
194
The sources used by the Parties in these estimates include publicly available information on the retailers’
liabilities and assumptions on the minimum turnover necessary to be viable or deriving sales from the number of
reviews on Amazon and eBay.
195
The actual sales data from third parties in our shares incorporates all retailers that the Parties identified as
their ‘main competitors’ and that, based on the Parties’ estimates, had sales of more than £2 million in 2019 (with
the exception of Powered Blinds, a retailer only supplying electric blinds).
196
[]. We note that, in the context of the CMA investigation on Hunter Douglas/Hillarys, an email exchange
started by David Sonnenberg (Co-President & CEO of the Hunter Douglas Group) and Mark Bramley (Blinds2Go
Director) shows that Hunter Douglas used an approach to the eBay sales estimation similar to the one used in
87
While, due to the inaccuracy of the Parties’ estimates, we do not include
these estimates in our market share calculations, we nevertheless
assessed what the Parties’ position would have been had we included
these estimates (see sensitivity analysis discussed in paragraph 8.11).
Overall, and for the reasons set out above, we consider that our market
share estimates, calculated using actual sales data, are more reliable than
the Parties’ estimates. Nonetheless, as part of a sensitivity analysis, we
have considered the impact of including the Parties’ sales estimates for the
retailers for whom we did not received actual sales data.
Market share estimates
Table 3 below shows the estimated market shares for the online retail
supply of M2M blinds in the UK for 2019.
Table 3: 2019 market shares for the online retail supply of M2M blinds in the UK
£m
%
Retailers
Revenues
Shares
Blinds2go (incl. Web Blinds)
[]
[50 - 60]
247
[]
[5 - 10]
Combined
[]
[60 - 70]
Interior Goods Direct (incl. Wilsons)
[]
[10 - 20]
Swift Direct Blinds
[]
[0 – 5]
Bloc Blinds
[]
[0 – 5]
Next
[]
[0 – 5]
MakeMyBlinds
[]
[0 – 5]
Dunelm
[]
[0 – 5]
Order Blinds Online Ltd
[]
[0 – 5]
Blinds4UK
[]
[0 – 5]
Meadow Blinds Ltd / Lifestyle Blinds Ltd
[]
[0 – 5]
John Lewis
[]
[0 – 5]
eBay (incl. ready-made and shutters)
[]
[5 - 10]
Amazon Marketplace (incl. ready-made)
[]
[5 - 10]
Others (with turnover <0.5m)
[]
[0 – 5]
Total
175.9
100.0
.
Notes: The Parties’ sales exclude any sales from Velux, namely £[] for Hunter Douglas and £[] for 247.
From these market share estimates, we note that:
(a) Blinds2Go is by far the largest retailer and is several times larger than
Interior Goods Direct, the second largest retailer;
the current investigation. In the same email exchange, John Andrews, proposing a methodology for estimating
eBay’s sales, writes: ‘I'll try again with a different tool and see if I am able to get a bigger number’.
88
(b) 247 is the third largest retailer and is approximately three times larger
than Swift Direct Blinds, the fourth largest retailer;
197
(c) The combined market share of the Parties is very high, and the increment
from 247 is significant in the context of an already concentrated market;
(d) Other than the Parties and Interior Goods Direct, there are no other online
M2M retailers with a market share above 5%, and few retailers with a
market share above 1%;
(e) None of the multi-channel retailers has a market share above 2%; and
(f) While the market shares of the marketplaces eBay and Amazon are not
insignificant, their shares are likely to be significantly over-estimated as
sales of ready-made blinds (and in the case of eBay also shutters) are
included, and these are likely to account for a large proportion of their
sales.
198
Further, we note that, as marketplaces, their share is attributable
to collections of individual retailers rather than reflecting their share as a
single retailer; as such, their individual shares are likely to be much lower.
Both of these factors mean that the shares of Amazon and eBay in the
above table are likely to significantly overstate their competitive constraint.
The Parties submitted that it is not reasonable to assume that the sales of
other competitors (ie retailers for whom we do not hold actual sales data)
are zero.
199
The Parties also submitted that adding in these omitted
retailers would reduce the Parties’ combined market share significantly.
200
As noted above, we have conducted a sensitivity assessment of our market
share estimates which includes these other retailers. Based on the Parties’
estimates, these retailers would jointly account for only £39 million in
revenue. Even if we added this £39 million to the total market size, the
Parties’ combined market share would still exceed 50%. As such, we find
that the addition of these retailers does not significantly change the position
of the Parties or the competitive landscape set out in paragraph 8.10
above.
201
The Parties argued that Velux should be included in the market share
calculations as it is a competitor to the Parties. However, we do not agree
because Velux sales do not appear to constitute a M2M product. In
197
Although Amazon and eBay appear to have larger shares than Swift Direct Blinds in Table 3, we note that
these are marketplace retailers, not a single retailer, and that their respective shares are likely to be significantly
over-stated, as set out at 8.10(f).
198
While unable to provide a clear split, eBay told us that M2M blinds account for only a small proportion of its
overall blind sales. In particular, eBay noted that only around 18% of blind listings on eBay relate to M2M blinds.
199
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.7.
200
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.8.
201
As set out in paragraph 8.191 below, we also note that the BDRC Survey shows that there is very low
diversion from the Parties to any of these additional retailers.
89
particular, Velux told us that it does not sell any M2M blinds and that it
considers its products to be ready-made rather than M2M products. This is
consistent with the Parties’ own submission that the dimensions of Velux
blinds are determined by a window code, rather than being specified by the
customer, which indicates that they are ready-made rather than M2M
products. Accordingly, we excluded Velux as a competitor from the market
share estimates and any sales of Velux products from the Parties’ sales.
202
While we did not calculate market shares for previous years, the CMA’s
assessment of the Hunter Douglas/Hillarys merger in 2017 showed that, in
2017, Blinds2Go was already, by far, the largest retailer of M2M blinds
online, with 247 and Interior Goods Direct being the next largest
respectively.
In sum, our assessment of market shares above indicates that the Parties,
together with Interior Goods Direct, are the largest suppliers of online M2M
blinds and that other suppliers are, by comparison, of much smaller scale.
Further, the Merged Entity is of significant scale in comparison to both
Interior Goods Direct and all other retailers in the market.
How competition works in the market
As discussed in previous sections (see paragraphs 4.22 to 4.25), the retail
supply of online M2M blinds involves customers measuring their window
and then ordering their preferred choice of blinds directly from retailers’
websites. As a result, a significant aspect of competition between retailers
of online M2M blinds occurs online.
In order to be successful, retailers of online M2M blinds need both to attract
customers to their websites, and to convert these visits into sales. A
number of factors determine whether a customer ‘converts’ once they have
navigated to a given website (including price, perceived quality of product,
quality of website, etc), but prior to that, a significant aspect of competition
in this market is competition for visibility in web search results, given the
importance of traffic from this channel in generating revenues.
203
In the section that follows, we assess the Parties’ submissions on the
customer journey and the importance of online traffic. We then present our
views on how competition for traffic works, including our assessment of the
202
We note that sales of Velux products may be included in the sales of third parties. However, we did not
consider this undermined our ability to give weight to the shares set out in Table 3, because the inclusion of any
Velux product sales would mean that the share estimates for the Parties were understated.
203
Search engines account for 75% of traffic by revenue across the online M2M retailers based on those
responses received by the CMA.
90
evidence we have seen on; sources of traffic for online M2M blind retailers,
their marketing spend, and customer search behaviour.
Parties’ submissions on the customer journey and importance of online
advertising
The Parties submitted that advertising in search engines is important for
their businesses, since, as small online mono-line retailers, they do not
have strong brand names, with this particularly being the case for 247.
204
However, the Parties also argued that customers’ purchasing journeys are
not limited to paid Google Ads for generic terms such as blinds,
205
and that
customers will conduct extensive research to decide upon the product that
they think will best suit their window, often visiting large home stores such
as Next, John Lewis and Dunelm, high street blind and curtain specialists
and online blind and curtain sites.
206
In support of their argument, the
Parties submitted the following points:
(a) The BDRC Survey shows that approximately three quarters of the Parties’
customers spend more than an hour researching and comparing products
before purchasing.
207
The BDRC Survey further suggests that customers
compare products / prices across many websites and that customers do
not just visit websites and physical stores as part of the purchase journey
(with a significant number of respondents also reading reviews online and
discussing with friends/family and a smaller number of respondents using
social media, browsing magazines and using in-home fitting
services).
208209
With respect to these survey results, the Parties argued
that, due to the survey being based only on the Parties’ customers
(including repeat customers), survey results would tend to under-report
the amount of ‘shopping around’ by customers.
210
(b) Google Analytics data shows that the average customer makes their
purchase 10 days after the first website interaction and after having visited
the website on three different occasions.
211
(c) The low session conversion (ie, proportion of single website visits that end
up in a purchase) the Parties achieve (ranging from []%) indicates that
204
Main submission, 20 May 2020, paragraph 6.27.
205
Main submission, 20 May 2020, paragraphs 6.27 and 6.28.
206
Main submission, 20 May 2020, paragraph 4.4.
207
Main submission, 20 May 2020, paragraph 6.22.
208
Main submission, 20 May 2020, paragraph 6.40.
209
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.28.
210
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.29.
211
Main submission, 20 May 2020, paragraph 6.22.
91
customers shop around before buying, and ultimately are not making
purchasing decisions based on Google prominence.
212
In assessing the Parties’ submissions, we have considered the importance
of traffic and marketing spend, as well as search behaviour and customer
journeys when purchasing online M2M blinds.
Traffic and marketing spend
There are several channels that can be used by customers to reach a
retailer’s website, including:
(a) Internet search results – arrivals at the website by entering search terms
into search engines (such as Google) and clicking on the results
generated. Internet search results can be either:
(i) Paid search – paid search is where the retailer has paid to feature
prominently in search results.
(ii) Organic search – results which are not influenced directly by
advertising expenditure (although retailers may improve their ranking
by investing in search engine optimisation (‘SEO’));
(b) Direct – Arrivals at the website by entering the retailer’s address in the
URL bar;
(c) Email – Referrals to the retailer’s website from email marketing sent to
customers;
(d) Referral – Referrals to the retailer’s website from external websites; and
(e) Affiliates – Referrals to the retailer’s website from external websites who
are paid commission.
For the reasons set out at paragraph 8.32 below, paid search is a
significant source of traffic for online M2M blinds retailers. On Google,
where the majority of this search activity takes place, there are several
options advertisers can select including:
213
(a) Google Ads – Links shown on both the top and bottom of a page of
search results. Advertisers set up ‘campaigns’ where they select which
search words they would like to target, as well as targeting specific
locations, devices, times of day, and customer profiles. Advertisers ‘bid’
212
Main submission, 20 May 2020, paragraph 6.36.
213
See Appendix E for more information.
92
on key words by indicating what they would be willing to pay for each click
on its advertisement. Advertisers pay every time a consumer clicks on
their advertisement – such advertising is referred to as ‘pay-per-click’
(PPC) advertising.
(b) Product Listing Ads (PLA) – Short advertisements that include images and
which are shown next to search results. Ads are for specific products
rather than brands or websites, and advertisers must provide Google a
products ‘feed’. Like Google Ads, this also works on a PPC basis –
advertisers pay Google every time a customer clicks on their
advertisement.
(c) Google Shopping – Like PLA, Google Shopping allows advertisers to
advertise products rather than websites or brands. Consumers access
Google Shopping results by clicking on the ‘shopping’ link at the top of the
search results page. Users can filter listings by price, product category,
and brand.
In order to understand the relative importance of the above marketing
channels, we have looked at two important sources of evidence, namely:
(a) the sources of traffic to a retailer’s website; and
(b) a retailer’s expenditure on marketing.
With respect to sources of traffic, it appears that, for each of the main
online M2M blind retailers, paid search generated the largest proportion of
traffic in 2019 (both by traffic volume and revenue).
214
More specifically, it
appears that:
(a) Paid search accounted for []% of revenue for Blinds2Go and []% of
revenue for 247 in 2019. This confirms that paid search is a significant
source of traffic for the Parties.
(b) While we do not hold data to split paid search into Google and other
search engines, the widespread use of Google indicates that traffic from
paid search almost exclusively comes from Google.
(c) Only a relatively small proportion of paid search revenue (less than 15%)
relates to Google Shopping and PLA.
214
In each of 2017 and 2018, paid search generated the largest proportion of traffic across the main online M2M
blind retailers in aggregate, while it ranked either first or second for each of these retailers individually in these
years.
93
After paid search, organic search is the next biggest source of traffic,
accounting for []% of revenue for Blinds2Go and []% of revenue for
247 in 2019.
215
Table 4: Proportion of revenues by channel in 2019
Channel
[]
[]
[]
[]
[]
Combined
Paid search
[
]
[
]
[]
[
]
[
]
56%
Organic
search
[
]
[
]
[
]
[
]
[
]
19%
Direct
[
]
[
]
[
]
[
]
[
]
10%
Email
[]
[]
[]
[]
[]
6%
Referral
[]
[]
[]
[]
[]
7%
Affiliates
[
]
[
]
[
]
[
]
[
]
1%
Other
[
]
[
]
[
]
[
]
[
]
2%
Total revenue
[
]
[
]
[
]
[
]
[
]
£[]
Source: CMA analysis of data from Parties and competitors Parties’ response to CMA Questionnaire, 14 May 2020, Annex
0115, Annex 0116 and Annex 131; [] response to CMA Questionnaires, 22 May 2020.
Notes: All retailers use the same attribution model
216
except [] – so the data is comparable against all retailers except
[].Retailers’ revenues include VAT and shipping costs, except for []. However, this should not affect the distribution across
channels.
Revenue from samples is unlikely to affect these results.
217
While we do not hold the same source-of-traffic data on multi-channel
retailers, the third-party evidence we have obtained indicates that paid
search is an important source of traffic for them, too. For example:
(a) [].
218
(b) [] told us that it occasionally bids for key search words related to M2M
blinds to help improve customers’ awareness, but that its marketing is
often more general.
219
(c) Next told us that it does not use paid search specifically for its blinds, it
uses it in a more general way for its entire product range.
220
The Parties submitted that multi-channel retailers and marketplaces do not
have to rely on PPC advertising to generate traffic to their websites. In
particular, the Parties argued that multi-channel retailers receive many
direct visits and a lot of traffic through brand searches and that 90% of UK
shoppers use Amazon and 70% of these use Amazon as their first point of
215
A relatively high proportion of organic search does not necessarily suggest good performance in organic
search, but could also be an indication of bad performance in other channels.
216
An attribution model is ‘the rule, or set of rules, that determines how credit for sales and conversions is
assigned to touchpoints in conversion paths.’ See https://support.google.com/analytics/answer/1662518?hl=en
217
The only case in which this could potentially affect our analysis is if samples are mainly requested through
some particular channels versus others. However, revenues from samples are likely to constitute a very small
percentage of revenues and, therefore, this is unlikely to create any significant bias.
218
[]
219
[].
220
[].
94
call.
221
We agree that, in general, multi-channel retailers obtain a significant
number of direct visits to their website. However, the Parties have not
explained to what extent this also specifically applies to customers wanting
to purchase online M2M blinds, nor have they submitted any evidence
specific to traffic generation for online M2M blinds. It is therefore important
to take into account the comments from multi-channel retailers – and on the
basis of these comments, PPC advertising appears to be an important
source of traffic. Overall, we therefore find that the evidence suggests that
PPC advertising is an important source of traffic for multi-channel retailers,
albeit less important than for online M2M blind retailers.
In relation to retailers’ expenditure on marketing, Table 5 sets out the main
online M2M blind retailers’ expenditure by marketing channel. Consistent
with the data on traffic, we also find that, for each of the main online M2M
blind retailers, Google PPC accounted for the majority of marketing spend
in 2019, often amounting to around or more than 80% of their total
marketing spend. While there is some variation over time, Google PPC is
always by far the largest source of spending. Further details are set out in
Appendix E.
Table 5: Proportion of marketing spend by channel in 2019
Channel
[]
[]
[]
[]**
[]
[]
Google PPC
[]
[]
[]
[]
[]
[]
Bing/Yahoo PPC
[
]
[
]
[
]
[
]
[
]
[
]
Facebook
[
]
[
]
[
]
[
]
[
]
[
]
Agency
[
]
[
]
[
]
[
]
[
]
[
]
Other categories
[
]
[
]
[
]
[
]
[
]
[
]
Total
[
]
[
]
[
]
[
]
[
]
[
]
Source: CMA analysis of data from Parties and competitors. Parties’ response to CMA Questionnaire, 17 April 2020 and 14
May 2020, Annex 0007, Annex 0054(Revised), Annex 0115, and Annex 0116; Interior Goods Direct, Swift Direct Blinds and
MakeMyBlinds response to CMA Questionnaires, 22 May 2020.
Notes: []** []
The proportion of Google Shopping and PLA spend varies by retailer, with
[] and [] spending 5% or less of their online marketing expenditure on
these categories, Web Blinds, 247 and [] around []% and Blinds2Go
almost []%.
In 2019, marketing expenditure by the main online M2M blind retailers as a
proportion of turnover tended to be around 15% (or higher for []), with the
exception of [] spending only around 9%.
Overall, the above source-of-traffic and marketing spend analysis shows
the importance of online marketing for online M2M blind retailers. Paid
221
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.35.
95
search appears to be the most important source of traffic, particularly for
the Parties for whom traffic obtained through paid search accounts for over
half of their revenues. Organic search appears to be the second most
important source of revenues for online M2M blind retailers. Consistent with
the source-of-traffic data above, it appears that, for each of the main online
M2M blind retailers, Google PPC accounted for the majority of marketing
spend in 2019, often amounting to around or over 80% of their total
marketing spend.
Search behaviour and customer journey
Given the importance of online search for retailers of online M2M blinds, we
have reviewed the following sources of evidence on customer search
behaviour:
(a) the CMA’s literature review on online search;
222
(b) the EC’s Google Search (Shopping) decision
223
; and
(c) The BDRC Survey submitted by the Parties.
Both the CMA’s literature review and the EC’s Google Shopping decision
consider general search behaviour, rather than searches specific to blinds.
However, the BDRC Survey specifically addresses customer search
behaviour in relation to M2M blinds sold online.
The CMA’s literature review considers a number of different search tools,
including search engine results (assessing the entire page of search
results, ie jointly assessing paid and organic search) and finds that, despite
searching on the internet appearing to be fairly easy and simple, customers
consider on average 2.1 to 3.0 brands when wanting to purchase a product
online.
224
It further finds that there is evidence that customers spend more
time searching for complex or differentiated products.
225
While we
acknowledge that these findings do not necessarily mean that customers
only consider up to three retailers when searching for online M2M blinds,
222
CMA, April 2017, ‘Online search: Consumer and firm behaviour - A review of the existing literature’.
223
EC, June 2017, Case AT.39740, Decision on Google Search (Shopping). This decision focuses on the
positioning and display by Google, in its general search results pages, of its own comparison shopping service
compared to competing comparison shopping services.
224
The finding that customers consider on average 2.1 to 3.0 brands is based on evidence in paragraphs 4.9 to
4.12 of the CMA’s literature review. It constitutes the aggregate view based on a number of studies including a
variety of products such as, but not limited to, CDs, music, books, travel products, airline tickets, hardware,
phones, cars, banking products and groceries.
225
CMA, April 2017, ‘Online search: Consumer and firm behaviour - A review of the existing literature’, paragraph
1.6.
96
we do find that these results support the view that customers searching for
online M2M blinds are likely to consider relatively few retailers.
The review also noted that consumers often use multiple channels for a
single search before they make a purchase and that the ‘path’ to the final
purchase is often quite complicated.
226
Across different digital channels
such as search engine results and price comparison websites, the review
found that customers disproportionately focus their attention, clicks and
purchases on links at the top of the returned search results. In particular:
(a) On average, the first three links account for 40-65% of the total clicks on
desktop devices.
227
(b) On mobile devices, this tendency is even more accentuated, with the top
three links, on average, accounting for more than 70% of the total
clicks.
228
(c) The focus on links at the top of the results page is not simply driven by the
fact that top links are more likely to be relevant to customers’ searches,
but also by the fact that customers seem to display an inherent bias
towards clicking on links in higher positions.
229
The EC’s Google Shopping decision found that organic search results
generate significant traffic to a website when ranked within the first 3 to 5
organic search results, while customers pay little attention to the remaining
results. However, the EC decision does not make clear how clicks for
organic search compare to paid search, in particular Google Ads, which
tend to be positioned above the organic search results.
The EC’s decision also found that the rank of a given link in organic search
results on the first general search results page has a major impact on the
click rates of that link, irrespective of the relevance of the underlying page.
In line with this, the EC’s decision further found that moving the first ranked
result to the third ranked resulted in a reduction in clicks of about 50%,
while the effect was even stronger if the website was moved lower in the
search rankings.
Overall, both the CMA’s literature review and the EC’s Decision indicate
that customers do not tend to click beyond top results on the results page.
226
CMA, April 2017, ‘Online search: Consumer and firm behaviour - A review of the existing literature’, paragraph
4.20-4.29.
227
CMA, April 2017, ‘Online search: Consumer and firm behaviour - A review of the existing literature’, paragraph
1.6 c)
228
CMA, April 2017, ‘Online search: Consumer and firm behaviour - A review of the existing literature’, paragraph
1.6 c)
229
CMA, April 2017, ‘Online search: Consumer and firm behaviour - A review of the existing literature’,
paragraphs 4.50 to 4.57.
97
This implies that if a website does not feature in the top results, it is unlikely
to obtain significant traffic and therefore will be unable to effectively
compete.
The BDRC Survey submitted by the Parties provides an overview of the
customer search journey for online M2M blinds. In particular, the survey
shows the following:
(a) Large proportions of customers found the websites of Blinds2Go and 247
through generic search words such as ‘blinds’ or similar ([]% for
Blinds2Go / []% for 247); though a notable number of customers were
aware of the brand prior to purchase and searched by name or went
directly to the Parties’ websites ([]% of Blinds2go customers / []% of
247 customers).
(b) Over 70% of Blinds2Go and 247’s customers spend more than an hour on
research prior to purchase.
(c) 52% of Blinds2Go’s customers and 65% of 247’s customers looked on
other websites (naming Blinds Direct, Dunelm, John Lewis, Blinds4UK,
Amazon Marketplace, Next and eBay) prior to purchase. The proportions
of respondents who looked in-store prior to purchase were lower: 19% of
Blinds 2Go and 24% of 247 customers (naming Dunelm, John Lewis,
B&Q, Ikea, Next and local blinds/curtains shops).
(d) Customers tended to visit at most three other websites or fewer: Of the
customers who looked at other websites before purchasing (52% for
Blinds2Go and 65% for 247), most visited a maximum of three other
websites (71-77% visited three other websites or fewer; half of Blinds2Go
customers and 41% of 247 customers visited only one or two websites
before purchasing).
230
With respect to the time spent on research prior to purchase, the BDRC
Survey does not provide further details of what this research entails. The
response options suggest that the question was not necessarily exploring
continued search, but time between first search and purchase. Moreover,
time spent researching does not necessarily reflect time comparing the
offering (in terms of eg price, quality and customer service) across retailers;
it may also include significant within-website research (for instance, time
taken to choose a design), and therefore, in our view, does not provide
230
There may be some under-reporting as it is possible that some respondents may not have been able to recall
all the websites they visited.
98
reliable insights into the extent to which the offering of other retailers is
assessed by customers.
In response to this point, the Parties submitted that the BDRC Survey
provides concrete evidence that shows that consumers engage in
significant research before they make their purchasing decisions and that
this research entails visiting competitor websites. The Parties submitted
that this evidence is dismissed based on pure speculation on the CMA’s
behalf about what exactly customers may do.
231
We note that we do not
dismiss the survey’s finding that customers spend significant time
researching, and we also do not dismiss the survey’s finding that
customers visit certain other websites. However, and as set out above the
time spent researching does not give reliable insights into the time spent
comparing different retailers’ offerings. The Parties have not provided
additional evidence to support that the research primarily entails comparing
different retailers’ offerings.
With respect to search behaviour, and as noted above (at paragraph 8.19),
the Parties submitted that the BDRC Survey shows that customers
compare products / prices across many websites and further consult other
channels – noting that the BDRC Survey would tend to under-report the
amount of ‘shopping around’ by customers in the wider market as it was
conducted only amongst the existing customers of the Parties, including
repeat purchasers. The Parties also submitted Google Trends data for
2019 with respect to the search term ‘blinds’, which shows that users were
also searching for Dunelm, Ikea, and Argos in the same search session,
with 247 being outside the top 20 associated search terms.
However, we note that the BDRC Survey shows that the large majority of
customers visited three other websites or fewer prior to purchasing. While
we acknowledge that the inclusion of repeat customers may lead to the
survey under-reporting the amount of ‘shopping around’ by customers, we
also note that the finding on the number of websites visited is consistent
with the findings of the CMA research on customer search behaviour.
Additionally, we note that visiting several websites does not necessarily
mean that those websites all exert a competitive constraint on the Parties –
it may simply indicate that there is a long customer journey in which
customers visit various websites which they might dismiss altogether once
they have a better view of the offering. With respect to the Google Trends
data, we note that, as explained in paragraph 8.43, we have not received
231
Parties' response to Provisional Findings, 7 August 2020, paragraphs 4.40 and 4.41.
99
any evidence that this data is driven by customers considering both ready-
made and M2M blinds as part of their purchase journey.
The Parties submitted that the BDRC Survey does not support the finding
that, as per the CMA research and the EC Google Shopping decision,
being ranked highly on search engines (paid and organic) is likely to be
important in order to be successful in this market. The Parties note that
among the websites listed by the survey respondents as sites visited
(Dunelm, John Lewis, Amazon and Next) are those that do not rank highly
on Google according to our analysis. The Parties also submitted that the
CMA’s interpretation of the available evidence overstates the importance of
Google rankings, reiterating that customers spend a significant amount of
time researching blinds across multiple websites (and stores) before they
make their purchase.
232
With respect to websites being visited that are not ranked highly on Google,
we note that those listed by the Parties only relate to multi-channel retailers
or marketplaces. However, the survey also shows that the majority of
customers visited the websites of online M2M blind retailers (eg Blinds2Go
or Blinds Direct) and, as per our analysis, these retailers feature
prominently on Google. Regarding the multi-channel retailers, we further
acknowledge that, as pointed out in paragraph
8.27, PPC advertising
appears to be less important for these suppliers compared to online M2M
blind retailers. Additionally, we note that, as shown by our ranking analysis
of organic search results (see paragraph 8.103), Dunelm does frequently
rank highly on Google, while Next and Amazon feature to a small extent.
With respect to the customers spending a significant amount of time
researching, we note our discussion in paragraphs 8.40 to 8.43 above.
Additionally, we note that a potentially longer customer journey does not
contradict the importance of Google. We also note that the BDRC Survey
supports the importance of online search, showing that 44% of Blinds2Go’s
customers and 59% of 247’s customers found the respective website
through a generic search.
Our finding on the importance of ranking high on search engines is in line
with comments we received from third parties:
Swift Direct Blinds told us that the number of advertisements and
ranking position on Google has a massive impact on visibility and thus
customer conversion.
233
Swift Direct Blinds further told us that without
232
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.42 and 4.44.
233
[].
100
visibility on the first page of Google search results, ‘you are at a very
large disadvantage’.
234
Blinds4UK told us that the ‘only way to survive’ as an e-commerce site
is to be on the first page of Google’s organic search results.
235
Additionally, we note that the number of search results shown on mobile
devices without scrolling is very limited, given the typically small screen
sizes of such devices. As such, positioning is likely to be even more
important for traffic originating from mobile devices. With mobile devices
playing an increasing role in generating traffic, this would suggest ranking
is likely to become even more important going forward.
Conclusion
Overall, we consider that there is evidence which demonstrates that the
customer journey in selecting online M2M blinds may be long and may
involve customers looking at different products and/or consulting different
channels. However, notwithstanding this position, we do not have sufficient
evidence to conclude that alternatives considered as part of this journey
ultimately impose a material competitive constraint on online M2M blinds.
We further find that paid search is the most important source of traffic for
online M2M blind retailers (with paid search primarily relating to Google
Ads), followed by organic search and, of considerably less importance,
direct search. We also find that online M2M blind retailers spend the
majority of their marketing spend on Google PPC, underlining the
importance of paid search as a source of traffic.
The importance of these sources of traffic is also confirmed by the BDRC
Survey, which shows that customers primarily reach the Blinds2Go’s and
247’s websites through either searching for a generic term such as ‘blinds’
(44% of Blinds2Go customers and 59% of 247 customers) or searching for
the brand / going to the website directly (41% of Blinds2Go customers and
26% of 247 customers).
The BDRC Survey also shows that customers typically visit only a limited
number of websites prior to making a purchasing decision. This is
consistent with the findings from the CMA’s literature review on online
search and the EC’s Google Search (Shopping) decision, which indicate
234
[].
235
[].
101
that customers focus their attention and clicks on links at the top of the
returned search results.
Taken together, these factors suggest that ranking highly on search
engines is likely to be an important factor in order to be able to effectively
compete. Price, quality, range and service (PQRS) are also important
aspects of competition. Our assessment of the closeness of competition
between the Parties and the strength of the competitive constraint of other
suppliers therefore takes into account PQRS, the competition for online
traffic, as well as the interaction of the two.
Closeness of competition
Our guidelines note that the closeness of competition between the Parties
is one factor to be taken into consideration for the assessment of horizontal
unilateral effects. The guidelines state: ‘If the products of the merger firms
are close substitutes, unilateral effects are more likely because the merged
firm will recapture a significant share of the sales lost in response to the
price increase, making the price rise less costly.’
236
The Parties submitted that 247 is a minor and undifferentiated rival and not
a particularly close competitor to Blinds2Go.
237
The Parties further
submitted that 247 is ‘not of any unique or special importance’.
238
Additionally, they submitted that the competitive pressure 247 exerts on
Blinds2Go is, at most, limited and declining.
239
Before setting out our analysis, we note as a preliminary response to the
Parties’ submission that mergers may give rise to an SLC in situations
where merging parties are not each other’s closest competitors. In
particular, a merger may lead to an SLC in situations where the merger
removes an important competitive constraint, even if there is another
competitor that is as close a competitor to the Parties as the Parties are to
each other, or indeed closer.
In this section we assess how closely the Parties compete with one
another, as well as relative to the competitive constraint from other retailers
of M2M blinds online. In our assessment, we consider:
(a) The Parties’ service proposition;
236
Merger Assessment Guidelines (OFT1254), paragraph 5.4.9.
237
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.2. Main submission, 20 May 2020,
paragraph 6.3.
238
Main submission, 20 May 2020, paragraph 8.6.
239
Main submission, 20 May 2020, paragraph 1.7.
102
(b) The Parties’ online presence;
(c) Evidence from the Parties’ customers in the form of the BDRC Survey
submitted by the Parties;
(d) Internal documents;
(e) Pricing data included in Blinds2Go’s internal documents; and
(f) Views of the Parties’ competitors.
We further discuss the Parties’ submission that the competitive pressure
247 exerts on Blinds2Go is declining.
240
The Parties’ service proposition
We have looked at several different aspects of the Parties’ service
proposition with respect to online M2M blinds.
At a general level, the Parties’ product and service propositions are very
similar to one another. Both Parties offer the same types of M2M blinds (ie
roller blinds, wooden blinds, vertical blinds, roman blinds, venetian blinds,
day and night blinds and conservatory blinds) through the online sales
channel.
At a more granular level, we assessed the Parties’ offering in terms of
price, quality and product range. These are key characteristics that
customers consider when purchasing online M2M blinds. This is evidenced
by the BDRC Survey, which shows that more than 40% of respondents
indicated that each of ‘good prices/offers’, ‘product quality’ and ‘good/wide
product range’ were reasons for their choice of retailer. Each of these
factors were also listed amongst the top five in terms of the ‘one main
reason’ that respondents identified for their choice of retailer.
241
Price
With respect to price, Blinds2Go monitors a number of retailers in the
ordinary course of business, namely []. For each of these retailers it
240
Main submission, 20 May 2020, paragraph 1.7.
241
Similarly, two market reports submitted by the Parties indicate the following: the Verdict HRS Window
Dressing Report 2016 identifies ‘price’, ‘quality’ and ‘design’ as the three main factors that customers consider
important when shopping for M2M blinds, while the Global Data report indicates the same but with respect to
shopping for window dressings. We further note that these three factors are largely considered the main factors
for customer choice by the Parties and third parties. Parties’ main submission, 20 May 2020, paragraph 6.17.ii
and 6.20 and [].
103
collects their prices [] (alongside its own prices).
242
On the basis of this
data, we compared the prices of different retailers. The details of this
analysis are set out in Appendix D.
From this analysis, we see that the Parties’ prices are generally
competitive, in the sense of their prices being generally at the lower end of
the market, but that they do not consistently stand out as the cheapest
retailers. We also note that the Parties’ prices are neither consistently
closer to each other than to those of other retailers, nor are they
consistently further apart from each other than from those of other retailers.
We further note that, with the exception of [] (which tends to be among
the more expensive retailers), none of the other competitors included in the
price data collected by Blinds2Go stands out as consistently more
expensive.
243
However, some retailers are more expensive for certain
products. For example, Wilsons Blinds (part of Interior Goods Direct) is
significantly more expensive than the Parties for a number of products
(including 25mm slats white venetian, standard white roman and blackout
white roman blinds), Interior Goods Direct’s prices are also significantly
higher than those of 247 and Blinds2Go for faux wooden blinds (with the
exception of prices in 2020, when they align with the Parties), while Swift
Direct Blinds’ prices are significantly higher than those of either of the
Parties for venetian blinds.
Our observation that certain retailers are more expensive for certain
products than Blinds2Go and 247 is consistent with comments we received
from third parties. In particular, Swift Direct Blinds told us that it sometimes
cannot match the prices of Blinds2Go or 247 because of higher costs to
obtain the blinds, while MakeMyBlinds told us that it no longer matches the
prices of Blinds2Go and 247.
Quality
With respect to quality, the Parties submitted that product quality is ‘key’
and an important consideration for customers.
244
This is consistent with the
results from the BDRC survey: 44% of respondents quoted product quality
as one of the reasons that influenced their choice of retailer, with 16% of
respondents for Blinds2Go and 13% of respondents for 247 citing product
quality as the main reason.
242
The data covers the period from May 2016 until February 2020 (although for some competitors, only more
recent data is recorded). [].
243
The Parties submitted that Next has recently cut its prices on roman blinds and rollers blinds and that Next
has therefore become much more competitive on price. We discuss this submission in paragraph 8.199.
244
Main submission, 20 May 2020, paragraph 6.17.
104
We have received mixed evidence regarding the product quality of
Blinds2Go and 247. On the one hand, the Parties submitted that [] (and
other competitors) and noted that they largely attribute the success of
Blinds2Go to delivering quality products at the lowest possible price.
245
On
the other hand, Interior Goods Direct told us that Blinds2Go and 247 both
use lower quality fabrics than they do.
We note that customers may take into account a number of other aspects
of overall quality, including, for example, website quality, customer reviews
for the retailer and a retailer’s position on Google search results. This is
either because the customers value these aspects in and of themselves or
because they treat them as proxies for product quality. We assess these
different aspects of overall quality below.
(a) Four third parties indicated that website quality matters to customers,
specifically referring to ‘clean design’ and ‘look’ Similarly, the Parties
submitted that visual merchandising and imagery is key, which implies
that website quality is important.
246
Additionally, we note that a website’s
quality is likely to impact its ranking on Google search results (see
Appendix E). We note that while the Parties’ websites have similar setups,
looks and functionality (eg chat functionality or customer service number),
this is also the case for each of the other six online M2M retailers with
sales of more than £1 million in 2019. However, the websites of some of
the smaller online M2M retailers have a significantly different (and
arguably less appealing) look.
247
(b) Two online M2M retailers told us that customer satisfaction scores,
especially those on Trustpilot, are important. Based on evidence
submitted by the Parties, each of Blinds2Go and 247 has more than
10,000 reviews and excellent Trustpilot ratings, and there are only two
other online M2M retailers with more than 10,000 reviews, namely Swift
Direct Blinds and Interior Goods Direct – both of which also have an
excellent rating. There are also at least four further online M2M retailers
with excellent Trustpilot ratings and more than 1,000 reviews.
248
(c) Customers may also see the ranking of a website on Google search
results as an indicator of the quality of the website. When asked whether
appearing high up Google rankings was a signal of quality to customers,
247 said that appearing at the top of organic listings ‘create[s] some
credibility’, whereas with respect to paid rankings ‘there is an increasing
245
Main submission, 20 May 2020, paragraphs 5.4 and 5.5.
246
Main submission, 20 May 2020, paragraph 6.17.
247
For example, https://www.conceptblindsdirect.co.uk/.
248
This statistic excludes Bloc Blinds, a niche player for specialised roller blinds.
105
awareness now of the way paid search marketing works and they [ie the
customers] know what they are actually looking at is an advert’. We note
that the Parties, alongside Interior Goods Direct and Dunelm, feature
frequently in the top positions on Google’s organic search results (see
table 7).
Range
The BDRC Survey indicates that customers consider range to be important,
with 42% of respondents for Blinds2Go and 43% of respondents for 247
indicating a ‘good/wide product range’ as one of the reasons for choosing a
retailer. Additionally, the survey shows that ‘had what specifically wanted’
was selected as the one main reason for choosing a certain retailer by the
second highest proportion of respondents (namely by 46% of respondents
for Blinds2Go and 50% of respondents for 247).
In line with this, the Parties submitted that product choice is key.
249
This is
also consistent with comments received from Swift Direct Blinds who told
us that a wide range of products relative to its competitors contributed to its
growth between 2014 and 2016.
However, the Parties also submitted the importance of range as a
competitive constraint should not be overstated and that retailers do not
need to sell a significant number of SKUs to be an effective competitor to
the Parties given the concentration of sales among a relatively small
number of popular blinds.
250
The Parties submitted that the top nine
variants of blinds account for []% of sales, while the top 243 SKUs for
Blinds2Go account for []% of sales by value.
251
252
In response to these submissions, we first note that the presented analyses
do not necessarily mean that sales are particularly concentrated, although
we acknowledge that a relatively limited number of SKUs accounts for the
majority of sales. With respect to the analysis of the top nine variants, we
note that the Parties aggregated different SKUs based only on the blind
type and a typically high-level colour (eg white, grey, blue).
253
This
assessment does not take into account other differences in colour, different
fabrics or other options available for M2M blinds.
254
Therefore, these top
249
Main submission, 20 May 2020, paragraph 6.17.
250
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.62.
251
The number of 243 SKUs was chosen because it equates to the total number of ready-made SKUs sold by
Dunelm.
252
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.86.
253
For example, for a certain type of blind, any blind that in its description contained the word ‘white’ was
aggregated as the ‘white’ variant of such blind type.
254
For example, the presence of patterns, the colour of some parts, the dimension of the slats for venetian blinds,
and the option such as ‘Enjoy’ or ‘DuoShare’ for rollers.
106
nine variants may actually include a large number of SKUs. With respect to
the analysis of the top 243 SKUs, we note that it remains unclear how the
remaining more than 40% of sales are distributed. Second, we consider
that identifying that a relatively limited number of SKUs accounts for the
majority of sales does not mean that range is not important, and the Parties
have not demonstrated how competitors can compete effectively if only
offering a comparatively small number of SKUs. The Parties’ submission
also does not consider the extent to which a wide range is in and of itself
attractive to customers (ie increases sales even if customers primarily buy
a limited number of SKUs).
255
On balance, we find that the evidence received suggests that range
matters, and have therefore compared the overall number of SKUs as well
as the number of SKUs relating to wooden venetian blinds (the top-selling
type of blind for each of the top four online M2M blind retailers) offered by a
number of different online M2M blind retailers, including each of Blinds2Go
and 247. Table 6 shows this comparison.
Table 6: Number of SKUs for different online retailers of M2M blinds
Retailers
# of SKUs
(all blinds)
# of SKUs
(wooden venetian blinds)
Blinds2Go
4,497
225
Interior Goods Direct
4,014
40
247
2,823
245
Swift Direct
2,454
222
Wilsons Blinds*
1,778
96
MakeMyBlinds
700
64
Blinds4uk
2,228
97
So Easy Blinds
3,730
90
Blinds UK
2,570
117
Concept Blinds
1,637
73
English Blinds
1,305
182
Terrys Fabrics
1,300
48
Unbeatable Blinds
668
58
Source: Data based on Parties’ desk research.
Note: [] submitted data on number of SKUs that was broadly in line with these estimates; however []submitted that it sells
a much larger number of SKUs (~[]) in total (source: Blinds2Go, response to CMA Questionnaire, 20 April 2020, paragraph
2.5, and Interior Goods Direct, Swift Direct Blinds and MakeMyBlinds, responses to CMA Questionnaires, 28 April 2020).
Note: *Part of Interior Goods Direct since April 2020.
Each of Blinds2Go and 247 appears to offer a very wide range of products
and several competitors offer a much narrower range than the Parties,
albeit there are also at least six competitors that offer a broadly comparable
overall range. Each of Blinds2Go and 247 also appears to offer a
255
We note that a similar argument was raised by the Parties, namely that []. Main submission, 20 May 2020,
paragraph 6.17(i).
107
particularly wide range of wooden venetian blinds, which is the top-selling
type of blind for each of the top four online M2M blind retailers. With
respect to the ten competitors for whom we hold information on the number
of SKUs relating to wooden venetian blinds, only two competitors have a
broadly similar range to the Parties for this type of blind.
256
While we acknowledge the Parties’ submission that the comparison is
limited to a simple count of SKUs and that it does not take into account
that, unlike 247, Blinds2Go sells a range of designer blinds,
257
we still find
that the SKU analysis shows that both Parties offer a wide range of
products and that their offering is hence similar in terms of range.
Conclusion on price, quality and range
Overall, we find that the Parties’ offerings in terms of price, quality and
range are similar. While, as illustrated by paragraphs 8.168 to 8.176 below,
Interior Goods Direct and Swift Direct Blinds have a broadly similar offering
to the Parties, we find that the offerings of other smaller online M2M blind
retailers (including MakeMyBlinds) is either differentiated, in at least some
respects, to that of the Parties or we have insufficient evidence that their
offerings are similar to the Parties’.
258
Online presence
As explained above (see paragraphs 8.15 to 8.16), competition between
retailers of online M2M blinds is likely to be heavily influenced by how
effective a retailer is at attracting customers to its website. This in turn
depends on how well a retailer ranks and performs on search engines
(especially Google).
We therefore undertook a range of analyses of Blinds2Go and 247’s online
presence, to understand how effective each of them is when competing in
the market as compared to their competitors:
(a) Our main type of analysis assesses how well Blinds2Go, 247 and their
competitors rank on Google paid search and Google organic search.
(b) Additionally, we compare Blinds2Go and 247’s Google Ad campaigns
performance with those of their competitors.
256
These two retailers are Swift Direct Blinds and English Blinds.
257
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.62.
258
For some of the smaller online M2M blind retailers, we do not hold sufficient evidence because these retailers
did not respond to our requests and the Parties’ submissions only included limited information on these retailers.
108
(c) As a supplementary analysis, we also assess the ad search term bidding
behaviour of Blinds2Go, 247 and their competitors.
The results provide evidence on the closeness of competition between
Blinds2Go and 247 and the strength of the competitive constraint exerted
by other M2M blind suppliers.
In relation to the relevance of these analyses, the Parties submitted that:
(a) Google is neither the only source of traffic nor the only or most important
determinant of competitive outcomes;
(b) attracting traffic does not mean that the website will be successful at
converting this traffic into sales; and
(c) it is wrong to consider competition for traffic and other parameters of
competition (on product offering or price, quality and range) separately
without a proper consideration of how they interact.
259
In response, we note that:
(a) Google remains a very important route to market, not just when it comes
to attracting traffic but also in terms of generating conversions – albeit not
being the only one, as explained in paragraphs 8.49 to 8.53. We note that
Blinds2Go and 247 generate most of their revenue from customers
sourced through paid and organic Google searches – from []% (for
Blinds2Go) to []% (for 247).
260
For the other main online M2M blind
retailers, the proportion of revenue generated from these sources ranges
between []% and []%.
(b) We recognise that attracting traffic to a website is only one (albeit an
important) way in which retailers compete, and elsewhere in our
competitive assessment, we have analysed other aspects of competition
in the market for the retail supply of M2M blinds online. Our analysis of
Blinds2Go’s, 247’s and other retailers’ online presence therefore only
constitutes part of our evidence basis.
(c) We acknowledge that it is important to consider the interaction between
competition for traffic and other parameters of competition. However, we
note that we are not viewing the Google analyses presented below in
isolation.
259
Parties' response to Provisional Findings, 7 August 2020, paragraphs 4.49.
260
We note that these values include all search engines, not just Google.
109
The Parties also submitted that our analyses failed to take into account the
fact that non-Google spending (or non-PPC spend) is likely to affect
retailers’ rankings and performance on Google and that, therefore, any
comparisons between Blinds2Go, 247 and other retailers must take this
into account. As an example, the Parties noted [].
We acknowledge that non-Google spending may be impacting the results
of our analyses to some extent, including the analysis of ranking (see
paragraph 8.78 above and paragraphs 8.87 to 8.107 below), the analysis of
performance on Google Ads (see paragraph 8.66(b) above and paragraphs
8.109 to 8.118 below) and the analysis of ROI/CPA (see Appendix F).
However, in our view, it is unlikely that non-Google spending is the only or
main driver for the results we observe. While we acknowledge that such
spending may influence results to some extent, we also note that with
respect to the analysis of ranking (which constitutes our key analysis with
respect to closeness of competition), the ranking position a retailer
achieves is in and of itself highly relevant (ie it is relevant independent of
how exactly the retailer obtained such ranking).
Below, we describe the results of our analyses of online presence and our
respective conclusions.
Ranking in Google search results
As explained above (paragraphs 8.32 to 8.49 ) the search behaviour of
customers indicates that if a website is not well represented on the first
page (and to some extent even the first two or three results) it is unlikely to
attract significant traffic and therefore be able to effectively compete in the
market.
We therefore conducted an analysis of Google rankings. This analysis, set
out below, covers analysis of rankings in search results, separately with
respect to:
(a) Google Ads; and
(b) organic links.
Google Ads
We analysed four different sets of data on rankings in Google Ads:
(a) The first consisted of rankings we were able to observe directly from
Google web results with respect to the top 10 search words in the industry
110
related to blinds (not M2M specific) based on Google Trends in 2019 as
explained in paragraph 8.88 below;
(b) The second consisted of ranking data provided by Blinds2Go and 247 and
two other retailers with respect to the top 10 search words used in their
Google Ads campaigns for the period of 2017-2019.
(c) The third consisted of ranking related data (but including other metrics
than the data in (b)) provided by Blinds2Go and 247 and two other
retailers with respect to the search words used in their Google Ads
campaigns for the period of 2019.
(d) The fourth consisted of ranking data found in Blinds2Go’s internal
documents, for a number of standalone periods covering 2016, 2017 and
2018.
The first piece of analysis consisted of recording the position of the Parties
and competitors in the first page of Google Web results. In particular, this
analysis systematically recorded results for:
261
(a) 10 search words - ‘blinds’, ‘blind’, ‘window blinds’, ‘blinds uk’, ‘roller
blinds’, ‘venetian blinds’, ‘roman blinds’, ‘vertical blinds’, ‘wooden blinds’,
‘blackout blinds’;
262
and
(b) Four different times per day for a period of one week - 18 May 2020 to 24
May 2020 at 10h, 13h, 16h, and 19h (except Saturday and Sunday, when
only 16h and 19h were considered).
263
Details of this analysis are presented in Appendix E. The results for Ads
shown at the top of the page are summarised below.
261
This analysis therefore considerably expanded on the analysis of rankings done at P1 which covered only one
search word and two dates.
262
We obtained this list of search words by analysing the most popular related search queries (excluding
branded search queries) for the word ‘blinds’ (based on Google Trends data for 2019, UK). We also checked that
these search queries broadly coincide with the non-branded Google Ads search words the Parties and their
largest competitors (ie Interior Goods Direct and Swift Direct Blinds) get most impressions from and spend the
largest amount of money on in 2019.
263
To establish timings for the analysis, we looked at times of day there seemed to be more traffic on Parties’
sites and simultaneously including other times in case those times corresponded to heavier traffic in the case of
other retailers. The timings recorded also were chosen to accommodating team availability.
111
Table 7: Google Ads on top of page - Proportion of time and search word combinations that a
given retailer ranked in a certain position in first page of results
Retailer
Position
1
2
3
4
Blinds2go
38%
39%
7%
247
51%
22%
11%
Wilsons*
5%
5%
4%
5%
Terrys Fabrics
3%
Blinds Direct*
1%
12%
30%
6%
Make My Blinds
1%
3%
7%
2%
Blindsbypost
1%
1%
1%
Swift Direct Blinds
3%
22%
BlindsUK
6%
No Ad**
19%
37%
54%
Source: CMA analysis of publicly available data.
Notes:
Other retailers that never feature in the top position and have a share of lower than 5% in all other positions are not shown
*Part of Interior Goods Direct
**While we have excluded time and search word combinations where no ads were shown at all, combinations where just
certain ad position were missing were kept in the analysis.
Table 7 above sets out, for each retailer, the proportion of times its ads
appeared in first to fourth position in Google Ads, across the 10 search
words we searched for. From these results, we note the following:
(a) Blinds2Go and 247 rank very frequently among the top two positions of
Google Ads (with respect to the first ad position, Blinds2Go shows up
38% of the times, whereas 247 appears 51% of the times; regarding the
second ad position, Blinds2Go appears 39% of times and 247 22% of
times). Interior Goods Direct (in this case, comprising Blinds Direct and
Wilsons Blinds) is the only competitor that has a notable share of the two
top positions (appearing in the first ad position 6% of times and second
position 17% of times). Interior Goods Direct outperforms all other
retailers regarding the third ad position, whereas Swift Direct Binds
outperforms all other retailers regarding the fourth ad position.
(b) Blinds2Go and 247’s brands tend to appear in several of the top ads
results.
264
When more than one ad is shown on the first page, Blinds2Go
and 247 feature in at least 2 of them in most cases (corresponding to a
minimum of 50% share of top ads for most cases).
265
For completeness, we have also assessed the ranking of retailers with
respect to Google Ads at the bottom of the page. While the details of this
264
Including the several Hunter Douglas’ brands (Blinds2Go, Web Blinds, Hillarys, Thomas Sanderson and
Tuiss) and 247.
265
This is because there can be only a maximum of four ads show on the top ads of Google search results
pages. This might however include situations where two brands of Hunter Douglas are present rather than both
Parties per se.
112
analysis are set out in Appendix E, we note that there appear to be no
further competitors that feature frequently in Google Ads at the bottom of
the page. Additionally, featuring in Google Ads at the bottom of the page is
less relevant than featuring in Google Ads at the top of the page for the
reasons discussed in paragraphs 8.32 to 8.42 above. We have taken this
into account when weighing this evidence.
The second piece of analysis involved looking at ranking metrics for the top
10 search words by number of impressions for each of Blinds2Go, 247,
[] in the period 2017 to 2019 with respect to their Google Ads
campaigns.
266
This analysis therefore covers the [] in the market in terms
of market share, although, as shown in Table 3 we note that Swift Direct
Blinds’ market share is significantly smaller than the market shares of the
other retailers considered. In conducting this analysis, we looked at the
following metrics, which are held by each retailer both in relation to itself
and other retailers:
267
(a) Impression share – Percentage of impressions that a retailer’s ad
receives compared to the total number of impressions that the retailer’s
ad is eligible for.
268
An impression share provides information on how
often an ad is shown rather than on the positions in which it is shown.
(b) Overlap rate – How often another retailer’s ad received an impression
while a given retailer also received an impression
(c) Top of page rate – How often a retailer’s ad shows at the top of the page
(within the group of top ads – rather than the first ad that shows) above
the organic results for a particular search word.
(d) Absolute top of page rate – How often a retailer’s ad shows at the
absolute top of the search page (first ad that shows) above the organic
results for a particular search word.
The main results of this analysis are summarised below (further details are
presented in Appendix E):
269
266
This includes Blinds-related search words only. Number of impressions means how often the ad is shown. An
impression is counted each time the ad is shown on a search result page or other site on Google. For the
avoidance of doubt, this analysis focused on the search words of each retailer as opposed to just looking only at
the search words of Blinds2Go and 247. The top 10 search words assessed represent a large proportion of
overall impressions for each retailer in this period ([]% for 247, []% for Blinds2Go, []% for Web Blinds,
[]% for Blinds Direct and []% for Swift Direct Blinds).
267
Additional metrics that were provided are ‘position above rate’, ‘absolute top of page rate’ and ‘outranking
share’, but we did not think they were as insightful and have therefore not analysed them in further detail.
268
‘Eligible for’ in this case means that the given retailer participated in the auction for this search word.
269
The analysis is for the period 2017 to 2019 combined (we do not hold the data separately for each year),
meaning that any recent growth of competitors may be underrepresented. We understand that both Interior
Goods Direct and Make My Blinds have grown significantly in the last three years.
113
(a) Blinds2Go consistently (ie across all retailers’ data) has the highest
average impressions share ([]%), average top of page rate ([]%) and
average absolute top of page rate ([]%).
(b) 247 is typically the runner up to Blinds2Go, although it is sometimes
outperformed by Blinds Direct (part of Interior Goods Direct) in terms of
impression share and top of page rate. Blinds Direct (part of Interior
Goods Direct) always outperforms 247 when it comes to absolute top of
page rate.
(c) Other than Blinds2Go, 247 and [], there are no other retailers that
perform highly on any of the metrics analysed.
The third piece of analysis involved analysing metrics related to ranking for
the search words of each of Blinds2Go, Web Blinds, 247, [] and [] in
the year of 2019.
270
In particular, we assessed:
(a) Impression share.
(b) Quality score – this provides insights on the quality of a certain retailer’s
ad. The score has values of 1-10 (1 lowest, 10 highest) and is reported for
each search word, providing an estimate of the quality of a retailer’s ad
and the landing pages triggered by it. Three factors determine quality
score: expected click-through rate, ad relevance, landing page
experience.
271
We focused our analysis on each retailer’s top 3, 5, 10, 30 and 50 search
words (by number of impressions in the year) as well as all search words in
2019.
Impression shares provide information on the extent to which a retailer’s
ads are shown and therefore allow inferences to be drawn about the extent
to which a retailer is able to reach visibility, but they do not provide
information on how well the retailer ranked. Quality scores are also
informative in the sense that, despite not being directly used in the ad
ranking algorithm,
272
they are based on much of the same data that is
270
This includes blinds-related search words only. For the avoidance of doubt by this we mean that the analysis
focused on the search words of each retailer as opposed to just looking only at the search words of Blinds2Go
and 247. There are several reasons why the results of this analysis may differ from the results of our previous
analysis discussed above in relation to the metrics they both cover (in this case, impression shares). These
reasons include that: (i) the top 10 search words for each retailer do not necessarily need to be the same when
considering overall period of 2017 to 2019 versus the top 10 words in 2019 only; (ii) the previous analysis shows
simple averages of impression shares, whereas this analysis shows weighted averages (weighted on the basis of
impressions); and (iii) there might in general be some differences in impression shares even for same search
words over the years.
271
https://support.google.com/google-ads/answer/7050591
https://support.google.com/google-ads/answer/2454010
272
https://support.google.com/google-ads/answer/140351?hl=en-GB
114
considered regarding quality in the Ad ranking algorithm, as explained in
Appendix E.
With respect to the top 10 search words of each retailer, [] and
Blinds2Go are the best performing retailers when it comes to both
impression shares ([]) and quality scores ([]). 247 performs
comparatively worse on both bases ([]% impression share and []
quality score). [] performs the worst, with a []% impression share.
We found similar results when we ran the analysis for all search words of
each retailer, rather than just the top ten. Further details are set out in
Appendix E.
The fourth piece of analysis involved a review of the ‘Auction Insights’
reports and ‘Account Health’ reports contained in Blinds2Go’s internal
documents. These reports show a number of different metrics related to
ranking.
We have only identified three complete ‘Auction Insights’ reports and two
‘Account Health’ reports, each covering a week to a few months within the
period from 2016 to 2018. As such, this evidence is likely to show an
incomplete and outdated picture, and we have therefore placed limited
weight on it. []. More details on the analysis of these internal documents
can be found in Appendix E.
In response to the results of our analyses on ranking, the Parties submitted
that there are various online retailers with non-negligible sales (eg
MakeMyBlinds) that do not feature frequently among the top Google search
results according to the CMA’s own analysis.
273
However, we note that the
frequency at which MakeMyBlinds appears in the top Google Ads is in line
with its relatively small sales and market share of []%: MakeMyBlinds,
according to our ranking analysis, appears 1% of times in the top, 3% in the
second, 7% in the third and 2% in the fourth ad position (see Table 7
above).
Organic search
With respect to rankings in organic search results, we considered the
rankings we were able to observe directly from Google Web results with
respect to top 10 search words in the industry related to blinds (ie not
made-to measure specific search words) based on Google Trends in 2019,
as explained in paragraph 8.88 above.
273
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.35.
115
The details of this analysis are set out in Appendix E. Below, we
summarise the results of this analysis:
(a) Each of Blinds2Go, 247, Dunelm and Interior Goods Direct features
frequently in the top three positions. The only other competitors that
rank frequently, but only in lower positions, are Argos, Swift Direct
Blinds, The Range and Next. However, we note that some of these
retailers (Dunelm, Argos, The Range, and Next) have a strong
ready-made blind offering (and some of these, eg Argos and The
Range, do not sell M2M blinds).
(b) In the vast majority of cases (94%), the Parties’ brands show in at
least 2 of the top 5 organic links and show in 3 of the 5 top organic
links in 30% of the cases.
274
We also found an internal document commissioned by 247 []. [] (see
paragraph 8.177). More details on this can be found in Appendix E.
Conclusion on Google rankings
Overall, this analysis shows that both Parties’ brands consistently rank
highly in paid search, and across a number of different metrics, indicating
that the Parties’ brands are both highly effective at competing for the top
positions. They also rank highly in organic search, though to a lesser
extent. Additionally, the analysis shows that the Parties’ brands tend to
consistently rank the highest relative to other retailers, although with
Interior Goods Direct outperforming 247 in some cases.
The results of our analysis on online presence, namely that the Parties
consistently appear in the highest rankings relative to other retailers, with
only Interior Goods Direct outperforming 247 in some cases, closely align
with the market shares we found for different retailers in the online M2M
blinds market (see Table 3). They further indicate that (i) the Parties,
together with Interior Goods Direct, are the leading suppliers in this market
and close competitors to each other and (ii) other online M2M blinds
suppliers (including Swift Direct Blinds) may be a less effective constraint.
Comparison of performance on Google Ads
Given the importance of Google Ads in obtaining traffic in this market, the
extent to which retailers are able to perform well in Google Ad campaigns
274
The Parties’ brands include the several Hunter Douglas’ brands (Blinds2Go, Web Blinds, Hillarys, Thomas
Sanderson and Tuiss) and 247. This might however include situations where two brands of Hunter Douglas are
present rather than both Parties per se.
116
can provide an indication of how effective the Parties and their competitors
are at obtaining traffic, which in turn provides an indication of closeness of
competition. For example, if two retailers within a market are able to
perform well on this metric whereas others are not, this might suggest that
the former two will be closer competitors with each other and that the
others pose a relatively less significant constraint in the market and are
likely to be more distant competitors.
In the previous section, we have assessed the extent to which retailers
perform differently with respect to relative positioning in Google Ad ranks,
including considering the impact of quality on rankings.
However, it is also relevant to look at other aspects of performance that, to
some extent, will reflect the impact of such rankings (through clicks and
click-through rates, which are likely to depend on ad position, and
consequently impact conversions).
We have therefore also analysed the performance of different retailers'
Google Ad campaigns by looking at a number of performance metrics for
these campaigns. In particular, we assessed data on Google Ad search
words from each of Blinds2Go, Web Blinds, 247, Blinds Direct (part of
Interior Goods Direct) and Swift Direct Blinds. We focused our analysis on
the following performance metrics:
275
(a) Costs;
(b) Impressions, ie views – how often the ad is shown. An impression is
counted each time the ad is shown on a search result page or other site
on Google.
276
(c) Interactions, ie clicks – the main user action associated with an ad format,
in this case, clicks.
277
(d) Conversions, ie sales – an action that has been defined as valuable to the
business, such as an online purchase.
278
(e) Costs per impression, per interactions and per conversions.
275
Blinds-related search words only. For the avoidance of doubt by this we mean that the analysis focused on the
search words of each retailer as opposed to just looking only at the search words of Blinds2Go and 247.
276
See https://support.google.com/google-ads/answer/6320?hl=en-GB
277
See https://support.google.com/google-ads/answer/6281923?hl=en-GB
278
See https://support.google.com/google-ads/answer/6365?hl=en-GB
117
We focused our analysis on the performance of each retailer’s top 3, 5, 10,
30 and 50 search words (by number of impressions in the year) as well as
all search words in 2019.
Details of our analysis are set out in Appendix E. Below we summarise the
main results with respect to the top 10 search words of each retailer in
2019:
(a) Blinds2Go spends the most on Google Ads ([]), whereas 247 spends
slightly over £[]. [] spends a similar amount to Blinds2Go and
significantly more than 247. [] spends much less than all other retailers
(less than £[]).
(b) Blinds2Go’s and 247’s ads outperform those of both [] and [] on
number of views by a large extent ([]). [] has the most interactions
and conversions by far, followed distantly by []. [] has a very small
number of interactions and conversions compared to the other retailers
considered.
(c) Although [] has a much smaller number of interactions than [], it is
the retailer that performs best when it comes to generating traffic from
views (as measured by interactions as a proportion of impressions).
However, and more importantly, [] is the best at converting when
considering conversions as a proportion of interactions, distantly followed
by [] and [].
(d) The other retailers considered are paying similar or higher amounts per
interaction to those of the Parties. However, they pay much more per
conversion than the Parties, especially when compared to [],
suggesting that other retailers are not as successful at converting (and
that their marketing spend is less effective) compared to the Parties [].
The Parties submitted that the fact that 247 – according to some metrics –
is a less effective PPC spender than some of its competitors is ‘somehow
brushed aside’.
279
As set out in paragraph 8.112 above, we acknowledge
that on the basis of certain metrics, 247 is indeed outperformed. However,
we find that on balance (ie considering the evidence on all metrics in the
round), 247 tends to perform better than the two other competitors
considered. Additionally, we note that one competitor performing better on
certain metrics does not mean that the Parties are not close competitors.
279
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.55.
118
The Parties further submitted that we should focus our analysis on all
search words instead of the top 10 search words only.
280
We note that, as
set out in paragraph 8.111 above, we have conducted the analysis not only
on the basis of the top 10 search words, but also on the basis of all search
words (as well as the top 3, 5, 30 and 50 search words). When considering
all search words, we find broadly similar results to when the analysis is
done for the top 10 search words (see Appendix E for details).
We note that retailers’ values for conversions were not all provided on the
same basis. In particular, conversions included revenue from providing
samples to customers for all retailers except [] and 247 (for which
samples were excluded from conversions). This suggests that the number
of conversions for the other retailers is likely to be overestimated compared
to [] and 247 – leading, in turn, to an underestimation of costs per
conversion for Blinds2Go, Web Blinds and [] and 247. However, when
we adjust conversions to exclude sample sales,
281
the costs per conversion
are still considerably smaller for the Parties than for other competitors.
Similarly, the findings on the absolute and relative numbers of conversions
(see paragraphs 8.112(b) and 8.112(c)) also continue to hold when
removing samples from conversions.
Overall, this analysis indicates that Blinds2Go tends to perform better on
Google Ads than the other online M2M blind retailers (247, []). With the
exception of cost per impression, Blinds2Go performs best on all metrics
analysed (number of views, interactions, conversions and costs per
conversion and interaction) and has the highest overall spend. Blinds2Go is
further the best at generating conversions from clicks, although [] is the
best at generating traffic out of views. 247 performs less well than [], but
better than [], in terms of number of interactions and conversions;
however it outperforms [] and [] when it comes to number of views and
costs per conversion. In our view, these findings indicate that the Parties
are both effective competitors for online M2M blinds, giving some indication
that they are likely to be close competitors.
Ad search words bidding behaviour
We assessed the Parties’ and other retailers’ bidding behaviour, with
respect to:
280
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.54.
281
Conversions can be adjusted to exclude the provision of samples to customers as we have information on the
percentage of conversions that represent samples for these retailers. Consistent with their responses to the
CMA’s Questionnaires on number of sales and samples in 2019, we have assumed that []% of conversions
corresponded to samples for both Blinds2Go and Web Blinds and []% for [].
119
(a) the extent to which the search words
282
they bid on overlap; and
(b) the extent to which branded search words are bid on and the non-brand
bidding agreements (‘NBBAs’) are present in the market.
Details of these analyses are set out in Appendix E.
With respect to search word overlap, the analysis does not give a clear
indication that there is one retailer that has a particularly high overlap with
either of Blinds2Go and 247. Blinds2Go and 247 overlap with [] to a
broadly similar extent as they overlap with each-other. Therefore, this
analysis does not allow us to draw any clear conclusions with respect to
relative closeness of competition between Blinds2Go and 247 and between
either of them and [].
Similarly, the analysis of brand bidding and NBBAs, does not allow us to
draw any clear conclusions on the effectiveness and strength of
competition between the Parties and between the Parties and other
retailers. Although we find that [],
283
this does not provide information on
the strength of such retailers as a competitive force in the market or the
extent to which they are able to successfully compete against the either
Blinds2Go or 247.
Overall, we therefore agree with the Parties’ submission that the ‘ad search
words bidding behaviour’ analysis does not give any clear indication that
the Parties are competing more closely between each other than with other
retailers.
284
However, this does not imply that the analysis provides
evidence that the Parties are not close competitors.
Conclusions on online presence
On the basis of our analyses of the Parties’ online presence, we find the
following:
(a) Our analysis of Google rankings, which includes all retailers present
online, shows that both Parties consistently rank highly, indicating
that the Parties are both highly effective at competing for the top
positions, with only Interior Goods Direct outperforming 247 in some
cases.
282
We have considered blinds-related search words only.
283
According to the Parties, all these agreements are informal rather than formal written agreements.
284
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.54.
120
(b) Our analysis of Google Ads performance indicates that the Parties
are both effective competitors for online M2M blinds.
Taken together, these analyses indicate that the Parties are likely to be
close competitors.
285
Survey evidence
As previously noted, the Parties submitted a survey, the BDRC Survey,
which they commissioned to understand the purchase journey and choices
of the Parties’ customers.
286
The Parties submitted that the BDRC Survey
indicates that only a small proportion of Blinds2Go’s customers, 12%,
would switch to 247 if Blinds2Go stopped selling blinds – and that diversion
to Dunelm (at 17%), Blinds Direct (part of Interior Goods Direct) (14%) and
John Lewis (12%) are higher. The Parties submitted that this is further
evidence supporting their view that 247 is not an important competitor to
Blinds2Go.
287
The Parties further submitted that with respect to diversion from 247 to
Blinds2Go, one would expect relatively high diversion from small
competitors to market leaders, simply reflecting relative market shares. The
Parties argued that if 247 were to have a 7% share of online M2M blinds
and Blinds2Go 55% and out-of-market constraints are very limited, then
based on relative market shares one would expect diversion of about 59%
(55%/(1-7%)). The Parties noted that substantially fewer of 247’s
customers, namely 33%, stated that they would switch to Hunter Douglas
brands (predominantly Blinds2Go) if 247 stopped selling blinds.
288
With respect to diversion from Blinds2Go to 247, the BDRC Survey implies
a diversion ratio of 13%. We note that, while this diversion ratio is not
particularly high, it is still the second highest diversion to an online M2M
blind retailer, closely behind Blinds Direct (part of Interior Goods Direct).
Additionally, we note that the third highest diversion to an online M2M blind
retailer, namely Blinds4UK, only amounts to 4%. In our view, this confirms
that there are only a small number of credible online M2M blind retailers,
namely Blinds2Go, 247 and Interior Goods Direct.
While we acknowledge that the survey results show that diversion to
Dunelm and John Lewis (combining diversion to both their online and their
285
As noted above, our analysis of search words overlap and brand bidding/NBBAs does not allow us to reach
clear conclusions on the closeness of competition between the Parties and effectiveness of other retailers as
competitors.
286
Main submission, 20 May 2020, paragraph 1.8.
287
Main submission, 20 May 2020, 8.18.
288
Main submission, 20 May 2020, paragraph 8.19.
121
offline offering) is higher or similar to the diversion to 247, we consider that
reported diversion to large multi-channel retailers (and, in particular, the
diversion to the online offering of Dunelm and John Lewis) is likely subject
to an upward bias (and hence that diversion to Blinds2Go and 247 and
other online M2M blind retailers is likely subject to a downward bias) for the
following reasons:
(a) Customers are likely to be more familiar with the brands of large multi-
channel retailers, given these are household names and customers may
have shopped at them before for other products. In contrast, the brands of
online M2M blind retailers tend to be less well known, and, indeed, the
Parties submitted that there is relatively little brand recognition with
respect to online M2M blind retailers.
289
Accordingly, when asked to
indicate which other retailer they would have purchased from,
respondents, not being able to do further research in the course of
responding to the survey, are more likely to have selected a familiar brand
name than an unfamiliar brand name (or a brand name they encountered
during their research process but subsequently forgot about).
(b) However, respondents may have chosen Dunelm or John Lewis on the
basis of brand familiarity, unaware of the true nature of their offerings. For
example, Dunelm’s website requires a virtual appointment and does not
have an immediate option to order online M2M blinds while John Lewis
only has a very limited offering online compared to its in-store range. In
this regard, we note that the BDRC Survey shows that only 38% of
Blinds2Go’s customers and 43% of 247’s customers that said they would
divert to Dunelm indicated that they had visited Dunelm’s website.
Similarly, only 43% of Blinds2Go’s customers and 48% of 247’s
customers that said they would divert to John Lewis indicated that they
had visited John Lewis’ website.
The Parties responded to this argument by submitting that the CMA’s
reported statistics only include customers who visited the website, but not
researched in-store (or indeed via other channels).
290
Additionally, the
Parties submitted that because multi-channel retailers are trusted high
quality retailers, can offer consumers broader design inspiration and a
much wider range of products than M2M blinds, ‘the CMA cannot
reasonably conclude that their offerings would not in practice be chosen’.
291
With respect to customers researching in-store, we note that such research
is unlikely to reveal the true nature of the online offering of these retailers.
289
Main submission, 20 May 2020, paragraph 7.73 and Figure 7.10.
290
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.78.
291
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.79.
122
For example, a customer visiting a John Lewis store may wrongly assume
that the same offering is available online (and hence indicate diverting to
the online offering of John Lewis), even though, and as noted above, John
Lewis only has a very limited offering online compared to its in-store range.
As such, we find that it is appropriate to focus on those customers that
visited the respective website.
With respect to our wider conclusion, we do not suggest that multi-channel
retailers would not be chosen at all. Our finding is that on the basis of the
above, we consider that reported diversion to large multi-channel retailers
(and, in particular, the diversion to the online offering of Dunelm and John
Lewis) is likely subject to an upward bias. We also note that at least some
of the factors the Parties listed on why multi-channel retailers would still be
chosen (eg a much wider range of products than M2M blinds) do not
necessarily apply to the online offering. Additionally, and as noted in
paragraph
8.244, we find that the in-store offering of multi-channel retailers
does not pose a material competitive constraint on the Parties, implying
that their in-store presence is not significantly strengthening their constraint
on the Parties.
With respect to diversion from 247, the BDRC Survey implies a diversion
ratio of 34% to Blinds2Go and of 36% across all mentioned Hunter Douglas
brands (Blinds2Go, Hillarys and Web Blinds). This diversion ratio is, as the
Parties submitted, smaller than market shares suggest. However, we note
that the survey results still imply that by far the largest proportion of 247
customers would divert to Blinds2Go / Hunter Douglas brands, suggesting
that Blinds2Go / Hunter Douglas is the key competitive constraint on 247.
Additionally, we consider that, as discussed in paragraph
8.126, the
reported diversion to large multi-channel retailers is likely subject to an
upward bias, implying that reported diversion to the Blinds2Go and 247
(and other online M2M blind retailers) is likely subject to a downward bias,
which may at least in part explain the discrepancy between diversion ratio
and market shares.
The BDRC Survey also indicates which websites the Parties’ customers
visited prior to placing their order. The results show a similar picture to the
diversion ratios. While a higher proportion of Blinds2Go’s customers visited
the website of Blinds Direct (part of Interior Goods Direct) and Dunelm, a
significant proportion, namely 12%, visited 247’s website. The survey also
shows that Blinds2Go was the website visited by the highest proportion of
247’s customers, namely 33%.
Overall, we find that the results of the BDRC Survey are consistent with the
Parties being close competitors. We additionally note that the competitive
123
constraint from Blinds2Go on 247 appears to be stronger than the
constraint from 247 on Blinds2Go.
Internal documents
In the course of our inquiry we have reviewed more than 1,200 documents
provided by the Parties at phase 2.
292
These were provided in response to
requests for internal documents relating to:
(a) Internal e-mails from the Parties discussing competitors, monitoring, and
the competitive landscape;
(b) Data and internal documents that show the monitored products and
respective prices of competitors;
(c) Documents and correspondence relating to marketing, and in particular
Google advertising; and
(d) A small number of Board reports and strategic documents related to the
rationale for the Transaction.
The Parties have told us that neither of the Parties’ strategic decision-
making is usually documented in board reports or other strategic
documents, but rather that decisions tend to be taken ad-hoc and by small
groups of individual decision-makers. In general, the internal documents
have been of a day-to-day rather than strategic nature.
We have received a large number of internal documents on the Parties’
monitoring activities. These internal documents show, in line with the
Parties’ submissions, []. Blinds2Go records the prices of [] for a
number of different M2M blinds of a certain size on a weekly basis. 247
scrapes the website of [], collecting the prices of [].
293
These internal documents further show that the Parties also monitor certain
other retailers. In particular:
(a) Blinds2Go records similar price information to that collected for 247 for
[];
294
and
292
While we set out below our analysis of the internal documents we have reviewed as they relate to the
closeness of competition between the Parties, we also note that the internal documents do not in general support
the Parties' submissions that ready-made blinds, blinds sold through alternative channels and alternative window
covering pose a competitive constraint on the Parties.
293
The Parties described this monitoring as ‘ad hoc’. While we agree that monitoring is not performed at regular
intervals, there appears to be extensive documentation recording such monitoring: [].
294
[].
124
(b) [].
295
While this implies that Blinds2Go and 247 do not only monitor the prices of
each other, the set of other retailers monitored is relatively small. In
particular, 247 is one of [] retailers whose prices Blinds2Go regularly
monitors (for non-motorised M2M blinds), while Blinds2Go is one of []
retailers that 247 monitors.
The Parties further submitted [].
296
The Parties further submitted that 247 only monitors Blinds2Go and []
because of resource and time constraints (noting that coding/setting up the
price scraping tool requires a significant upfront investment and that there
is also a significant ongoing manual task associated with the exercise) and
because it sees them as a ‘proxy’ for all of its competitors. In our view, this
approach to monitoring does not contradict that 247 sees Blinds2Go and
[] as its key competitive constraints and in fact supports it.
Overall, we find that the evidence on the Parties’ monitoring is consistent
with the Parties being close competitors, and consistent with there being
few other retailers that the Parties view as a significant competitive
constraint.
We have identified two further internal documents that speak to the
closeness of competition between the Parties. We find that these also
indicate that the Parties are close competitors:
An email from [] (both 247) from December 2018 requests that 247’s
prices for entry level roller blinds should be adjusted [].The email also
refers to a similar adjustment for []. These adjustments show that 247
used the price scraping information it collects to adjust its prices,
positioning itself as marginally cheaper than Blinds2Go and Interior
Goods Direct.
An internal 247 email from July 2020 notes that ‘[]’. The expectation
that a coordinated price increase by the Parties would not result in a
decrease in sales indicates that the main competitive constraint on each
of Blinds2Go and 247 is the respective other Party.
297
295
247’s [].
296
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.69.
297
The CMA notes that any agreement between Hunter Douglas or Blinds2Go and 247 to fix or coordinate their
retail prices may breach the Chapter I prohibition in the Competition Act 1998 (as well as the equivalent EU
prohibition). This report does not consider whether this correspondence discloses behaviour that is contrary to
the Competition Act 1998.
125
The internal documents form part of a wider evidence base. The
documents are consistent with our finding that Blinds2Go and 247 closely
monitor online M2M blind retailers, and do not monitor other retailers
closely. They are also consistent with the Parties being close competitors.
Pricing analysis
The Parties submitted that they used the pricing information collected by
Blinds2Go through its monitoring of competitors (see paragraphs 8.137 and
8.138) to assess pricing trends over time:
(a) The Parties assessed the extent to which weekly price changes by
Blinds2Go correspond with other retailers moving their prices in the same
direction across the various product categories.
298
(b) The Parties calculated the correlation between the weekly percentage
changes of prices for each retailer pair.
Our assessment of this analysis, as well as our own analyses, are set out
in Appendix D.
Overall, in our view the analysis submitted by the Parties is not suitable for
determining the extent to which the prices of different retailers move
together. We also find that our own analyses do not provide strong
evidence that Blinds2Go and 247’s prices follow each other more closely
than the prices of other retailers. However, there is no indication that there
is another retailer that Blinds2Go and 247 follow more closely.
In response to this, the Parties submitted that, given our findings, they do
not understand how it could be concluded that Blinds2Go’s prices are in
any way constrained by 247’s prior to the 2019 Transaction, nor that there
would be an appreciable loss of price competition as a result of the 2019
Transaction.
299
As set out above, we find that the price correlation is low
between all retailers which we assessed. This finding highlights the
limitations of looking at price correlation and shows that it is important to
consider all factors in the round, rather than focusing on price correlation.
Third party evidence
As part of our investigation, we have contacted all third parties for which
the Parties provided contact details. We received written responses on how
and with whom they compete from eight third-party retailers of online M2M
298
Main submission, 20 May 2020, paragraph 6.81.
299
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.60.
126
blinds and further conducted detailed calls with eight third parties. We
therefore consider that we have extensively sought evidence from third
parties.
Comments from third parties on how they perceive Blinds2Go and 247
suggest that they are close competitors with each other:
(a) Interior Goods Direct told us ‘Most online customers value a good value
product and will then shop from Blinds2Go or 247 because of their lower
prices in comparison to competitors.’
(b) Others, including MakeMyBlinds, said that Blinds2Go and 247 are
frequently ‘ranked number one and two on Google for most search
queries’.
(c) We asked four third parties whether they viewed Blinds2Go and 247 as
close competitors. All four confirmed that they did.
All online M2M blind retailers that submitted information on which retailers
they monitor told us that they monitor both Blinds2Go and 247. Notably, the
only other retailer apart from Blinds2Go and 247 that was monitored by all
of these online M2M blind retailers is Interior Goods Direct. While this alone
does not necessarily imply that the Parties are close competitors, it shows
that Blinds2Go and 247 are, alongside Interior Goods Direct, considered by
other online M2M blind retailers to be the key retailers of online M2M
blinds.
Lastly, we note that no third party we spoke to provided evidence that
conflicted with our view that Blinds2Go and 247 are close competitors.
Competitive significance of 247
The Parties submitted that the market share of 247 has steadily declined.
They submitted that 247’s market share was in a range of between 10 to
20% in 2016, and that it declined to a range of between 5 to 10% in 2019.
With respect to the constraint from 247 on Blinds2Go, the Parties further
submitted that such constraint is limited and declining. The Parties
submitted that the CMA has failed to address how a retailer as small as
247 imposes any significant constraint on the market leader Blinds2Go.
300
Additionally, the Parties submitted that [].
301
300
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.9.
301
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.113.
127
Finally, the Parties submitted that by basing its provisional conclusion on
the Merged Entity’s combined market share and the merger removing a
direct competitor, the CMA has asked the wrong questions, as all mergers
between competitors have ‘the effect of removing a direct competitor from
the market’ and it is not sufficient in and of itself for the CMA to find an
SLC.
302
In our view, 247’s decline in market share is not particularly significant for
the following reasons:
(a) First, despite a loss in market share, 247 is still the third largest retailer in
the market for online M2M blinds, and in terms of sales, is relatively close
behind the second largest retailer, Interior Goods Direct.
(b) Second, in absolute terms, 247’s sales of online M2M blinds have grown
by more than []% in each of the last two years. While the Parties
submitted that the absolute growth of 247 is irrelevant,
303
we find that
247’s ability to grow its sales is a relevant observation.
(c) Third, the decline in market share has been relatively limited, from [10-
20]% in 2016 to [5-10]% in 2019. While the Parties submitted that this is a
decline by []% and hence [],
304
we note that when expressed in
percentage points, the decline only amounts to less than [] percentage
points over a period of three years.
We also note that the Parties’ submission on the limited significance of 247
does not seem to align with the Parties noting that a merger between 247
and Interior Goods Direct, a retailer much smaller than Blinds2Go, would
raise competition concerns.
In response to the Parties’ submissions on the constraint from 247 on
Blinds2Go, we note that while significantly smaller than Blinds2Go, 247 is
still the second largest competitor to Blinds2Go, closely behind Interior
Goods Direct. The Parties have also not submitted any evidence in support
of the claim that [].
Finally, we note that our assessment is not merely based on the Merged
Entity’s combined market share and the merger removing a direct
competitor. We assessed the extent to which the 2019 Transaction has a
significant effect on rivalry and the competitive pressure on the Parties by
taking into account a range of evidence, with this evidence showing that the
302
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.2.
303
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.13.
304
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.13.
128
Parties are two of the three leading retailers of online M2M blinds and that
the Parties are close competitors.
Conclusion on closeness of competition
Based on the assessment set out above, we find that:
(a) Blinds2Go and 247 have a similar service proposition, in terms of price,
quality and product range;
(b) Both feature prominently in Google search results;
(c) Survey evidence on the reported diversion of Blinds2Go’s and 247’s
customers is consistent with them being close competitors with each other
(high rates of diversion from 247 to Blinds2Go at 34% and lower but still
significant diversion from Blinds2Go to 247 at 13%), with the constraint
from Blinds2Go on 247 appearing to be stronger than the constraint from
247 on Blinds2Go;
(d) Blinds2Go and 247 monitor each other’s prices and each is part of a
relatively small set of retailers that the respective other regularly monitors;
and
(e) Third parties have told us that they see Blinds2Go and 247 as being close
competitors to each other.
Overall, we therefore find that the Parties are close competitors.
Remaining constraints
Unilateral effects are more likely where customers have little choice of
alternative supplier. In this section, we consider the strength of the
competitive constraint on the Merged Entity from alternative retailers. In
their submissions, the Parties identified the following constraints:
(a) Other online M2M blind retailers, including Interior Goods Direct, Swift
Direct Blinds, MakeMyBlinds and a number of smaller online M2M blind
retailers;
(b) Multi-channel retailers, including Next, Dunelm and John Lewis;
(c) Online marketplaces such as Amazon and eBay; and
(d) Out-of-market constraints.
129
We assessed the level of constraint from each of these alternatives by
taking into consideration:
(a) The similarity of their service proposition compared to the Parties;
(b) Their online presence;
(c) Evidence from Blinds2Go and 247’s customers in the form of the BDRC
Survey submitted by the Parties;
(d) Evidence on monitoring; and
(e) Views from third parties.
Competition from other online M2M blind retailers
With respect to competition from other online M2M blind retailers, the
Parties submitted that Blinds2Go and 247 are constrained not only by
larger online M2M blind retailers, but also by smaller ones, that can and do
compete with Blinds2Go and 247 given the lack of economies of scale and
capacity constraints. The Parties specifically highlighted ‘the breadth of
competition on quality and the strength of a number of competitors’.
305
The
Parties also submitted that the Parties are one of many effective
competitors in the retail of online M2M blinds.
306
We assess each of Interior Goods Direct, Swift Direct Blinds and
MakeMyBlinds in turn below. With the exception of BlocBlinds, a niche
retailer for specialised roller blinds, these are the only online M2M blind
retailers with a market share of more than 1%. We subsequently also
consider the level of constraint from other smaller online M2M blinds
retailers.
Interior Goods Direct
Interior Goods Direct is the second largest retailer in the online M2M blinds
market, with sales of online M2M blinds amounting to £[]million in 2019.
Interior Goods Direct has grown significantly over the last years, with its
sales of online M2M blinds growing by []% and []% in 2017 and 2018,
respectively. Additionally, in April 2020, Interior Goods Direct acquired
Wilsons Blinds, a smaller retailer with online M2M blind sales of £[] in
2019. Interior Goods Direct is vertically integrated and has been trading
blinds exclusively online since 2003.
305
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.61.
306
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.64.
130
The Parties submitted that Interior Goods Direct is an important and
growing competitor, with survey evidence indicating significant diversion
from the Blinds2Go and 247 to Interior Goods Direct.
307
The Parties further
submitted that the vertical integration of Interior Goods Direct allows it to
stock over 1,000 M2M products offering next day delivery. The Parties
noted that they expect that Wilsons Blinds will benefit from accelerated
growth through the benefits of Interior Good Direct’s vertical integration.
308
The offering of Interior Goods Direct is broadly similar to that of Blinds2Go
and 247. In particular:
(a) Our analysis of prices (see paragraph 8.146) indicates that the prices of
Interior Goods Direct are similar to those of Blinds2Go and 247.
(b) In terms of quality, Interior Goods Direct markets itself as a slightly higher
quality producer, noting that manufacturing its products in the UK helps in
this regard. However, Interior Goods Direct has a lower Trustpilot score
than either Blinds2Go or 247.
309
(c) While the Parties claimed that Interior Goods Direct offers next day
delivery, a review of the Interior Goods Direct website showed that it only
offers a ‘fast track’ option – which guarantees dispatch within 2 to 3
working days for orders made before 12pm at an extra cost (£4.95).
Interior Goods Direct also states on its website that the estimated dispatch
time does not include the delivery time (one day for deliveries to the UK
Mainland).
310
(d) In terms of range, while Interior Goods Direct has a larger overall range
(in terms of SKUs) than Blinds2Go and 247, it offers significantly fewer
SKUs for wooden blinds – which is, as discussed in paragraph 8.63
above, the top-selling type of blind for each of the top four online M2M
blind retailers.
311
Interior Goods Direct appears to be less successful than either Blinds2Go
or 247 at ranking highly on Google paid search, although it does
outperform 247 in some cases: our analysis of Google paid search
(paragraph 8.90) found that Interior Goods Direct features significantly less
frequently (and in lower positions) in the paid Google Ads search results
than Blinds2Go and, to a lesser extent, 247 (Interior Good Direct does
307
Main submission, 20 May 2020, paragraphs 6.38, 6.46, 6.60, 6.61, and 6.62.ii.
308
Main submission, 20 May 2020, paragraphs 6.46.
309
Interior Goods Direct has a Trustpilot score of 87%, while Blinds2Go and 247 have a Trustpilot score of 97%
and 91%, respectively Parties’ response to Issues Letter, paragraph 3.62, 4 March 2020.
310
Parties’ main submission para 6.46, and Direct Blinds website, accessed on 30 June 2020.
311
In particular, while the Parties offer more than 200 SKUs of wooden blinds each, Interior Goods Direct only
offers 80.
131
outperform 247 in some cases). Interior Goods Direct’s performance
regarding organic search is better, with it tending to appear more frequently
than 247 (though still less frequently than Blinds2Go) in the top positions.
The BDRC Survey submitted by the Parties shows significant diversion
from each of Blinds2Go and 247 to Interior Goods Direct – 16% for
Blinds2Go customers and 14% for 247 customers. We further note that
34% of Blinds2Go customers and 25% of 247 customers indicated that
they visited the website of Blinds Direct (part of Interior Goods Direct) prior
to their purchase, while only very small proportions of either Blinds2Go and
247 customers indicated that they visited the website of Wilsons Blinds.
With respect to evidence on monitoring, we note that []. However, and as
discussed in Annex D, we found little correlation between the prices of
Interior Goods Direct and those of the Parties. Additionally, we note that
Interior Goods Direct is monitored by all online M2M retailers that submitted
information on which competitors they monitor. Interior Goods Direct is
further monitored by [] (it is the only online M2M retailer other than the
Parties that is monitored by []) but not by [].
Finally, we note that the conversion rates of Interior Goods Direct are
significantly lower than those of Blinds2Go and slightly lower than those of
247.
312
This suggests that Interior Goods Direct is significantly less
effective than Blinds2Go at converting visitors to its websites to sales.
Based on the evidence set out above, we find that Interior Goods Direct is
a significant competitive constraint on the Parties, and that its constraint on
Blinds2Go is similar to the constraint 247 poses on Blinds2Go: Interior
Goods Direct has a similar offering to the Parties, it is the second largest
supplier in the market and of similar size to 247, and survey evidence
shows significant diversion to Interior Goods Direct – although we note that
Interior Goods Direct is less often ranked highly on paid search than the
Parties and is also less effective than the Parties in terms of conversion.
Swift Direct Blinds
With sales of online M2M blinds amounting to £[]million in 2019, Swift
Direct Blinds is the fourth largest retailer in the online M2M blinds market
and has a considerably smaller share than 247 (the third largest retailer).
313
312
For user conversion, the rates are []% for Blinds2Go, []% for 247 and []% for Interior Goods Direct. For
session conversions, the rates are []% for Blinds2Go, []% for 247 and []% for Interior Goods Direct. While
Blinds2Go told us that the e-commerce conversion rates for Blinds2Go are not reliable as Google Analytics
counts sample orders as transactions, the figures shown above were submitted by the Parties without this
caveat, so we understand these to be reliable and not include samples as transactions. Main submission, [].
313
Main submission, 20 May 2020, para 7.47.ii and [].
132
Swift Direct Blinds’ sales of online M2M blinds decreased [] in 2019, by
£[]million compared to 2018, and [] [Another retailer] told us that Swift
Direct Blinds has been struggling over the last year or two.
314
Swift Direct
Blinds is a vertically integrated supplier. While it was active as an in-store
retailer in the 2000’s, Swift Direct Blinds moved online in 2010.
315
The Parties submitted that Swift Direct Blinds is a key online retailer of
M2M blinds and that it has registered significant growth since its launch in
2012.
316
The Parties further submitted that Swift Direct Blinds concentrated
its efforts on the Google organic channel and regularly featured in the
number one position.
317
Finally, the Parties submitted that Swift Direct
Blinds’ confidence in the quality of their blinds is such that they offer a five-
year guarantee on all products.
318
Swift Direct Blinds’ offering is broadly similar to that of the Parties. In
particular:
(a) Our analysis of prices (see Annex D) indicates that the prices of Swift
Direct Blinds are similar to those of the Parties, although we note that
Swift Direct Blinds’ prices are significantly higher than those of either of
the Parties for venetian blinds;
(b) In terms of quality, Interior Goods Direct told us that the product quality of
Swift Direct Blinds is similar to that of the Parties. We also note that Swift
Direct Blinds’ Trustpilot score is similar to those of the Parties.
319
We
acknowledge the Parties’ submission that Swift Direct Blinds offers a five-
year guarantee on all products. However, we note that offering several
years guarantee is a common practice in the market, and therefore we do
not consider it as a significant element of differentiation.
320
(c) In terms of range, we find that Swift Direct Blinds offers a similar number
of SKUs as the Parties, both overall and with respect to wooden blinds,
the top-selling type of blind for each of the top four online M2M blind
retailers.
Swift Direct Blinds is less successful than the Parties at ranking highly on
Google paid search, with our analysis of Google paid search showing that
314
[].
315
[].
316
Main submission, 20 May 2020, paragraph 7.47.ii.
317
Main submission, 20 May 2020, paragraph 7.19.
318
Main submission, 20 May 2020, paragraph 6.49.
319
Parties’ response to Issues Letter, paragraph 3.62, 4 March 2020.
320
247 offers three years guarantee, MakeMyBlinds four years and Swift Direct Blinds, Interior Goods Direct and
Blinds2Go five years.
133
Swift Direct Blinds very rarely features in the top positions of paid Google
Ads.
With regard to organic search, our analysis of Google organic rankings
shows that, contrary to the Parties’ submission, Swift Direct Blinds does not
feature prominently: the highest position in which Swift Direct Blinds
showed was the fourth position, and Swift Direct Blinds occupied this
position in only 4% of the conducted searches. We also note that Swift
Direct Blinds told us that its organic positioning on Google suffered after it
changed its website URL (from Direct Blinds to Swift Direct Blinds) in
August 2019, and provided us with the respective underlying data (see
Appendix F).
The BDRC Survey submitted by the Parties shows very limited diversion
from each of the Parties to Swift Direct Blinds – less than 1%. We consider
this to be strong evidence of the limited constraint imposed by Swift Direct
Blinds on the Parties. We further note that only 2% of Blinds2Go’s
customers and 3% of 247’s customers indicated that they visited Swift
Direct Blinds’ website prior to their purchase. This is further evidence that
Swift Direct Blinds only poses a limited competitive constraint on the
Parties.
With respect to evidence on monitoring, we note that [], other online
M2M retailers (including 247) and multi-channel retailers do not tend to
monitor Swift Direct Blinds. The only exception to this is.
321
[], and as
discussed in Appendix D, we found little correlation between the prices of
Swift Direct Blinds and Blinds2Go (and likewise 247).
Finally, we note that the conversion rates of Swift Direct Blinds are
significantly lower than those of the Parties. This implies that Swift Direct
Blinds is significantly less effective than the Parties at converting visitors to
its website to sales.
322
Based on the evidence set out above, we find that while Swift Direct Blinds
appears to have a similar offering to the Parties, it performs significantly
less well in terms of its online presence and its effectiveness in terms of
conversion. We also find that its revenue and market share are significantly
smaller than 247’s, and that its sales have significantly decreased in 2019.
321
[].
322
For user conversion, the rates are []% for Blinds2Go, []% for 247 and []% for Swift Direct Blinds. For
session conversions, the rates are []% for Blinds2Go, []% for 247 and []% for Swift Direct Blinds. While
Blinds2Go told us that the e-commerce conversion rates for Blinds2Go are not reliable as Google Analytics
counts sample orders as transactions, the figures shown above were submitted by the Parties without this
caveat, so we understand these to be reliable and not include samples as transactions. Main submission, 20 May
2020, Figure 6.8.
134
We therefore conclude that the current competitive constraint from Swift
Direct Blinds on the Parties is limited.
However, and as also pointed out by the Parties,
323
we acknowledge that
Swift Direct Blinds was recently acquired by Decora, and that this may
impact its future performance. We consider this in more detail in our
assessment of entry and expansion (see paragraphs 9.56 to 9.63 below).
MakeMyBlinds
MakeMyBlinds is a relatively new and small player in the online M2M blinds
market, having entered the market in 2015. In 2019, MakeMyBlinds’ sales
of online M2M blinds amounted to £[], leading to a market share of [0-
5]%.
324
The Parties submitted that MakeMyBlinds’ growth has been due, in part, to
an effective online marketing strategy that has focussed not just on PPC
advertising but also on social media, in particular Instagram. Consistent
with this, Interior Goods Direct told us that MakeMyBlinds targets younger
customers.
While MakeMyBlinds offers all main types of blinds and has a website with
similar functionalities to those of the Parties, there appear to be certain
differences between MakeMyBlinds’ offering and those of the Parties:
(a) While we do not hold detailed pricing data from MakeMyBlinds,
MakeMyBlinds told us that it ‘used to be very price conscious against
Blinds2Go and 247’, but that it was not making the required margin and
therefore no longer price matches. Instead, it tries to position its blinds as
quality products.
(b) MakeMyBlinds’ range in terms of SKUs is significantly more limited than
those of the Parties: MakeMyBlinds offers 820 SKUs, while Blinds2Go
and 247 offer 4,497 and 2,823 SKUs, respectively.
(c) MakeMyBlinds’ Trustpilot score is comparable to that of the Parties.
However, we note that MakeMyBlinds has significantly fewer reviews,
namely around 1,000 while Blinds2Go and 247 each have more than
10,000 reviews. This is likely to be a function of MakeMyBlinds having
been in the market for less time. However, the fact that they have been
active for less time and hence have fewer reviews is itself likely to
disadvantage them to some extent over more established firms.
323
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.71.
324
[].
135
MakeMyBlinds appears to be less successful than the Parties at ranking
highly on Google search: our analysis found that MakeMyBlinds only
features sporadically in the paid Google Ads search results and never
appears among the organic Google search results on the first page. While
we acknowledge that MakeMyBlinds spends a higher proportion of its
marketing spend on social media as compared to other online M2M blind
retailers, we note that Google paid and organic search are still the largest
two sources of revenue for MakeMyBlinds, with Google paid search being
the channel on which MakeMyBlinds spends the largest proportion of its
marketing spend.
With respect to evidence on monitoring, we note [] even though []
(see paragraph 8.138).This indicates that MakeMyBlinds is not seen as a
strong competitive constraint by either Blinds2Go or 247. For
completeness, we note that MakeMyBlinds is monitored by [].
The BDRC Survey submitted by the Parties shows very limited diversion
from each of the Parties to MakeMyBlinds – less than 1%. We consider this
to be strong evidence that MakeMyBlinds imposes only a limited
competitive constraint on the Parties. We further note that only 2% of
Blinds2Go’s and 1% of 247’s customers indicated that they visited
MakeMyBlinds’ website prior to their purchase. This is further evidence that
MakeMyBlinds only poses a limited competitive constraint on the Parties.
Based on the evidence set out above, MakeMyBlinds appears to have a
similar offering to the Parties, albeit its range is more limited.
MakeMyBlinds also appears to be less successful than the Parties at
ranking highly on Google search and it currently is a small player in the
online M2M blinds market.
325
We therefore find that MakeMyBlinds
currently only poses a limited competitive constraint on the Parties.
However, we note that MakeMyBlinds has experienced significant growth in
the past years, and that it has plans for significant further growth. We
consider this in more detail in our assessment of entry and expansion (see
paragraph 9.64 below).
Other smaller online M2M blind retailers
In addition to the online M2M blind retailers discussed above, the Parties
submitted that there is a large number of other online M2M blind retailers
with a similar offering and that the aggregate competitive constraint from
‘many smaller players’ should not be ignored. The Parties provided a
325
We do not hold data on MakeMyBlinds’ conversion rates and are hence unable to comment on the
effectiveness of MakeMyBlinds in terms of converting visitors to its website into sales.
136
selection of screenshots of the websites of such other online M2M retailers
and further specifically discussed the following retailers:
326
(a) The Parties submitted that Blinds4UK has over 35 years’ experience in
window coverings, that its blinds featured on BBC Grand Designs and In-
House Beautiful magazine and that it has an excellent rating on Trustpilot,
with over 3,000 reviews.
327
The Parties further submitted that the BDRC
survey shows that approximately a quarter of customers visited Blinds4UK
before making their purchase with the Parties, and that this is evidence of
Blinds4UK’ competitive constraint.
328
(b) The Parties submitted that Terrys Fabrics was established 46 years ago,
that it offers payment through a credit provider and that it has been
featured on ITV’s This Morning and in the Good Housekeeping
magazine.
329
The Parties further submitted that in August 2019, Terrys
Fabrics released an augmented reality window dressing app to help
customers visualise its blinds in their home before placing an order.
330
(c) The Parties submitted that Concept Blinds is rated 4.9 out of 5 on
Trustpilot and has been online for 14 years.
331
The Parties further
submitted that Concept Blinds claims to have the lowest UK prices and
that it does its best to match any listed price for an equivalent product on
another online site 48 hours after purchase.
332
Whilst we acknowledge that there are a number of smaller online M2M
blind retailers, we note the following:
(a) With the exception of Interior Goods Direct, Swift Direct Blinds and
MakeMyBlinds (and the specialised retailer Blocblinds), all other online
M2M blind retailers are small and have remained small, despite having
been active in the market for a number of years. Even the largest of the
other online M2M blind retailers, OrderBlindsOnline, has a market share
of less than []%, with sales of online M2M blinds amounting to only
£[] million in 2019. The Parties submitted that the combined market
share of these smaller retailers was 20%. However, the evidence we have
seen does not support such a high estimate. Based on the actual sales
data we received from the smaller retailers that responded to our request
326
Main submission, 20 May 2020, Annex 0099.
327
Main submission, 20 May 2020, paragraph 6.47.
328
Main submission, 20 May 2020, paragraph 6.47.
329
Main submission, 20 May 2020, paragraph 6.50.
330
Main submission, 20 May 2020, paragraph 6.50.
331
Main submission, 20 May 2020, paragraph 6.51.
332
Main submission, 20 May 2020, paragraph 6.51.
137
for sales data, other online M2M blind retailers only account for a market
share of 2%.
333
(b) The Parties do not consider these smaller retailers in their strategic
decision making. For example, with the exception of [].
(c) The other online M2M blinds retailers do not tend to rank frequently in
paid Google Ad search results or the first page of Google organic search
results.
334
(d) The BDRC Survey submitted by the Parties shows limited diversion from
each of the Parties to the other online M2M retailers. The highest
diversion is to Blinds4UK (4% for Blinds2Go customers and 5% for 247
customers), with diversion to all other online M2M retailers in aggregate
amounting to less than 1% for Blinds2Go and 247 customers. Additionally,
the survey shows that the Parties’ customers did not tend to visit the
websites of the other online M2M retailers prior to their purchase. The
exception is Blinds4UK, although we note that, in contrast to the Parties’
submission that around a quarter of customers visited Blinds4UK, the
BDRC Survey actually shows the proportions were lower: 11% of
Blinds2Go customers visited Blinds4UK and 18% of 247 customers visited
Blinds4UK.
335
(e) The Parties have not provided evidence to show how they are constrained
by the smaller online M2M blind retailers, either individually or in
aggregate.
Based on the above, we find that the smaller online M2M blind retailers do
not, either individually or in aggregate, constitute a significant competitive
constraint on the Parties: they have not managed to grow despite having
been active in the market for a number of years, they are not able to attract
a significant number of potential customers and they do not appear to
influence the Parties’ behaviour.
Conclusion on constraint from online M2M blind retailers
We find that Interior Goods Direct is a significant competitive constraint on
the Parties, and that its constraint on Blinds2Go is similar to the constraint
247 poses on Blinds2Go.
333
See Table 3, where the market shares of Order Blinds Online Ltd, Blinds4UK, Meadow Blinds Ltd / Lifestyle
Blinds Ltd, and Others (with turnover <0.5m) add up to 2%.
334
See paragraphs 8.90, 8.93 and 8.103, as well as the more detailed discussion in Appendix E.
335
The Parties’ appear to only count customers that in a preceding question indicated that they visited other
websites, rather than counting all customers.
138
Apart from Interior Goods Direct, the constraint from other online M2M blind
retailers on the Parties is relatively weak. This is consistent with the view
that there are few credible sizeable online M2M blind retailers.
Competition from multi-channel retailers
Multi-channel retailers are retailers that are active in several channels,
namely both the offline channel (in-store and/or in-home) and the online
channel. In this section, we focus on the online M2M blind offering of these
retailers.
On the basis of the Parties’ submission and responses from third parties,
we understand that Next, Dunelm and John Lewis are the only multi-
channel retailers with an online offering of M2M blinds (although we note
that Dunelm’s website currently does not have click-to-order functionality,
see paragraph 4.34).
336
We note that, as set out in paragraph 8.10 above
the sales of online M2M blinds for each of these multi-channel retailers are
small: in 2019, the sales of online M2M blinds for Next, Dunelm and John
Lewis amounted to £[]million, £[]million and £[]million,
respectively.
337
Similarity of service proposition
We assessed the similarities between the online offering of the multi-
channel retailers and online M2M blind retailers. In particular, we focus on
the most important factors that affect a customers’ choice of retailer: price,
quality and range.
338
Overall, we find that multi-channel retailers’ service propositions tend to be
materially different from those of online M2M blind retailers when assessed
on the above parameters.
Multi-channel retailers tend to price higher than online retailers.
(a) Price comparisons submitted by the Parties suggest that the prices of
John Lewis are significantly higher than those of online M2M retailers
(between 70% more and more than double), although we note that these
comparisons only assess two M2M blind products for John Lewis, namely
roller blinds and roman blinds.
336
While the Parties further included Laura Ashley in the online M2M market shares, it does not appear to be
possible to order M2M blinds through the Laura Ashley website.
337
Corresponding to market shares of [0-5]%, [0-5]%, and [0-5]% respectively.
338
When asked about the reasons that influenced the choice of the retailer, the majority of the Parties’ customers
interviewed responded ‘Good prices/offer’, followed by ‘Had what specifically wanted’, ‘Website easy to use’,
‘Product quality’, and ‘Good/wide product range’.
139
(b) John Lewis noted that its products tend to fall within the mid-range to
higher end of the market pricewise due to a range of factors including its
use of higher quality materials relative to some of its competitors and its
higher overheads compared to online competitors.
(c) Next, the only multi-channel retailer included in the price monitoring data
of Blinds2Go, tends to be among the more expensive retailers of online
M2M blinds. The Parties submitted that Next has cut its prices of roman
blinds and roller blinds since the date of the last price scrape that the
Parties had submitted to the CMA, arguing that Next has therefore
become much more competitive on price.
339
However, the new Next
prices submitted by the Parties are similar to the ones used in our
comparison, implying that even on the basis of the new prices, Next
remains among the more expensive retailers of online M2M blinds.
(d) Blinds2Go states on its website that it compared its prices with the five
main M2M blind retailers on the high street and that ‘in almost 25% of
cases we were cheaper by 60% or more’. It further advertises on its
website that ‘you can also save up to 60% off High St. prices’.
340
While the Parties agreed that multi-channel retailers tend to charge higher
prices, the Parties submitted that the price differential does not suggest that
multi-channel and online M2M blind retailers do not compete, arguing that
customers are likely to make a quality-price trade-off as a brand like John
Lewis has built a reputation for the quality of their products over time. The
Parties also submitted that the CMA has advanced no evidence to support
that differentiation may limit the multi-channel retailers’ competitive
constraint on the Parties, and that the significant diversion to John Lewis
indicates that customers see the offerings as substitutes.
341
We note that
the presence of price and quality differences (which, according to the
Parties, customers trade-off against each other) is an indicator of
differentiation. All else equal, the more differentiated the offering of retailers
are, the lower the competitive constraint these retailers are likely to exert
on each other. While we further acknowledge that some customers may
still see the offering of certain multi-channel retailers as a substitute
(although, as discussed in 8.127, we consider that reported diversion to
large multi-channel retailers is likely to be subject to an upward bias), we
339
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.57.
340
Blinds2Go website, accessed 23 June 2020. Although these comparisons refer to the in-store prices of the
multi-channel retailers, the multi-channel retailers tend to apply the same price online as offline. Next told us that
it charges the same price for its M2M blinds online and in-store (Call with Next, 5 May 2020). John Lewis states
on its website that ‘We apply the same national price to products in our shops and online.’
https://www.johnlewis.com/customer-services/prices-and-payment/never-knowingly-undersold
341
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.82.
140
have not received evidence that multi-channel retailers exert a significant
competitive constraint on the Parties.
In terms of quality, multi-channel retailers may benefit from brand
recognition and the perception that they offer higher quality. For example,
John Lewis noted that it considers itself to be a very trusted brand, taking
pride in quality and design. It also noted that it uses higher quality materials
relative to some of its competitors.
342
Multi-channel retailers offer a significantly smaller online range than the
Parties. In particular, while Blinds2Go and 247 respectively offer 4,497 and
2,823 SKUs, the Parties submitted that Next offers 844 SKUs online, John
Lewis 511, and Dunelm 285.
We note that the more limited range of multi-channel retailers seems to be
driven by system and/or website functionality. For instance, one multi-
channel retailer ([]) told us that whilst it sold some types of blinds online,
it was still not able to provide a full offering via this channel. [] told us that
its webstore is not currently versatile enough to offer larger ranges of highly
bespoke products, such as M2M blinds, and that it currently offers only
[]% of its made-to-measure products online.
343
Both Next and John Lewis have told us that they would like to increase
their range going forward. However, the significance and/or timing of such
any increase remains unclear, particularly given the technical difficulties of
system and/or website functionality noted above, the competing priorities
resulting from selling products other than M2M blinds within these
businesses and the impact of the current COVID-19 pandemic. This is
discussed in further detail in paragraphs
9.76 to 9.80.
The Parties submitted that Next materially improved its offering in Autumn
2019. However, we note that Next was selling online M2M blinds
throughout 2019 and its sales did not change substantially in the course of
2019. Moreover, Next forecasted similar revenues for 2020.
We also note that Dunelm exited the market for the online retail supply of
M2M blinds, although it plans to re-enter. Dunelm currently does not offer
click-to-order functionality and is offering online consultations via Microsoft
Teams. Customers have to look through a catalogue of c.300 pages to
browse the selection and cannot place an order without the assistance of a
342
We acknowledge the Parties’ submission that that Blinds2Go uses the same supplier as John Lewis. We have
not verified whether this implies that Blinds2Go and John Lewis offer comparable quality. However, even if the
quality of Blinds2Go and John Lewis was comparable, it is our view that when considering all of the evidence in
the round, we do not find that John Lewis is a strong constraint on either of the Parties
343
[].
141
consultant.
344
Dunelm told us that they have put their plans for online click-
to-order functionality for M2M blinds on hold as a result of the current
COVID-19 pandemic.
345
Finally, and as a more general point, the Parties submitted that Next,
Dunelm and John Lewis have stated their intention to increase their online
presence.
346
As further discussed in the countervailing factors section
below (paragraphs 9.76 to 9.93), we note that the claims cited by the
Parties are not specific to M2M blinds and that the evidence submitted by
multi-channel retailers contains no significant expansion plans for their
online M2M blinds sales in the near future.
Brand recognition
The Parties submitted that the multi-channel retailers have the ability to
attract customers directly to their websites, that they have a natural
competitive advantage in that customers will frequently visit their websites,
and that they can run large scale email marketing campaigns.
We note that, while in principle this could be the case, in practice we
observe that the multi-channel retailers achieve very low sales in online
M2M blinds despite this advantage.
Additionally, we note the following comments from third parties:
(a) Only one third party considered that high street retailers with well-known
brands may have some advantages in online advertising because of
brand recognition.
(b) One multi-channel retailer told us that it appears in a relatively low
position in the organic search results on Google for the search word
’blinds’ and received only a limited number of site visits through that
search word.
Survey evidence
The Parties submitted that the BDRC Survey shows that the multi-channel
operators are a constraint on the Parties.
347
The Parties submitted that
17% of Blinds2Go’s customers and 15% of 247’s customers would divert to
Dunelm, that 12% of Blinds2Go’s customers and 9% of 247’s customers
would divert to John Lewis and that 5% of Blinds2Go’s customers and 4%
344
[].
345
[].
346
Parties' response to Provisional Findings, 7 August 2020, paragraph 1.12.
347
Main submission, 20 May 2020, paragraph 6.41.
142
of 247’s customers would divert to Next.
348
The Parties also submitted that
Dunelm’s and John Lewis’ websites feature strongly in the BDRC Survey
as websites where the Parties’ customers compare products and prices,
noting for example that 43% and 22% of 247’s customers make
comparisons against Dunelm and John Lewis, respectively.
349
We acknowledge that the BDRC Survey submitted by the Parties shows a
relatively high diversion to (at least some of) the multi-channel retailers.
However, and as set out in 8.138, in our view, that diversion to large multi-
channel retailers (and in particular the diversion to the online offering of
Dunelm and John Lewis) is likely subject to an upward bias, due to
customers being more familiar with these brands but potentially unaware of
the true nature of their offerings (ie that Dunelm required a virtual
appointment and that John Lewis only has a very limited offering online
compared to its in-store range).
Additionally, we note that the diversion to multi-channel retailers includes
diversion to M2M blinds bought online, M2M blinds bought in-store or in-
home, as well as ready-made blinds bought through any of these
distribution channels – and therefore aggregates a number of different
constraints. While we accept that overall diversion is relevant, in our view,
the cited diversion overstates the constraint from the multi-channel
retailers’ online M2M offering (the diversion to the online M2M offering of
these retailers amounts to 10% of Blinds2Go customers and 7% of 247
customers for Dunelm, 8% of Blinds2Go customers and 4% of 247
customers for John Lewis, and 3% of Blinds2Go and 247 customers for
Next).
With respect to the proportion of customers visiting websites, we note that
the Parties appear to only take into account customers that, in a preceding
question, indicated that they visited other websites, rather than taking into
account all customers. As a proportion of all customers, we note that it is
not 43% and 22% of 247’s customers that make comparisons against
Dunelm and John Lewis, respectively, but only 28% and 14%. For
Blinds2Go’s customers, the percentages as a proportion of all customers
are 25% for Dunelm and 16% for John Lewis.
Third party views
The competitors identified by multi-channel retailers were different to those
identified by online M2M blinds retailers, and included both online and in-
348
Main submission, 20 May 2020, paragraphs 6.53, 6.60 and 6.61; Parties' response to Provisional Findings, 7
August 2020, paragraph 4.75.
349
Main submission, 20 May 2020, paragraph 6.40.
143
store retailers. Multi-channel M2M blinds retailers generally identified other
multi-channel retailers and in-home retailers as their main competitors.
However, they also identified some online M2M blind retailers as
competitors:
(a) One multi-channel M2M blinds retailer, [], identified the Parties and
Interior Goods Direct as competitors, although [] also told us that they
were not significant competitors to either of the Parties due to their limited
online presence in M2M blinds.
(b) Another multi-channel retailer, Next, considered Blinds2Go to be the only
online M2M retailer that is a close competitor.
The Parties submitted that these third-party comments show the strength of
competition from multi-channel retailers. We acknowledge that the views of
multi-channel retailers suggest that multi-channel retailers compete at least
to some extent with online M2M retailers. However, we also note that they
do not show that multi-channel retailers are a strong constraint, and this is
consistent with the other evidence we received taken in the round.
Multi-channel retailers appear to have a different commercial focus than
online M2M retailers. In particular, rather than marketing online M2M blinds
specifically, their approach to marketing tends to encompass a broader
range of interior products:
(a) John Lewis told us that its online competitors can focus on M2M blinds
while John Lewis is a department store selling many products. Its
marketing efforts are focused more on the entire home as opposed to
blinds specifically.
(b) Dunelm told us that it has a wider range of products to promote using its
marketing spend.
(c) Next emphasised a customer’s ability to purchase from a whole range of
interior products. Next told us that its catalogue contains its entire product
range and that it is not specific to blinds. Next also told us that while it
performs general market research, it does not perform market research
specific to blinds. Next told us that it places emphasis on its ready-made
fabrics and in-house design. Finally, we note that Next told us that it is
focused on selling M2M products to expand the ready-made selection with
additional sizes.
The Parties submitted that the ability of multi-channel retailers to market
‘the home’ holistically is a distinct advantage as compared to a retailer
144
focusing exclusively on blinds.
350
The Parties further submitted that if
having a different focus implies that multi-channel retailers do not impose a
competitive constraint, then, having a different commercial focus would
mean that multi-channel retailers would never impose a constraint on single
product firms.
351
We acknowledge that the ability to use a wider marketing approach can
constitute an advantage. However, we note that despite this apparent
advantage, the online M2M blind sales of multi-channel retailers have
remained small. Additionally, we do not find that multi-channel retailers do
not impose a competitive constraint due to their different focus – we find
that their different commercial focus is one of many characteristics that
differentiate multi-channel retailers from online M2M blind retailers.
Internal documents and monitoring
The Parties do not consistently monitor the prices of multi-channel retailers.
[].
The Parties submitted that Blinds2Go []. At the same time, the Parties
submitted that [].
352
However, the Parties have not provided any evidence of such informal
monitoring. []., [].
The Parties further submitted that the latest AMA Report highlights the
constraint from multi-channel retailers.
353
We acknowledge that the
excerpts quoted by the Parties suggest at least some competition between
multi-channel retailers and online M2M blind retailers. However, this does
not imply that multi-channel retailers pose a significant competitive
constraint on online M2M blind retailers. In particular, the AMA report does
not provide any detail on the strength of the competition from multi-channel
retailers. Additionally, when taking the overall body of evidence in the
round, we find that there is not sufficient evidence for us to conclude that
the constraint from multi-channel retailers is material.
Conclusion on multi-channel M2M blinds retailers
Based on the evidence set out above, we note that multi-channel retailers
had very limited online sales of M2M blinds in 2019. We further note that
their online offering is materially different to that of the Parties and other
350
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.84.
351
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.84.
352
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.69.
353
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.88.
145
online M2M blind retailers: multi-channel retailers offer online M2M blinds
at an often substantially higher price point than online M2M retailers and
their online range is more limited. Compared to online M2M blind retailers,
multi-channel retailers have a different business model and commercial
focus, which is not specifically on the M2M blinds product category, and
they feature significantly less prominently in online search results.
While survey evidence indicates material diversion to Dunelm and John
Lewis online, these results are likely to be biased upwards due to
customers being more familiar with these brands and/or customers not
realising that Dunelm no longer offers a click-to-order functionality, while
John Lewis’ range online is very limited.
Finally, we note that multi-channel retailers are not consistently monitored
by the Parties and most other online M2M blind retailers (although Next is
monitored by Blinds2Go, and Interior Goods Direct monitors various multi-
channel retailers).
We therefore find that multi-channel retailers only exert very limited
competitive constraints on the Parties and are not an effective alternative
for most of the Parties’ customers. However, we assess in more detail in
paragraphs 9.76 to 9.94 whether the constraint they exert may increase
going forward.
Competition from eBay and Amazon
In the UK, the main marketplaces on which retailers sell M2M blinds are
eBay and Amazon. Marketplaces are primarily e-commerce sites that
provide a platform for other retailers (ie marketplaces do not tend to be
retailers in their own right). While we note that Amazon also directly sells
certain products on its platform (‘first-party sales’), Amazon does not have
any first-party sales of M2M blinds.
As set out in paragraph 8.7(a), eBay and Amazon submitted that sales of
blinds (including both M2M blinds and ready-made blinds, and in the case
of eBay further including shutters) on their marketplaces amounted to £[]
and £[], respectively, in 2019. While neither eBay nor Amazon was able
to provide a separate sales figure for M2M blinds only, eBay noted that only
approximately 18% of the blinds listed on eBay are M2M.
All of the main types of M2M blinds (ie venetian blinds, wooden and faux
wooden blinds, roller blinds, roman blinds and vertical blinds) appear to be
available on each of eBay and Amazon. However, it appears that eBay is
146
currently not configured to effectively sell M2M products in a way that is
comparable to the Parties’ own websites. In particular:
(a) eBay told us that within a sellers’ listing, it is not possible to have a data
field for buyers to specify the exact dimensions of the blinds. eBay further
told us that the only way to sell M2M products on its platform is for the
seller to state in its listing that the buyer should message the seller with
their exact measurement requirements. One retailer that used to sell on
these platforms told us that this is a slow and cumbersome way to sell.
Another retailer which currently sells on these platforms confirmed that it
considers this to be a ‘messy process’.
(b) While sellers can alternatively display several drop-down menus with a
large number of entries on the product pages (which would in principle
allow for an offering that to some extent resembles M2M), we consider
this option to not be particularly user-friendly. Indeed, eBay confirmed that
sellers do not tend to provide too many increments on their listings since
this is considered to be not appealing to customers.
(c) Additionally, eBay told us that it is not possible to offer free products on
eBay, and hence not possible to offer a free sample. While some sellers
may offer samples for a small fee and then reimburse that fee to the
customer later, eBay submitted that this process is less attractive for
customers than being able to receive a free sample.
(d) Finally, we note that contacting a seller on eBay is more complex than
contacting an online M2M blind retailer. First, it is only possible to contact
the seller on eBay if the buyer has an eBay account. Second, eBay does
not offer instant chat functionality. Third, eBay submitted that the platform
is not quite designed for potential buyers to contact sellers and that there
could be practical issues. However, we acknowledge the Parties’
submission that Blinds2Go and 247 introduced chat functionality on their
websites only recently and that only a very limited amount of customers
used it which may indicate that chat functionality is not a decisive factor
for many customers.
With respect to Amazon, we acknowledge that Amazon has recently
introduced a ‘customise now’ button, which allows customers to enter the
desired measurements of a product before the purchase. We also
acknowledge that this feature makes Amazon more attractive for selling
online M2M blinds.
354
354
This was also noted by the Parties. Parties' response to Provisional Findings, 7 August 2020, paragraph 5.72.
147
The Parties further submitted that Amazon offers a free sample service.
However, it appears that it is not possible to directly order a free sample:
Amazon told us that sellers wanting to offer free samples would have to do
so via the ‘contact the seller’ features. Amazon was not aware of a
functionality that enable free samples to be offered otherwise.
355
Additionally, we note that the options to filter results on Amazon are very
limited compared to websites of online M2M blind retailers. For example, it
is not possible to filter by colour, and other filters (eg for fabric) either do
not exist or are limited to few choices. In our view, this is an important
limitation given that the BDRC Survey shows that 45% and 47% of
respondents for Blinds2Go and 247, respectively, indicated that ‘website
easy to use’ is a factor that influenced their choice of retailer. Also, and
similar to eBay, Amazon does not offer instant chat functionality.
The Parties submitted that there is no specific evidence that filters and
[].
356
However, we find that the presence of filters is likely to be an
important characteristic given the number of different options (eg type of
blind, colour, fabric) and the resulting large number of SKUs. This is
supported by many websites for M2M blinds featuring several filtering
options in prominent positions. However, and as already highlighted in
paragraph 8.230(d) above, we acknowledge that chat functionality may not
be a decisive factor for many customers.
In line with the above, the Parties’ customers do not appear to see eBay
and Amazon as close substitutes for the offering of the Parties. In
particular, the BDRC Survey shows that only 1% of respondents for
Blinds2Go and 3% of respondents for 247 would divert to eBay if the
respective Party was no longer selling blinds. For Amazon, the percentage
is 4% for both Blinds2Go and 247.
357
Similarly, comments from third parties indicate that online M2M blind
retailers do not tend to see selling through eBay or Amazon as a substitute
for selling through their own website, highlighting a number of limitations
with these marketplaces:
(a) MakeMyBlinds, an online M2M blind retailer that used to sell on eBay and
Amazon, told us that one of the reasons for ceasing to sell on these
platforms was their refunding policies. The same retailer told us that it
does not see Amazon or eBay as useful long-term engagements,
355
[].
356
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.90.
357
The higher diversion figures quoted by the Parties, namely 5% for Blinds2Go customers and 6% for 247
customers, arise because the Parties appear to be aggregating diversion to Amazon and eBay. Parties' response
to Provisional Findings, 7 August 2020, paragraph 4.91.
148
although it did consider eBay and Amazon useful to ‘get started’, to
initially generate some extra volume and to gain experience of the
industry. It also told us that one advantage of selling on eBay as a new
entrant to the market was that it was able to use eBay as a testbed, as
poor reviews wouldn’t affect its Trustpilot score.
(b) One online retailer told us that it tried to sell via Amazon, but that it was
not successful because customers tried to make an excuse not to pay.
358
It further told us that it would rather keep Amazon as only a small part of
its revenue to reduce the risk of dependency on another business and
that it currently sells only end of line products that it wants to get rid of on
Amazon.
Additionally, it told us that using platforms such as Amazon and
e-Bay means losing control over the brands,
and that it considers Amazon
as generally offering ‘cheap and cheerful’ products.
(c) Two retailers of M2M blinds (Blinds4UK and Dunelm) pointed to the high
costs of selling on Amazon and eBay. For example, Blinds4UK told us
that the marketplaces take a big cut from sellers and Dunelm told us that it
considers that Amazon charges a high commission rate. We acknowledge
the Parties’ submission that the commission charged by the platforms is
comparable to any equivalent PPC spending,
359
and give accordingly less
weight to these third-party comments. However, we also note that, as the
Parties pointed out, multi-channel retailers like Dunelm do not have to rely
on PPC advertising to generate traffic to their websites.
360
As such, the
cost of selling on Amazon and eBay may deter multi-channel retailers
from selling through these marketplaces.
We further note that the Parties do not appear to monitor the offerings on
eBay and Amazon. In particular, we are aware of only two instances where
Blinds2Go conducted some research on Amazon and eBay sales.
However, we note that these were in relation to the CMA’s investigation
into the Hillarys’ acquisition and the current CMA investigation.
Similarly, none of the other M2M blind retailers we spoke to indicated that
they are monitoring eBay or Amazon or that they consider marketplaces as
competitors:
(a) One competitor that works with Amazon and eBay considered that these
marketplace platforms mainly ‘compete with Dunelm, Argos and B&Q’.
361
358
[].
359
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.91.
360
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.35.
361
[].
149
(b) Three retailers of M2M blinds (Blinds4UK, Dunelm, and Swift Direct
Blinds) told us that they think only a very limited amount of M2M blinds
are sold on Amazon or eBay. In particular, Blinds4UK told us that the
platforms are not big players in the market and that sellers on Amazon
and eBay sell only ready-made products. Dunelm told us that it does not
believe that there are any M2M products available on Amazon. Swift
Direct Blinds told us that Amazon and e-Bay are active in this sector but
more for ready-made products.
Finally, while [], we have otherwise not received any evidence that would
suggest that the role of eBay and Amazon would significantly change going
forward. In particular:
(a) [] told us that it does not expect its sales on Amazon of M2M blinds to
grow beyond £[] within the next year.
362
(b) eBay told us that it expects sales in the blinds category to grow at around
[]% per year, in line with its other categories.
(c) Amazon told us that blinds make up a very small part of its product range
and so very little time is devoted to it. It therefore appears unlikely that
Amazon would have strong incentives to improve the offering of these
products.
Overall, we find that Amazon and eBay only exert a very limited competitive
constraint on the Parties. Additionally, we note that marketplaces would, in
any case, not act as separate competitors, but merely constitute platforms
for other retailers to sell online M2M blinds.
Out-of-market constraints
As set out in paragraph7.8, the Parties referred to the constraint from (i)
different window coverings including curtains and shutters, (ii) ready-made
blinds and (iii) in-store and in-home channels to M2M blinds supplied
through the online channel.
363
The Parties submitted that even if these are
not considered as part of the relevant market, it is highly relevant to
consider how such out-of-market factors exercise a competitive constraint
on their activities.
364
The Parties submitted that it is necessary to consider the aggregate
constraint of out-of-market constraints on the Parties and that it is the
362
[].
363
Main submission, 20 May 2020, paragraph 6.76 and 6.77.
364
Main submission, 20 May 2020, paragraph 6.65.
150
aggregate constraint from alternatives that constrains the Parties.
365
As
evidence of this aggregate constraint, the Parties submitted the following
points based on the results of the BDRC Survey:
(a) First, according to the BDRC Survey, 33% of Blinds2Go’s customers
and 25% of 247’s customers stated that they would not have
shopped for M2M blinds online in the event that their chosen online
retailer (i.e. either Blinds2Go or 247) was no longer selling M2M
blinds. The Parties submitted that this out-of-market diversion is
larger than the diversion to the next largest competitor.
366
(b) Second, according to the BDRC Survey, the diversion ratio from
Blinds2Go to 247 amounts to 13%, and is thereby much lower than
247’s market share would suggest. The Parties submitted that this
highlights the fact that market constraints are wider than just online
M2M blinds.
367
In addition, the Parties submitted that it is evident that, in a growing market,
the Parties are primarily competing for new business from ‘outside the
market’ rather than winning/losing business from each other, and that this
highlights the need for the CMA to properly consider the customer journey
and where these new sales are coming from.
368
As set out in the market definition section, we assessed the constraint from
each of the alternatives listed by the Parties individually. For the same
reasons as set out in that section, we find that:
a) curtains and shutters do not pose a material competitive constraint on
the Parties;
b) ready-made blinds pose a weak competitive constraint on the Parties,
and
c) M2M blinds sold through the in-store and in-home channel do not pose
a material competitive constraint on the Parties.
With respect to the aggregate constraint from out-of-market alternatives,
we acknowledge the Parties’ submission that the BDRC Survey indicates
that a significant proportion of the Parties’ customers (34% for Blind2Go
and 25% for 247) would not have bought online M2M blinds if the
respective Party had stopped selling blinds.
365
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.20.
366
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.18.
367
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.24.
368
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.15.
151
Whilst we acknowledge that this aggregate diversion is material, we also
note the following:
(a) The BDRC Survey shows that diversion to other retailers’ online M2M
blinds amounts to 66% for Blinds2Go’s customers and 75% for 247’s
customers. This indicates that online M2M blinds sold by other retailers
are the main competitive constraint on the Parties.
(b) The diversion to other products and channels is small when compared to
the relative size of sales of these alternatives. In particular, market reports
suggest that online M2M blinds account for less than 10% of the overall
market for window coverings.
369
Despite this, the results of the BDRC
Survey show that the Parties’ customers are substantially more likely to
switch to other retailers selling online M2M blinds than to the other
alternatives (ie other window coverings, ready-made blinds, M2M blinds in
the in-store or in-home channel). This suggests that these alternatives
would at best be a distant competitive constraint.
370
While we also acknowledge that diversion from Blinds2Go to 247 is lower
than 247’s market share would suggest, we note that this is consistent with
our finding that ‘out-of-market’ constraints pose some competitive
constraint on the Parties. Additionally, and as discussed above, we
consider that diversion to 247 is likely subject to a downward bias (see
paragraph
8.127).
Additionally, we note that the constraint from each of the individual out-of-
market constraints is weak (as evidenced by the low individual diversion).
Whilst we fully acknowledge that, in terms of sales lost in the event of a
price increase, diversion is as relevant whether it is an aggregated
diversion or diversion to a single competitor, it does not necessarily follow
that the impact of this aggregate constraint on the Parties’ behaviour will be
as strong as the impact from a single competitor. While the aggregate
constraint will by definition impact on the extent to which the Parties can
increase their prices (or otherwise deteriorate their offering) post-merger in
a static way, in terms of the dynamics of competition and the process of
rivalry, the individual constraint from single competitors is much more
relevant than the aggregate constraint from a disperse and diverse group of
small competitors: if a competitor improves their offering going forward, the
369
While we acknowledge that market reports may not be reliable, we find that they are likely to give a directional
indication of the relative size of online M2M blinds compared to other products and channels.
370
In particular, if ready-made blinds were as close a competitor as online M2M blinds, we would expect
diversion rates to be in line with the magnitude of sales, ie we would expect higher diversion to ready-made
blinds than to online M2M blinds.
152
Parties will have much stronger incentive to react by also improving their
offer if diversion to this competitor is large.
The fact that the Parties do not monitor any of these alternatives (or at least
not in any significant way, see paragraphs 8.138 to 8.141) indicates that
these alternatives, even in aggregate, do not exert a significant competitive
constraint on the Parties, in the sense that it does not appear that the
conduct of any of the retailers offering such alternatives would cause the
Parties to change their competitive behaviour. The Parties have also not
provided any evidence that they give strong consideration of these
constraints on an individual or aggregate basis.
With respect to the Parties’ submission that the Parties are primarily
competing for new business from ‘outside the market’, we note that this
may be caused by a shift in customer preferences away from alternatives
to online M2M blinds. The observation that the market is growing therefore
does not show whether these alternatives pose any constraint on online
M2M blind retailers.
Overall, whilst we recognise that the out-of-market constraints, in
aggregate, impose some degree of constraint on the Parties’ ability to raise
prices due to the aggregate diversion to these alternatives, we find that this
is likely to only exert a weak competitive constraint on the Parties.
Conclusion on remaining constraints
Overall, we conclude the following on the remaining constraints on the
Parties:
(a) Interior Goods Direct is a significant competitive constraint on the Parties,
and its constraint on Blinds2Go is similar to the constraint 247 poses on
Blinds2Go.
(b) Other online M2M retailers are small and only pose a relatively weak
constraint on the Parties, while the online offering of multi-channel
retailers and marketplaces exert very limited competitive constraints on
the Parties. Even in aggregate, we find that these constraints are not
particularly high.
(c) Out-of-market constraints, in aggregate, impose some degree of
constraint on the Parties’ ability to raise prices due to the aggregate
diversion to these alternatives, but we find that this is likely to only exert a
weak competitive constraint on the Parties.
153
Impact of the Merger
On the basis of our assessment set out above, we find that the Parties are
two of the three leading retailers of online M2M blinds and that the Parties
are close competitors. We find that there are few remaining constraints:
while Interior Goods Direct is a significant competitive constraint on the
Parties and its constraint on Blinds2Go is similar to the constraint 247
poses on Blind2Go, other suppliers of online M2M blinds as well as
alternatives from outside of the market at most pose a relatively weak
constraint on the Parties.
Against this background, we have assessed the impact of the 2019
Transaction as a result of the change in Hunter Douglas’ interest in 247,
resulting in Hunter Douglas holding a controlling interest over 247 and an
increase in its profit share in the company to 100%. We find that, following
the 2019 Transaction, Hunter Douglas has both the ability and the incentive
to increase the price (or otherwise worsen the offering) of both 247 and
Blinds2Go.
In assessing the impact of the 2019 Transaction, we note the following
preliminary observations when considering the ability and incentive of
Hunter Douglas to increase 247 and/or Blinds2Go’s prices:
(a) We note that, through the 2019 Transaction (ie through acquiring 100% of
the shares in 247), and as discussed in paragraphs 5.31 to 5.33, Hunter
Douglas gained the ability to unilaterally determine all aspects of 247’s
competitive strategy (including the ability to set 247’s prices) – which it did
not have prior to the 2019 Transaction.
371
(b) We acknowledge that a shareholding of 65% in Blinds2Go implies that
any incentive to increase 247’s price and capture diversion to Blinds2Go
is lower than under a 100% shareholding (ie that Hunter Douglas only
benefits from 65% of sales diverted to Blinds2Go). We also note that
Hunter Douglas has the option to acquire an additional 15% in 2021 and
the remaining 20% in 2026.
372
[].
373
(c) The evidence discussed above indicates that the Parties are close
competitors, with high rates of diversion from 247 to Blinds2Go at 34%
(with this diversion being likely subject to a downward bias, see
paragraphs 8.131 and 8.132). Therefore, in this context, Hunter Douglas’
371
Notwithstanding any material influence that Hunter Douglas may have held over 247 prior to the 2019
Transaction, we note that the ability to unilaterally determine 247’s competitive strategy and the benefit from an
increase in 247’s profits are a direct consequence of the 2019 Transaction.
372
Main submission, 20 May 2020, paragraph 6.10.
373
[].
154
65% shareholding in Blinds2Go (potentially rising to 80% in the near-
future) creates an incentive to increase 247’s prices as Hunter Douglas
still benefits from a significant share of sales that will likely be diverted to
Blinds2Go in the event of an increase. Hunter Douglas’ ability to increase
247’s prices results directly from the 2019 Transaction.
374
(d) We note that the 2019 Transaction increased Hunter Douglas’ share in
the profits of 247. Given that the Parties appear to be close competitors, it
follows that the increase in Hunter Douglas’ share of 247’s profits implies
an incentive for Hunter Douglas to increase Blinds2Go’s prices, as Hunter
Douglas now benefits from 100% of profits from sales diverted to 247.
While we acknowledge that the diversion from Blinds2Go to 247 found by
the BDRC Survey is, at 13%, not particularly high, we also note that this
diversion is likely subject to a downward bias, see paragraph 8.131.
375
Additionally, we have found that the Parties are two of the three leading
retailers of online M2M blinds and that the Parties are close competitors,
and that they will face limited remaining constraints post-merger. We
therefore consider that, combined with an increase of 51% in a party’s
interest in the profits of a target, this implies that Hunter Douglas will have
the incentive to increase Blinds2Go’s prices following the 2019
Transaction.
376
For these reasons, we find that, following the 2019 Transaction, Hunter
Douglas has both the ability and the incentive to increase 247’s prices. This
may also result in a reduction in product quality or customer service, and/or
a reduction in the range of products or services. The incentive to increase
247’s prices (or otherwise worsen the offering of 247) arises from (i) the
Parties being close competitors and (ii) Hunter Douglas benefitting from a
significant [] share in Blinds2Go.
In addition, we find that, following the 2019 Transaction, Hunter Douglas
has both the ability and the incentive to increase Blinds2Go’s prices. This
may also result in a reduction in product quality or customer service, and/or
a reduction in the range of products or services. The incentive to increase
Blinds2Go’s prices (or otherwise worsen the offering of Blings2Go) arises
from (i) the Parties being close competitors and (ii) Hunter Douglas, as a
consequence of the 2019 Transaction, benefitting from a 100% interest in
247, rather than only 49% pre-merger.
374
To the extent that Hunter Douglas had an incentive to increase 247’s prices prior to the 2019 Transaction, it
lacked the ability to do so (paragraph 5.30).
375
[]. Parties' response to Provisional Findings, 7 August 2020, paragraph 4.108.
376
[]. Parties' response to Provisional Findings, 7 August 2020, paragraph 4.105.
155
Parties submissions on the impact of the merger
The Parties submitted that Hunter Douglas has a strong interest in ensuring
that 247 maximises its own profits and maximises its own cash generation,
and that the nature of Hunter Douglas’ shareholding in Blinds2Go and the
accompanying management incentives make this more likely. In particular,
the Parties pointed out that Hunter Douglas only has a 65% shareholding in
Blinds2Go and that the agreed buy-out calculates the purchase price for
the remaining shares based on Blinds2Go’s profits in the three years prior
to the exercise of a call option to acquire the remaining shares. According
to the Parties, these characteristics imply the following:
(a) [].
377
(b) [].
378
[].
379
Additionally, the Parties submitted that the 2019 Transaction has reduced
Hunter Douglas’ incentives to worsen 247’s offer, pointing to the fact that
Hunter Douglas’ share of profits stood at 49% prior to the 2019
Transaction.
380
The Parties also submitted that, following the 2019 Transaction, 247 would
remain an independent competitor and that Hunter Douglas would continue
its ‘hands-off’ approach of not interfering with the business decisions of the
companies it owns.
Finally, the Parties submitted that the 2019 Transaction completed in
February 2019, around nine months before the CMA imposed the IEO. The
Parties argued that, if the increase in ownership of 247 by Hunter Douglas
results (or has resulted) in an SLC, one would expect to see evidence of
that (eg in the form of price changes) in the nine months of data following
the completed 2019 Transaction. The Parties submitted that, on the
contrary, there has been no discernible impact of the 2019 Transaction on
either Blinds2Go’s or 247’s margins. Similarly, the Parties submitted that
there is no evidence that Blinds2Go has worsened its offering at all
following the transfer of Web Blinds from Hillarys, with this rival having a
similar 5-10% share to 247.
381
377
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.117.
378
Parties' response to Provisional Findings, 7 August 2020, paragraphs 4.119 and 4.120.
379
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.116.
380
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.111.
381
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.106.
156
Our assessment of the Parties’ submissions
We consider each of the Parties arguments in turn.
The effect of Hunter Douglas’ 65% interest in Blinds2Go and buy-out formula
We have significant doubts about the Parties’ assessment of the impact of
the 2019 Transaction with respect to its current shareholding in Blinds2Go
and subsequent buy-out formula:
(a) First, the Parties’ submission only assesses the position in the short-term
and does not take into account the subsequent profits Hunter Douglas
would obtain if it has increased its shareholding in Blinds2Go (as
described above). [].
(b) Second, the calculations submitted by the Parties imply that any increase
to Blinds2Go’s profit would necessarily result in a decrease in Hunter
Douglas’ profits, through the impact it has on the price at which it would
then acquire the remaining Blinds2Go shares. In other words, the
calculations submitted by the Parties imply that Hunter Douglas would be
better off if Blinds2Go was less profitable. This would therefore imply that
Hunter Douglas would have an incentive to actively decrease Blinds2Go’s
profits. We disagree with this implication on the basis that it is inconsistent
with Blinds2Go having grown substantially in recent years, despite being
controlled by Hunter Douglas. It is also unclear why this would be of
commercial benefit to Hunter Douglas and it has not explained how it will
ultimately benefit from any such strategy.
(c) Finally, we note that, insofar as the buy-out formula dampens Hunter
Douglas’ incentive to divert profits from 247 to Blinds2Go, by the same
logic, this would also increase Hunter Douglas’ incentive to divert profits
from Blinds2Go to 247. Therefore,
[].
In response to our arguments regarding diverting profits from Blinds2Go,
the Parties submitted that this ‘completely disregards the whole purpose of
the incentive arrangements, which is to incentivise Blind2Go’s
management to pursue profitable growth’.
382
However, we note that Hunter
Douglas has a majority stake and thus a controlling interest in Blinds2Go.
Hunter Douglas can therefore control the composition of Blinds2Go’s board
and thus, indirectly, the behaviour of Blinds2Go’s management as well as
determining Blinds2Go’s competitive strategy. As such, in our view, it is the
382
Parties' response to Provisional Findings, 7 August 2020, paragraph 4.109.
157
incentives of Hunter Douglas, not the management of Blinds2Go, that will
determine Blinds2Go’s future behaviour.
Overall, and for the reasons discussed above, we do not find the Parties’
submissions in relation Hunter Douglas’ incentives as a result of its current
shareholding in Blinds2Go and the subsequent buy-out formula to be
persuasive.
While we acknowledge the Parties’ submission that it is [].
Impact of the 2019 Transaction on the incentive to increase 247’s prices
As noted above, []. However, prior to the 2019 Transaction, Hunter
Douglas did not have the ability to act upon any incentive to increase
247’s prices or otherwise worsen the offering of 247, as it was not able to
unilaterally determine the competitive strategy of 247. Hunter Douglas only
gained this ability as a direct consequence of the 2019 Transaction.
As such, it is not relevant whether Hunter Douglas’ incentive to increase
247’s prices (or otherwise worsen the offering of 247) decreased as a
consequence of the 2019 Transaction. Instead, it is relevant whether
Hunter Douglas has an incentive to increase 247’s prices (or otherwise
worsen the offering of 247) compared to the prices 247 would have set
unilaterally in the absence of the 2019 Transaction. As discussed in
paragraph
8.255 above, and due to the high rates of diversion from 247 to
Blinds2Go and Hunter Douglas’ 65% shareholding in Blinds2Go, we find
this to be the case.
Hunter Douglas’ management style
As set out above, the Parties submitted that 247 would remain an
independent competitor and that Hunter Douglas would continue its ‘hands-
off’ approach of not interfering with the business decisions of the
companies it owns. For the reasons discussed above, we consider that
following the 2019 Transaction Hunter Douglas has both the ability and
incentive to increase both 247 and Blinds2Go’s prices, and accordingly we
consider that it is not appropriate to give weight to the Parties’ statements
about their past chosen management style and how they might or might not
choose to adapt that in the future absent legal or other constraints.
Observed competitive outcomes following the 2019 Transaction and the Hunter
Douglas/Hillarys transaction in 2017
We now address the Parties’ submission that there has been no discernible
impact of the 2019 Transaction on either Blinds2Go’s or 247’s margins,
158
despite the transaction completing nine months before the CMA imposed
the IEO.
In our view, an absence of evidence of price/margin increases in the period
immediately following the completion of a transaction is not determinative
and does not imply a lack of ability or incentive to increase prices in the
future. Nor is it particularly compelling evidence that parties will not act on
any such incentive in the future. In addition, we find that there is a material
risk that this data is affected by the prospect of a CMA investigation. This
prospect would have militated against any incentive to increase prices until
the risk of an investigation had passed, or the 2019 Transaction had
received clearance from the CMA.
383
With respect to the Parties’ submission that there is no evidence that
Blinds2Go has worsened its offering following the transfer of Web Blinds
from Hillarys, we note that it is not necessarily possible to obtain insights on
the likely competitive effects of one merger by looking at a different merger.
In the case at hand, the competitive conditions in 2017 were different to
those observed now, including for example the fact that in 2019, Hunter
Douglas’ position was already strengthened by it owning Web Blinds. This
implies that insights from 2017 cannot necessarily be transferred to 2019.
Additionally, when assessing whether the Hunter Douglas / Hillarys
transaction resulted in Hunter Douglas worsening its offering, we only
examined prices, meaning that the analysis did not control for changes in
any other factors, including costs.
Conclusion on impact of the 2019 Transaction
In light of the above findings, we find that following the 2019 Transaction,
Hunter Douglas has both the ability and the incentive to raise both 247 and
Blinds2Go’s prices (or otherwise worsen the offering of 247 and Blinds2Go,
including through a reduction in product quality or customer service, and/or
a reduction in the range of products or services). We note that this holds in
the short-term, with the incentive to increase 247’s prices increasing over
time (should Hunter Douglas acquire an additional 15% and the remaining
20% in Blinds2Go).
Conclusion on competitive assessment
On the basis of our competitive assessment, we find that:
383
In this regard, we note Hunter Douglas’ submission that it has in relation to previous acquisitions taken
preparatory steps to mitigate the risk and effects of CMA intervention.
159
(a) The market shares we have calculated indicate that the Parties have a
high combined share of [60-70]%. Blinds2Go is by some distance the
leading supplier in this market with a share of [50-60]% and 247 is the
third largest at approximately [5-10]%. Interior Goods Direct is the only
other competitor of any meaningful scale, with a market share of [10-
15]%. The remainder of suppliers in the market have a limited presence,
with no other supplier holding a market share in excess of 5%. In the
context of this competitive landscape, the 2019 Transaction represents a
material increment to Blinds2Go’s existing scale and reduces the number
of established suppliers of scale.
(b) The Parties are close competitors and pose a significant competitive
constraint on each other;
(c) Interior Goods Direct is the only other significant constraint on the Parties,
but is not a closer competitor to either of the Parties than the Parties are
to each other;
(d) While there are other constraints (from smaller online M2M retailers, from
the online offering of multi-channel retailers, from marketplaces and from
out-of-market constraints), these alternatives pose a relatively weak
competitive constraint on the Parties; and
(e) Our assessment of the impact of the merger shows that Hunter Douglas
has acquired the ability to increase 247’s prices as a direct consequence
of the 2019 Transaction. We find that Hunter Douglas will have an
incentive to increase 247’s prices, as Hunter Douglas will benefit from a
significant share of sales that would likely be diverted to Blinds2Go in the
case of an increase through its 65% shareholding in Blinds2Go. At the
same time, we consider that, following the 2019 Transaction, Hunter
Douglas also has the ability and the incentive to increase Blinds2Go’s
prices. This may also result in a reduction in product quality or customer
service, and/or a reduction in the range of products or services.
We therefore find that the Merger may be expected to result in an SLC in
relation to the supply of online M2M blinds in the UK.
In the next section, we have considered whether there are any
countervailing factors (such as entry or expansion by other rivals) which
would be timely, likely and sufficient to outweigh the SLC.
Further, in paragraph 6.20 above, we have identified a number of possible
counterfactual scenarios to the 2019 Transaction. We note that our finding
of an SLC in relation to the supply of online M2M blinds in the UK (which is
160
subject to our views on countervailing factors), may be expected to result in
each of the counterfactual scenarios considered.
161
9. Countervailing factors
Our guidelines state that, in considering whether a merger may be expected
to result in an SLC, the CMA will consider factors that may mitigate the initial
effect of a merger on competition (often known as countervailing factors),
which in some cases may mean that there is no SLC. These factors include:
(a) the responses of others in the market (rivals, customers, potential new
entrants) to the merger, for instance the entry into the relevant market of
new providers or expansion by existing providers;
(b) the ability of customers to exercise countervailing buyer power; and
(c) the effect of any rivalry-enhancing efficiencies arising as a direct
consequence of the merger.
384
With respect to these countervailing factors, the focus of our inquiry has
been the assessment of the potential entry and/or expansion of rivals. The
Parties have not made any submissions suggesting that their customers
have significant buyer power or that there are any rivalry-enhancing
efficiencies. Accordingly, we do not consider these additional factors any
further in this section.
Entry and Expansion
CMA framework for assessing entry and expansion
As part of the assessment of the effect of a merger on competition, we look
at whether entry by new firms or expansion by existing firms may mitigate or
prevent an SLC.
385
We have considered whether such entry and expansion
would be likely to outweigh the SLC we have found in relation to the retail
supply of online M2M blinds. In assessing this, we have considered whether
entry and/or expansion would be timely, likely and sufficient.
386
We have considered the extent to which there are barriers to entry and
expansion in the retail supply of online M2M blinds. We discuss the
384
MAGs, para 5.8.1.
385
MAGs, para 5.8.1.
386
The timeliness of entry/expansion is assessed on a case-by-case basis, however the CMA would normally
consider entry or expansion within two years to be timely (MAGs 5.8.11). In assessing whether entry or
expansion may be likely, the CMA will consider the scale of any barriers to entry and/or expansion that may
impact on the likelihood of entry or expansion but also whether firms have the ability and incentive to enter the
market (or the intent to do so) (MAGs 5.8.8). Finally, with respect to sufficiency, the CMA will assess whether
entry or expansion is of sufficient scope so as to deter or defeat any attempt by the Parties to exploit any
lessening of competition (MAGs 5.8.10).
162
presence of such barriers first, before examining the possible sources of
entry and expansion from rivals.
387
The Parties’ views
The Parties submitted that there are no material barriers to entry or
expansion in the market for the retail supply of online M2M blinds. They
argued that ‘there are no material economies of scale that would constitute a
significant barrier to an entrant or smaller rival that made a determined effort
to increase its sales rapidly.’
388
It is the Parties’ view that the market is highly dynamic as demonstrated by
examples of successful recent entry both in the UK and other countries. The
Parties recognise that the market has matured since their own respective
entries, however, they submitted that prospective entrants could readily
replicate the Parties’ approach.
389
The Parties consider that UK manufacturers are likely to enter the retail
market either through launching their own retail operations or through
purchasing an existing retailer.
The Parties further submitted that evidence of entry and expansion should
be considered in aggregate when assessing if entry and expansion could
replicate the lost rivalry resulting from an incremental change in Hunter
Douglas’ influence over 247.
390
Our assessment of barriers to entry and expansion
We note that our SLC finding is based on Hunter Douglas acquiring the
ability to determine all aspects of 247’s competitive strategy as a result of
the 2019 Transaction (which it previously did not have). Therefore, our
assessment of entry and expansion is based on whether the loss of 247 as
an independent competitor can be replicated by competitors in the market.
Our assessment of barriers to entry and expansion will first consider what
potential barriers may exist with respect to generating traffic to retailers’
websites. We then consider what barriers may exist with respect to website
costs, before finally considering other potential barriers to entry and
expansion.
387
Barriers to entry and expansion are specific features of a market that give incumbent businesses advantages
over potential competitors. Where such barriers are low, the merged entity is more likely to be constrained by
entry. Conversely, this is less likely where barriers are high (MAGS 5.8.4).
388
Main submission, 20 May 2020, para. 7.4.
389
Main submission, 20 May 2020, para. 7.5 and 7.6.
390
Parties’ response to Provisional Findings, paragraph 5.3.
163
We also assess the extent to which a potential barrier to entry or expansion
may change depending on the nature of rival.
Generating traffic
As set out in paragraphs 8.15 to 8.30above, a significant aspect of
competition in this market is competition for visibility in web search results,
given the importance of traffic from this channel in generating revenues. As
also discussed above, we find that online M2M blind retailers primarily
generate traffic to their websites through paid search and organic search. As
such, this section focuses on assessing the extent to which there are
barriers to entry and expansion with respect to generating traffic through
these channels, although we also discuss other traffic sources where
appropriate.
391
Our assessment of the extent to which the generation of traffic may
constitute a barrier to entry and expansion in the retail supply of online M2M
blinds is set out in Appendix F. Our findings can be summarised as follows:
(a) Developing sufficient ability to convert traffic into sales constitutes a
barrier to entry and expansion, with low conversion rates implying that
retailers need to spend significantly more on marketing per conversion.
We find that effectively converting traffic relies on overcoming certain
knowledge barriers and that brand recognition, which tends to be more
prominent for established websites, is likely to improve conversions.
(b) There appears to be some incumbency advantage with respect to paid
search, with Blinds2Go []. In addition to this being driven by
Blinds2Go’s ability to [], we have identified the presence of knowledge
barriers and the role of brand recognition with respect to PPC
advertising as a barrier to entry and expansion.
(c) It appears that a significant investment is required, at least for smaller
retailers, to attract significant traffic through organic search and that the
return from this investment is not immediate. While we acknowledge that
it is in principle possible to circumvent such investment by entering into a
profit-sharing agreement with an SEO provider, we have seen no
evidence on the potential cost effectiveness of such an agreement for
the retailer.
391
We accept that entry in the literal sense does not require the generation of significant traffic (ie in the sense
that a retailer ‘entered’ as soon as this retailer made ‘a sale’). However, in our view, the generation of traffic is
required to enter with at least some scale (as well as for expansion). As such, we refer to ‘barriers to entry and
expansion’ when discussing the barriers associated with generating traffic.
164
(d) New entrants and smaller retailers have difficulty in achieving sufficient
visibility through Google, with the limited number of available positions
on Google search results that obtain significant proportions of clicks
constituting a natural barrier to entry.
We note that the findings set out above primarily apply to online M2M blind
retailers. The evidence received from multi-channel retailers indicates that
multi-channel retailers are less likely to consider the cost of digital
advertising as a barrier to entry and expansion, although some did cite this
as a barrier.
392
Additionally, it appears that multi-channel retailers tend to
have a broader approach to online advertising, ie their marketing efforts are
focused more on the entire home category as opposed to blinds
specifically,
393
and that they benefit at least to some extent from brand
recognition. As such, generating traffic appears to be less of a barrier to
entry and expansion for multi-channel retailers, and this is reflected in our
assessment of the likelihood of entry and expansion from multi-channel
retailers discussed below.
Overall, we find that generating traffic constitutes a barrier to entry and
expansion in the retail supply of online M2M blinds, in particular for online
M2M blind retailers. We note that the impact of this barrier may vary
depending on the nature of the rival seeking to enter or expand, with the
barrier being more significant for new entrants than for existing retailers.
Website costs
In this section we consider the extent to which the cost of establishing and
maintaining a website may be a barrier to entry or expansion in the online
M2M blinds market.
394
Website costs have been identified as the primary
capital cost incurred by entrants to the market, with other typical costs such
as office premises and distribution centres not required at a large scale or in
the initial stage of development.
395
The Parties submitted that website costs are not a barrier to entry or
expansion. They submitted that, like 247, new entrants are able to use third-
party providers (such as Magento or Shopify) to set-up websites in order to
392
For example, Dunelm told us it was unable to match the marketing spend of the Parties in relation to M2M
blinds because it has a wider range of products to promote using its marketing budget – although we consider it
unlikely that Dunelm would not be able to afford such spending (given its overall size).
393
For example, John Lewis mentioned this.
394
We note that multi-channel retailers typically do not have a standalone website for the supply of M2M blinds
online and that these services are incorporated into their wider websites. As such, the focus of this section is on
websites selling online M2M blinds and we address potential costs specific to multi-channel channel retailers
below.
395
We note that the manufacturers of blinds typically send the final product direct to customers without the need
to go via the retailer.
165
sell M2M blinds. However, Hunter Douglas acknowledged that replicating
the Blinds2Go website, for example, which was developed in-house over
many years, would result in additional costs.
396
The Parties further submitted that the costs for development and continuous
improvement of a website would account for only a small share of revenue
for successful retailers and that the costs of developing and hosting a
website of comparable quality to that of 247 are not prohibitive.
Blinds2Gosubmittedthat its total website development costs amounted to
£[] in 2019.
The Parties separately submitted that new entrants could also choose to
avoid website development costs altogether by selling through established
marketplaces such as Amazon, eBay or Wayfair.
397
They submitted that the
commission charged by Amazon does not place retailers who choose this
route at a competitive disadvantage to incumbents; the commission covers
various cost elements that incumbents who have their own websites need to
cover through their gross margin.
Our assessment of evidence provided by online M2M blind retailers
indicated that website set-up costs are a significant barrier to entry for new
entrants given the level of sophistication of the market leading websites that
supply M2M blinds. In summary, they told us that:
(a) Developing a fit-for-purpose website for a new entrant would cost at
least £100k. MakeMyBlinds entered the online M2M blinds market in
2015, they submitted that their website cost in excess of £[].
(b) CGI imagery of products would also result in meaningful costs on an
ongoing basis for new and existing retailers. 247 launched high-quality
CGI imagery in 2016 at a cost of £[] per annum. This implementation
alone led to an increase in sales of [] times ROI. We currently
consider that incumbent websites could also be subject to further
technical advantages. Decora submitted that visualisation software will
become a key part of the online offering by retailers, adding an element
of differentiation for online retailers. Decora also submitted that the
bigger online retailers will probably be the first to implement this. This
type of technology will play a part in the expansion for online M2M
retailers.
(c) For an online blinds company to compete effectively with leading rivals it
is essential to constantly invest in its website. MakeMyBlinds submitted
396
Main submission, 20 May 2020, paragraph 7.9.
397
Parties Main Submission, paragraph 7.9.
166
that in addition to its upfront website development costs, it incurs an
ongoing cost of several thousand pounds per month to improve its
website and search ranking. Some retailers emphasised the importance
of website quality and the customer experience.
The above indicates that even once a retailer has an online presence, it is
likely to incur ongoing costs to improve its website (which may in turn impact
search ranking, see discussion in Appendix F, paragraph 23), albeit,
depending on scope, this could be developed on an incremental basis.
398
In assessing the effectiveness of open-source platforms such as Magento
and Shopify for the sale of online M2M blinds, we note the Parties’
submissions that they may be a cost-effective alternative to a bespoke
website. However, we also note the Parties’ submission [].
399,400
In light of
this, in our view, obtaining a low cost open source solution will allow entry to
the online M2M market, however, in order to compete effectively with the
leading rivals, further investment will likely be required in order to optimise
these standardised platforms, as is done so by 247 who outsource their
website operations.
401
Online marketplaces
With respect to online marketplaces, some third parties trade through these
marketplaces in addition to their main sites, albeit on a smaller scale. The
Parties submitted that they expect that online marketplaces will represent a
growing proportion of the online M2M blinds offering in future. [].
[] submitted that it initially sold products on eBay and Amazon until 2017,
commenting that ‘[t]he reason for selling on eBay and Amazon was to get
generate volume for their suppliers and gain experience of the industry.’
However, [] stopped selling on external platforms at the end of that year,
and it does not see Amazon and eBay as useful long-term engagements.
We discuss the constraint posed by marketplaces in our Competitive
Assessment (paragraphs 8.228 to 8.240). Below we list several issues, as
identified by four third parties, with the use of online marketplaces as an
alternative way to sell M2M blinds.
398
An example of ongoing website develop costs is the use of A/B testing, a type of user research methodology
where two versions of a webpage (or app) are compared against each other to determine which works better.
[],
398
[].
399
Main submission, 20 May 2020, paragraph 6.17(v).
400
We note that the Parties submitted in their response to the Provisional Findings that [].
401
Main submission, 20 May 2020, paragraph 5.18.
167
(a) Restricted functionality: third parties noted that, in order to customise
their blinds, customers must message sellers separately through a
‘contact the seller’ function. For example, MakeMyBlinds offered M2M
products on eBay in certain sizes but then received requests to modify
the product. MakeMyBlinds described it as a slow and cumbersome way
to sell and Swift Direct Blinds described it as a ‘messy process’.
402
Amazon’s recently added ‘customise now’ function allows customers to
input dimensions prior to check-out; however, this isn’t a widely used
function. [] also submitted that Wayfair had ‘been considered as a
sales channel, but it is not good at offering M2M products.’
403
(b) One competitor also commented that online platforms can dilute brand
image and can be seen as ‘cheap and cheerful’, as a result, they
currently only sell end of line products that they want to get rid of on
Amazon.
404
(c) Fees and policies: whilst listing is free, two third parties highlighted the
commission taken from sales by the marketplaces. For example,
Blinds4UK told us that the marketplaces take a big cut from sellers and
Dunelm told us that it considers that Amazon charges a high
commission rate
.
The evidence we have seen indicates that there are low cost options to
enable small scale entry into the online market for M2M blinds, however,
website costs may constitute a barrier to entry for a retailer wishing to
compete on similar terms to the more established players in the market. In
particular, we note that the costs of establishing a new website and
associated features are not immaterial for online retailers. Further, whilst
new entrants may be able utilise open-source options to create and host a
simple website for a relatively low cost, the evidence provided by the Parties
and third parties indicates that a knowledge barrier exists in order to set-up a
credible website to compete effectively with the largest retailers in the
market.
However, we also note that for current more established rivals, or smaller
retailers who have already established high quality websites, the costs of
maintaining and improving their existing websites may not be particularly
high relative to their existing revenues, suggesting that website costs may
not be an equally significant barrier to further expansion of existing online
M2M blinds retailers.
402
[].
403
[].
404
[].
168
Finally, while marketplaces such as Amazon and eBay were identified as
introductory low-cost routes to enter the market, we are unconvinced they
would facilitate effective entry or expansion for a sizeable online M2M blinds
retailer for the reasons discussed above. We assess the evidence on
website costs to multi-channel retailers separately below.
Other barriers to entry/expansion
In this section we discuss the possibility of other barriers to entry or
expansion, including: (i) the role of an existing customer base and brand
awareness; and (ii) supplier relationships.
Role of an existing customer base and brand awareness
The Parties submitted that an existing customer base and brand awareness
play a limited role in the online M2M blinds market. They argued that the
estimate for the share of revenue from repeat customers overstates actual
long-term brand loyalty, []. The Parties further argued that [].
Additionally, the Parties submitted that brand awareness for 247 in the
general population is limited when compared to Blinds2Go and multi-channel
retailers such as Dunelm and Next. The Parties cited evidence from the
BDRC Survey, which shows that only 26% of customers (who bought from
247) mentioned the reputation of 247 as an important factor affecting their
choice of retailer, with six other factors being identified as important by more
customers and with only 9% identifying brand as the most important factor.
Despite M2M blinds not being a very frequent purchase, the data we
received from the main online M2M blind retailers shows that the proportion
of revenues from repeat customers is relatively high, ranking between 26%
and 42% in 2019. In addition, and as also noted in the Parties’ submissions,
[]% and []% of purchases are from customers who returned to
Blinds2Go and 247, respectively, more than one year after their first
purchase. On the basis of this data, we find that at least a proportion of
customers appear to exhibit loyalty towards the retailer from whom they
previously made a purchase.
169
We also consider that retailers with an established customer base benefit
from word-of-mouth recommendations
405
and the ability to engage in more
effective e-mail marketing campaigns.
406
In line with there being benefits from an established customer base, an email
from 247 to Hunter Douglas in 2018 notes that 247 is [].’
We agree with the Parties’ submission that multi-channel retailers are likely
to have a more significant repeat customer base than online M2M blind
retailers and that they enjoy brand recognition. However, it seems unlikely in
our view that the brand recognition of multi-channel retailers and
marketplaces has been created on the back of M2M blind sales. While the
Parties submitted that this does not matter,
407
we find that it is likely to play
at least some role: just because a customer bought for example a TV from
John Lewis, this does not necessarily mean that this customer is more likely
to buy M2M blinds from John Lewis than a customer that had not shopped at
John Lewis before. By the same token, although almost 90% of UK
shoppers use Amazon and 70% of these indicate that Amazon ‘is the first
online retailer they go to’ (as submitted by Parties, referencing a Mintel
report), this does not mean that customers will choose Amazon for their
online M2M blind purchases. Accordingly, while we acknowledge that the
general brand recognition of multi-channel retailers and marketplaces is
likely to benefit these retailers’ online M2M blind sales to at least some
extent, we find that the benefits are likely to be less pronounced than if the
brand recognition would have been created on the back of M2M blind sales.
In light of the above, in our view, there is some evidence that established
suppliers of online M2M blinds benefit both from an existing customer base
and brand awareness. Although this may not appear to be a strong barrier, it
does indicate that a new entrant may initially struggle to compete against
existing established suppliers who benefit from these factors. At the same
time, it may also be the case that existing suppliers (including multi-channel
retailers) are able to use their customer base and brand awareness to assist
in further expansion.
405
This channel of ‘advertising’ was specifically mentioned by Blinds2Go and the BDRC Survey shows that
‘recommendations’ are, after previous purchases, the second most important way in which the Parties’ customers
became aware of their respective brands. See comment made by Blinds2Go during CMA site visit, 24 April 2020.
406
While we acknowledge that email marketing is not a major source of traffic for any of the main online M2M
blind retailers in percentage terms, it still generated revenues of more than £[] and c.£ [], respectively, for
each of Blinds2Go and 247 in 2019 – and, as noted in the Parties’ submissions, at a very low cost.
407
Parties' response to Provisional Findings, 7 August 2020, paragraphs 5.32 and 5.70.
170
Supplier relationships
The Parties submitted that there are no direct costs associated with the
establishment of supplier relationships. They stated that there are many
suppliers who can deliver direct to the customer, with no shortage of logistics
providers to support the direct-to-customer model, concluding that both
existing players and new entrants can operate entirely on a ‘drop ship’ basis
whereby they need to hold no stock and require no warehousing.
408
However, certain third parties referred to the need for new entrants and
smaller retailers to establish relationships with new suppliers and noted that
this can be difficult. In particular, one multi-channel retailer told us that there
are only a limited number of suppliers in addition to Hunter Douglas,
restricting their ability to move to a different supplier.
409
In addition, []
submitted that some suppliers will not supply goods to new businesses due
to the relatively high failure rate of these businesses and have encountered
some difficulties in finding new suppliers due to suppliers not wanting to
disrupt their existing relationship with Hunter Douglas.
Swift Direct Blinds also submitted that ‘to operate at maximum efficiency
relative to the larger players in the market (obtaining the best deals on
materials, supplies etc) a turnover of approximately £20m would be
required.’
We find that, while there may be some difficulties in establishing
relationships with suppliers, the evidence we have received does not
suggest that this is a significant barrier to entry and expansion.
Conclusion on barriers to entry and expansion
We have found that there is evidence of some barriers to entry and
expansion in the retail supply of online M2M blinds. These barriers relate to
generating traffic, and to a lesser extent to website costs, brand awareness
and customer loyalty. However, we also note that the impact of any such
barriers may vary depending on the nature of the rival seeking to enter or
expand.
Whilst individual barriers may, in some circumstances, be overcome, their
cumulative effect could be significant. With respect to existing online M2M
blind retailers, we note that barriers to further expansion may not be as high
as for new entrants, however the Parties’ existing strengths in the market for
online M2M blinds (as discussed in the competitive assessment section)
408
Main submission, 20 May 2020, paragraph 7.13.
409
[].
171
mean that it is likely to be difficult for rivals to achieve sufficient expansion to
replace the loss of 247 as an independent rival. Indeed, the fact that there
has been little change in the identity of the leading suppliers in the market in
the past few years indicates that there is a degree of incumbency advantage
in the market that may constrain further expansion.
Possible sources of entry and expansion
In this section, we assess possible sources of entry or expansion. In line with
our Guidelines and past experience, this assessment considers the plans of
actual and potential rivals as to how entry or expansion may be achieved.
410
In order to outweigh the SLC that we have identified, we must be satisfied
that the evidence available to us clearly shows that entry or expansion from
these sources will be timely, likely and sufficient.
411
Speculative entry or
expansion plans from unspecified sources therefore will not meet this test.
Given that the online market for M2M blinds is a growing market and is
forecast to continue to grow, potentially accelerated by the current COVID-
19 pandemic, we would expect that there would be expansion from most
retailers as the market grows. An increase in turnover in a growing market
does not necessarily represent a growing competitive constraint that may
offset the SLC identified.
As part of our assessment, we have considered the possibility of actual entry
or expansion by: (i) online M2M blinds retailers; (ii) overseas retailers; (iii)
online marketplaces; (iv) multi-channel retailers; and (v) blinds
manufacturers.
The Parties’ views
The Parties cited several examples of recent entry into the market for online
retail supply of M2M blinds as evidence of their claim that the costs of entry
and expansion are limited. The Parties submitted that:
(a) Blinds Direct (Interior Goods Direct), which was established in 2004 and
relaunched its website in 2018, grew rapidly in recent years and
overtook 247 as the second largest online-only M2M blinds retailer.
(b) Swift Direct Blinds also registered significant growth since its launch in
2012. Additionally, the Parties submitted that the recent acquisition of
410
MAGs, paragraph 5.8.12.
411
As part of this assessment, and in accordance with our guidelines, we may consider entry or expansion within
less than two years as timely, but this is assessed on a case-by-case basis, depending on the characteristics and
dynamics of the market, as well as on the specific capabilities of potential entrants. – See MAGs, paragraph
5.8.11.
172
Swift Direct Blinds by Decora will only strengthen Swift Direct Blinds and
its ability to compete.
412
The Parties also submitted recent press reports
as regards Swift Direct Blinds’ revenue growth citing ‘Monthly ex VAT
revenue figures have increased from around £500,000 prior to lockdown
to over £1 million during and post.’
(c) MakeMyBlinds was founded in 2015 and has exhibited rapid growth
since then. It has a similar business model to 247.
413
The Parties stated that overall they ‘believe that expansion from existing
online M2M retailers alone is sufficient to replicate the rivalry that would be
lost by 247 – [].’
414
Additionally, the Parties referenced their own expansion in other European
countries, such as the Netherlands and Ireland, where they submitted that
they generated profits in a short period of time.
415
The Parties also argued that multi-channel retailers have a strong incentive
to expand their offering of online M2M blinds in order to ‘defend their existing
sales’ and expand their presence.
416
They consider that the COVID-19
pandemic will ‘turbo-charge’ multi-channel retailers’ expansion to online
sales, as the Parties believe the pandemic will lead to a change in customer
behaviour that will result in multi-channel retailers rapidly developing their
online offerings.
417
The Parties also asserted that manufacturers have a strong incentive to
enter the market for the online retail supply of M2M blinds. They referred to
Interior Goods Direct and Swift Direct Blinds as examples of vertically
integrated online retailers of M2M blinds. The Parties stated that they expect
other UK manufacturers to enter in the retail market in the near future, either
through launching their own retail operations or through purchasing an
existing retailer.
418
Our assessment
Online retailers
As explained in paragraph 8.10 above, the Parties comprise two of the three
largest online M2M blinds retailers. Interior Goods Direct is the only other
412
Parties response to Provisional Findings, paragraph 5.86(ii), Parties’ email to the CMA, 7 September 2020.
413
Parties response to Provisional Findings, paragraph 5.87.
414
Parties response to Provisional Findings, paragraph 5.87.
415
Main submission, 20 May 2020, paragraph 7.48. [].
416
Main submission, 20 May 2020, paragraph 7.68.
417
Main submission, 20 May 2020, paragraph 7.74.
418
Main submission, 20 May 2020, paragraphs 7.75-7.77.
173
online M2M blinds retailer of meaningful scale (with a similar share to 247),
with no other supplier holding an individual market share in excess of 5%.
Further, we also note that 247’s market share is approximately three times
bigger than the next largest supplier in the market.
• Interior Goods Direct
Interior Goods Direct was first launched 17 years ago under a different
brand. Blinds Direct, its current brand, is only 3 years old whereby Interior
Goods Direct was able to leverage traffic from the smaller websites it owned
e.g. Wooden Blinds Direct, Roller Blinds Direct, into the new Blinds Direct
website.
419
Interior Goods Direct recently acquired Wilsons Blinds, with the aim of
increasing its market share and turnover. While Interior Goods Direct did not
actively seek out this opportunity, it submitted that it is very difficult to gain
more market share in this sector other than through acquisition. Interior
Goods Direct stated that it intends to keep the two websites separate as it
sees that repeat customers, and thus customer loyalty, is a key component
in ongoing sales.
Interior Goods Direct submitted that it has not experienced growth that
significantly exceeds growth in the overall market. [].Therefore, we find
that despite not having certain or detailed plans for growth, Interior Goods
Direct continues to look for growth in an expanding market.
However, we also note that in order to address the SLC identified in
paragraph 8.276, the growth of Interior Goods Direct would have to replace
the constraint currently posed by 247 on Blinds2Go as the third largest
competitor in the market. Accordingly, as the second largest competitor,
Interior Goods Direct would have to grow to such an extent that it posed the
combined constraint of both Interior Goods Direct and 247 together on
Blinds2Go. Whilst we acknowledge that Interior Goods Direct may have
some high-level growth plans, in the absence of certain or detailed plans, we
consider the evidence does not demonstrate that any expansion by Interior
Goods Direct will be timely, likely and sufficient so as to offset the loss of
competition identified in our SLC finding.
• Swift Direct Blinds
Swift Direct Blinds was recently acquired by Decora, a leading M2M blinds
manufacturer. As a preliminary matter, we note that Swift Direct Blinds
419
[].
174
currently is a third of the size of 247 (by sales value). Therefore, in order to
replicate the constraint of 247, significant growth by Swift Direct Blinds would
be required (beyond that of the market) in a relatively short period of time
and evidence would have to demonstrate that growth of such scale is likely
to be achieved.
We discussed with Decora its rationale for purchasing Swift Direct Blinds
and its future plans. [].We understand from Swift Direct Blinds that it
experienced a period of growth from 2014 to 2016. [].Swift Direct Blinds
submitted that an increased number of competitors in the market and
difficulties in finding the right price point and positioning on Google has
meant that the growth rate of its competitors has outpaced its own.
Therefore, Swift Direct Blinds’ [] experience provide evidence of negative
performance even in the context of a fast-growing market.
Decora submitted [].
In order to verify this position, the CMA required the production of internal
documents by Decora in relation to its acquisition of Swift Direct Blinds and
future growth plans. In response, Decora explained that it did not possess
significant volumes of documents in relation to this acquisition as ‘Decora's
acquisition of Swift did not require scrutiny by any third party (such as a
lender) [].
Decora submitted a planning document prepared by an external third-party
consultant as part of its pre-acquisition planning which forecasts possible
future scenarios of sales revenue growth for Swift Direct Blinds. Decora
submitted that these forecasts were prepared when Decora was assessing
its e-commerce strategy. [].
However, Decora noted that it does not consider this document as a strategy
document and that business plans and strategies for the online market for
M2M blinds has not been developed yet due to the current business
circumstances it faces, such as purchasing Swift Direct Blinds and the
current COVID-19 pandemic. Moreover, the document submitted to the CMA
does not detail, even at a high-level, how such growth may be achieved.
Indeed, the documents submitted to the CMA noted that ‘[].’
420
As
explained in paragraph 9.13 above, and in Appendix F, there are a number
of factors which affect the efficacy of marketing spend; doubling marketing
spend does not necessarily result in a doubling of sales.
420
[].
175
As a result, Decora submitted that while the evidence provided to the CMA
may form the basis for devising a future strategy, it does not currently have
firm plans or sales growth projections for Swift Direct Blinds.
We accept that some degree of growth may well be likely for Swift Direct
Blinds given the growing market and increased marketing spend as a result
of the Decora acquisition. However, given the absence of any firm strategy
or growth plans, we consider the evidence does not demonstrate with
sufficient certainty that Swift Direct Blinds’ expansion following Decora’s
acquisition of Swift Direct Blinds would be timely, likely and sufficient to
offset the SLC we have found.
• MakeMyBlinds
MakeMyBlinds submitted that it aims to become the number one online
blinds company in the UK. We understand that it has seen significant growth
year on year,
421
and in 2020 is forecasting revenues of over £[] million.
MakeMyBlinds has experienced increased turnover month on month and
year on year as a result of the COVID-19 pandemic
422
and therefore, [].
MakeMyBlinds is currently running its first TV advertising campaign, which it
said is boosting sales and brand awareness.
423
MakeMyBlinds submitted
that TV advertising is important to prove legitimacy and adds weight to a
company’s proposition, noting that 247 and Blinds Direct utilised TV
advertising in the past.
424
[].
425
[].
426
[].
Notwithstanding its stated ambitions, the evidence available to us does not
demonstrate that MakeMyBlinds itself is likely to experience sufficient growth
within two years in order to replicate the constraint posed by 247 within the
online market for M2M blinds. In particular, we note the following:
(a) MakeMyBlinds is currently the seventh largest online M2M blind
retailer (with a share of [0-5]%) and revenues of £[]million in 2019.
This compares to 247’s existing share of [5-10]% and revenues of
approximately £[] million in 2019. Indeed, the company would
421
Its last years of revenue were £[] in 2018, £[] in 2019, 2020 is projected to generate £[].
422
[].
423
[].
424
[].
425
[].
426
[].
176
need to experience significant growth from a small starting point to
replicate the level of competitive constraint posed by 247.
(b) Moreover, any growth on the part of MakeMyBlinds would require
not just keeping pace with the market but also winning market share
from incumbents over and above this.
(c) Whilst some degree of expansion appears likely for MakeMyBlinds
over the next two years (given its [] and its current TV advertising
campaign) the evidence does not show that growth of sufficient
scale is likely to be achievable within this timeframe. In particular,
the CMA requested internal documents from MakeMyBlinds to
support its growth plans, however, [].’ [].Therefore, given the
low base from which MakeMyBlinds would have to grow
substantially, in the absence of any more detailed strategy or plans
to achieve growth, the CMA has no reliable basis on which it can be
confident that MakeMyBlinds is likely to grow to offset the
competitive constraint of 247 lost as a result of the 2019
Transaction.
(d) Additionally, we note that MakeMyBlinds will also be focussing its
efforts, and to some extent resources, on []. As a result, growth
experienced by MakeMyBlinds will not be limited to increasing its
market share of online M2M blinds but also in the adjacent markets.
However, we also note that building the ‘MakeMy’ brand across
markets could also develop MakeMyBlinds’ position.
Overall, in the absence of a reliable basis demonstrating that
MakeMyBlinds’ stated expansion plans are achievable within the next two
years, and given the low base from which MakeMyBlinds would have to
grow substantially within that period, we do not consider that the evidence
demonstrates that MakeMyBlinds’ expansion plans will be timely, likely and
sufficient so as to offset the SLC arising as a result of the 2019
Transaction.
• Other smaller online M2M blind retailers
In addition to the above, we observe that there is a long tail of online M2M
blinds retailers with a limited share of the market. However, we also note that
there has been limited growth by these retailers in recent years and they
remain very small. As we have seen only limited growth from these smaller
retailers on an individual basis, we do not consider it likely that they will
collectively account for a material aggregate competitive constraint on the
Parties going forward.
177
• Aggregate constraint
Whilst we have found that, for each specialist online M2M blind retailer
identified above, their expansion would not be timely, likely and sufficient to
offset the SLC on an individual basis, we have also considered the
aggregate position. We note, in this regard, that some degree of expansion
may be timely and likely for each of Interior Goods Direct, Swift Direct Blinds
and MakeMyBlinds (based on their high level plans) and if both Swift Direct
Blinds and MakeMyBlinds were to achieve their aspirational high level
growth plans, this could theoretically be sufficient to replicate the constraint
exerted by 247 prior to the 2019 Transaction. However, each of these
retailers lacks detailed and firm plans as to how such growth would be
achieved and, accordingly, we do not consider any of them likely to be fully
achieved in practice. Moreover, given the significant growth that would be
needed across all three of these retailers to be able to replicate the
constraint from 247, we consider that only a small proportion of such growth
is likely to be achieved in a timely manner. Accordingly, even on an
aggregate basis we consider that the expansion of Interior Goods Direct,
Swift Blinds Direct and MakeMyBlinds would not be timely, likely and
sufficient to offset the SLC we have found.
Overseas retailers
The Parties submit that there is evidence of successful entry in other
European countries and cite their own experience expanding into Europe
as evidence. However, we do not have any evidence of overseas retailers
looking to expand into the UK market.
In relation to an overseas retailer as a potential purchaser of 247, Hunter
Douglas submitted limitations on the success of an overseas retailer
entering the UK online M2M blinds market, citing ‘they would have to be
interested in coming into the UK, knowledgeable about the UK: []. Each
market is just different: the taste is different, the structure is different. It
would be a challenge for someone from outside the UK coming into the
UK.’ Therefore, even in the event that an overseas retailer were to enter
the UK market for online M2M blinds, we have no reason to believe that
they would grow to the extent to pose a constraint on the Merged Entity.
Online marketplaces
With respect to marketplaces, Amazon’s and eBay’s current sales of online
M2M blinds are low and both retailers submitted that they had no plans to
grow their respective offering of M2M blinds.
178
We note that as of the beginning of 2020 Amazon has introduced new
functionality such as its ‘customise now’ function, and as such the platform
may increasingly be used for M2M blinds. However, since, other than [],
we have not received any evidence that would suggest that the role of
Amazon would significantly change (see paragraph 8.239), we have no
reason to believe that Amazon’s sales of online M2M blinds will
substantially increase going forward.
Multi-channel retailers
As highlighted in the Industry Background section, multi-channel retailers
offer M2M blinds online to varying extents, with not all retailers offering the
same click-to-order functionality as the Parties. We recognise that multi-
channel retailers may not face the same barriers to entry or expansion as
online-only retailers and may also benefit from greater financial resources
and brand recognition. However, the online M2M blind sales of multi-channel
retailers are currently low, implying that any expansion of these retailers
starts from a low base. We have also not received any evidence to suggest
that multi-channel retailers are using their brand recognition to grow their
share of the online M2M blinds market.
Additionally, multi-channel retailers face different challenges with the online
retail supply of M2M blinds. In particular, all of the main multi-channel
retailers identified by the Parties as competitive constraints have
emphasised the complexity of adding and integrating website functionality
for M2M blinds with their existing websites (where they sell other products
and services not related to the window coverings sector). They noted that,
as a result, new functionality tends to be added as part of a wider
overhaul/relaunch of a retailer’s website and therefore the timescales are
much longer and costs much higher. This suggests that reputational and
operational risks are a greater consideration for multi-channel retailers than
other sources of entry and expansion.
Moreover, multi-channel retailers have emphasised that the current COVID-
19 pandemic has impacted them to a significant extent. Given that their in-
store or in-home offering of M2M blinds has been halted in recent times, this
has in turn impacted their ability to expand further. This also indicates that
multi-channel retailers may have competing priorities in their business ahead
of online M2M blinds.
The Parties, however, refute the stated impact of COVID 19 on the multi-
channel retailers. The Parties submitted that:
179
(a) Dunelm has seen year-on-year increases of over 100% and that it
will invest in its digital capabilities.
(b) John Lewis stated that ‘[b]efore the virus struck, 40 percent of John
Lewis sales were online. This could now be closer to 60 to 70
percent of total sales this year and next’. The Parties further
submitted that John Lewis is also conducting a digital transformation
and significantly expanding its click & collect service.
(c) Next has returned its next day delivery service and rumoured to
launch an online platform that will allow other brands to sell products
through its website.
427
We assess below the evidence submitted to us by each retailer in relation
to any plans to expand or enter the market, including how these plans may
be impacted by the current COVID-19 pandemic.
• John Lewis
John Lewis told us that it offers a variety of blinds online with click-to-order
functionality, although it currently has only []% of its M2M products online.
Indeed, many of the blinds advertised are listed as in-store only. We
understand that there has been a drive to increase this for the last [],
however John Lewis said that ‘it would need significant investment in its
online systems and website update to achieve this.’
428
[]. We note that
these costs are far in excess of those submitted by other parties, including
other multi-channel retailers. John Lewis submits that these costs are
specific to John Lewis and relate to setting up the required functionality and
do not include the cost of making the products available online.
429
[]’
430
John Lewis noted further that, while its online businesses and
factories for M2M blinds were closed due to the COVID-19 pandemic, many
online competitors remained open for business.
431
John Lewis submitted that
while online sales of M2M blinds would restart once the COVID-19 pandemic
guidance changed, absent its website upgrade project, John Lewis does not
expect any growth and sales are likely to go down due to its limited online
offering of M2M blinds.
432
427
Parties’ response to Provisional Findings, paragraph 5.90.
428
[].
429
[].
430
[].
431
[].
432
[].
180
The John Lewis Partnership is currently undergoing a business-wide
strategic review of its digital offering, as referenced by the Parties (see
paragraph 9.79(b)),
433
with the aim of improving general performance within
3 to 5 years.
434
However, this review is not specific to online M2M blinds. In
terms of how this could impact blinds, John Lewis submitted that [].
Therefore, notwithstanding the impact of COVID-19 on John Lewis’ sales of
online M2M blinds, the evidence submitted to us by John Lewis []. As a
result, we are not satisfied that any further expansion by John Lewis in the
online M2M blinds market would be timely or likely. [].
• Next
Next integrated its click-to-order functionality for online M2M blinds with its
main website in Autumn 2019. Next explained that it changed supplier for
M2M blinds in 2019 and brought its entire M2M blinds offering in house,
enabling more control of what products it was able to sell.
Next told us that the current COVID-19 situation had initially prevented both
ready-made and M2M blinds being sold both in-store and online due to
Next’s two-man delivery service. However, both products can now be
purchased online once again. [], however, this is currently on hold due to
the team responsible for this currently being unavailable, although Next
expects that this will continue once the COVID-19 pandemic subsides. Next
reconfirmed that the current situation still remains and that there has been
no change in their plans.
Crucially, however, Next believes its share of the market is small compared
to other M2M blind retailers and is not forecasting an increase in sales from
2019 to 2020 on a like for like basis. As confirmed in our market share
estimates, Next’s share of the online M2M blinds market is modest at [0-5]%
and significantly smaller than the Parties. Therefore, given Next’s position in
the market and the fact that it has no significant plans to expand its online
retail offering of M2M blinds, we do not consider that any further expansion
by Next would be timely or likely. Given Next’s limited share of the market
we also consider that even if further expansion by Next was timely and likely,
substantial growth would be needed in order to be timely, likely and sufficient
to offset the SLC we have found.
433
Parties response to Provisional findings, paragraph 5.90(ii).
434
https://www.johnlewispartnership.co.uk/media/press/y2020/progress-update-on-the-strategic-review.html
181
• Dunelm
Dunelm does not currently offer click-to-order M2M blinds online. As
described in paragraph 4.34, Dunelm currently offers M2M blinds online
through virtual consultations. However, for the reasons described in that
section, we do not consider that Dunelm’s current offering is comparable to
that of the Parties or their rivals in the online M2M blinds market.
Dunelm previously had a platform offering M2M blinds online, however this
was removed as a part of a major update of its website in Autumn 2019
requiring development work. [].
435
Dunelm also estimated that the third-
party cost of developing the online blinds element was in the region of
£1.5m, which does not include internal costs or time’.
436
We therefore
consider that the costs of Dunelm integrating online M2M blinds into its web
platform are considerable.
[]. A date to implement M2M online functionality has not been set.
437
The Parties have indicated that they expect re-entry by Dunelm to be timely
and significant, stating their view that Dunelm ‘[].’ In response to this, we
note that Dunelm submitted that sales of online M2M blinds were £[]in
2019, compared to in-store sales of £[].
The Parties further submit that Dunelm made the following statements: ‘We
continue to invest in our digital capabilities […]. We will also be investing in
supply chain capacity to meet the high growth ambition for our home delivery
channels.’ Additionally, the Parties have submitted recent press reports
citing Dunelm’s positive performance as a result of the COVID-19 pandemic.
However, we note that the referenced growth and investment is not limited to
Dunelm’s online offering of M2M blinds [], additionally that its reported
growth also relates to sales earned in-store across all products.
Whilst we note that []. Therefore, we do not consider that re-entry by
Dunelm would be timely or likely. Moreover, Dunelm’s online sales of M2M
blinds were low prior to its exit in 2019 and significantly below those of the
Parties. We therefore consider that any re-entry by Dunelm would need to
be likely to achieve significant scale within a short timeframe in order to be
sufficient to offset the SLC we have found.
435
[].
436
[].
437
[].
182
• IKEA
IKEA has been identified as a current competitor by the Parties and also
described as an entrant, or re-entrant, to the blinds market by the Parties,
and a ‘serious competitor’. However, IKEA has confirmed to us that it has no
plans to enter the online M2M blinds market.
Manufacturers
The Parties identified Decora as a recent entrant to the online M2M blinds
retail market. In support of this, the Parties noted Decora’s recent acquisition
of Swift Direct Blinds, as well as its investment in the Netherlands (via the
acquisition of Coolblinds). The Parties also cite that ‘Decora and its
investment partners plan to ‘transform a family owned business into a
European leader’.’
438
As a manufacturer and distributer for M2M blinds, Decora submitted that it is
interested in entering the online retail market for M2M blinds, but only
through acquisition of an existing online platform. This is consistent with
evidence available to the CMA, [], as well as the successful acquisition of
Swift Direct Blinds described above. The CMA also understands that Decora
has held, and will continue to hold, informal discussions with other potential
targets. However, these discussions have not resulted in any opportunity
that Decora has considered commercially viable, or, only initial inquiries
have been made. In response to the CMA’s questions related to Decora’s
potential acquisition strategy, Decora submitted that it has no current plans
to acquire any additional companies active in the online M2M blinds market
in the UK (with the exception of 247). Decora also submitted that it has no
internal documents discussing or assessing a future acquisition strategy
and/or potential targets.
439
Therefore, whilst Decora may be interested in
acquiring companies active in the market in the future, in the absence of any
current plans to acquire companies, the CMA does not consider that future
expansion by Decora will be timely, likely and sufficient to offset the SLC we
have found.
Separately, we also note Decora submitted it would take at least five to
seven years for a new company to build any significant market share and
such new company will likely take losses in five to seven years of
operation.
440
438
Quote from Don Harrington (director of Goodbody Corporate Finance, who advises Decora’s parent company)
from an article in the Irish News dated 8 August 2020.
439
[].
440
[].
183
Conclusion on sources of entry and expansion
Whilst we understand that certain competitors do have plans to grow, their
growth plans would have to considerably outperform an already fast-growing
market in order to provide a sufficient constraint to mitigate the effects of the
Merger between the first and third largest retailers. The evidence available to
us does not reliably indicate how these growth plans would be achieved so
as to result in these competitors being a significantly increased individual or
aggregate constraint on the Parties post-Merger.
The evidence above shows that expansion from individual existing online
retailers of M2M blinds will not be timely, likely and sufficient to offset the
SLC we have found. In particular, whilst we acknowledge that Interior Goods
Direct (being an online retailer of comparable scale to 247) may have some
high-level growth plans, it does not have certain or detailed plans on how
such growth will be achieved, and therefore we do not consider its
expansion would be timely, likely and sufficient to offset the loss of 247 as
an independent competitor. []. Whilst MakeMyBlinds has told us that it
also intends to expand further, given it is currently the [] largest online
M2M retailer, the evidence available to us does not demonstrate that it is
likely that it will achieve sufficient scale in a timely manner in order to be a
sufficient competitive constraint on the Parties in the near future given the
significant level of growth required from a small starting base.
The evidence from multi-channel retailers does not demonstrate that any
expansion (in the case of John Lewis and Next) or re-entry (Dunelm) will be
timely or likely. Further, the fact that multi-channel retailers have previously
been able to achieve only a limited presence in the online M2M blinds
market means that, even if expansion or re-entry were timely and likely, we
do not consider that it would be timely, likely and sufficient to offset the SLC
we have found. In addition, whilst they may have expressed a previous
interest in developing a presence in this market, all of these retailers have
indicated to us that these plans have been significantly impeded by the
current COVID-19 pandemic. Current plans for future growth through the
online channel for multi-channel retailers encompass their entire online
product offering, of which M2M blinds comprise a small part. When
questioned, multi-channel retailers confirmed that there were no specific
plans, or unlikely to be, for M2M blinds. Moreover, the significant
development costs and lead times incurred by multi-channel retailers
indicates that they face specific challenges in developing a market presence
that is consistent with their wider reputational and operational objectives. In
this respect, it is notable that Dunelm has exited the market previously
having entered in 2018. Dunelm’s current online M2M blinds offering is still
some way off the functionality offered by the online M2M blind retailers,
184
given that it does not offer click-to-order M2M blinds. Similarly, John Lewis
told us that development in this area is significantly more costly than other
retailers and reflects its company-specific considerations. We also note that
the evidence available from all of these multi-channel retailers indicates that
they have been unable to develop a market presence comparable to that of
the Parties. Accordingly, our view is that any expansion or entry by multi-
channel retailers would not, individually, be timely, likely and sufficient to
prevent an SLC from arising.
Finally, we note that the evidence from manufacturers does not demonstrate
that entry or expansion would be timely, likely and sufficient. In particular,
whilst we note Decora’s entry into this market through acquisition, the
evidence available to us does not indicate that it has any current plans to
acquire future targets. Moreover, additional evidence indicates to us that de
novo entry by a manufacturer would be risky, costly and with no guarantee
of success.
Aggregate constraint
The CMA also has considered the extent to which these potential sources
of entry and expansion may pose a competitive constraint on the Parties
when considered in aggregate, including in the relevant market and across
channels. However, we note that in each instance in which potential entry
or expansion has been assessed, we have not obtained sufficient evidence
to give us confidence that the requisite growth will be achieved. We also
note that, except for Interior Goods Direct, each potential rival is of
significantly smaller scale than 247. Therefore, any aggregate constraint
on the Parties would require several rivals to achieve significant growth
beyond both their previous performance and any current growth in the
market. In the absence of evidence showing that rivals are likely to achieve
this individually, we do not consider it likely that this constraint will be
achieved on an aggregate basis.
Conclusions
Based on the evidence as set out above we have found there to be evidence
of barriers to entry and expansion in the supply of online M2M blinds. Whilst
individually, these barriers are not insurmountable, the cumulative effect of
these barriers could be significant for new entrants. We have also found that
barriers to entry may be higher for entry at scale than barriers to expansion
for existing rivals. However, the fact that there has been little change in the
identity of the leading suppliers in the market in the past few years suggests
185
that there is a degree of incumbency advantage in the market that may
constrain expansion by existing rivals.
Notwithstanding this finding, we have further found there to be limited
evidence of actual and/or planned entry or expansion from third parties in
this market. This reflects the following:
(a) Some online retailers may have plans to expand, however, the evidence
available to us does not reliably indicate how these growth plans would
be achieved so as to result in these competitors being a significantly
increased individual or aggregate constraint on the Parties post-Merger.
Furthermore, we observe limited growth from the tail of smaller retailers
in recent years. Plans for expansion, for both Swift Direct Blinds and
Make My Blinds, have, in our view, been more of a statement of intent,
as opposed to developed or detailed plans.
(b) Whilst MakeMyBlinds has grown in recent years, as noted at paragraph
9.64 its current market position indicates that it is unlikely to expand
sufficiently to offer a significant competitive constraint on the Parties in
the near future and we do not have sufficient evidence as to how this
company may achieve its stated growth plans.
(c) While manufacturer Decora has entered the retail market for M2M blinds
via Swift Direct Blinds, evidence received as part of our investigation
indicates that it is unlikely to represent a sufficient competitive constraint
on the Parties []. In addition, even if Swift Direct Blinds’ growth
significantly outpaced that of the market, it would still not be comparable
in size to 247 for several years.
(d) The evidence received from multi-channel retailers also does not
demonstrate that any expansion or re-entry into the market will be timely
or likely as they deal with the consequences of the COVID-19 pandemic.
(e) Where there has been entry, or increased presence from multi-channel
retailers, we note they have achieved only a limited market share and
sales as compared to the Parties. Where multi-channel retailers have
stated their ambitions to expand their online offering for customers, we
have no evidence that this is relevant to M2M blinds. Therefore, even if
entry or expansion from these retailers was timely and likely (which we
do not consider to be the case), the evidence currently available to us
does not allow us to conclude that this would be sufficient (either
individually or in aggregate) to constrain the Merged Entity.
186
In summary, the evidence available to us does not demonstrate that entry or
expansion will be timely, likely and sufficient, either on an individual or
aggregated basis.
In light of the foregoing, we conclude that the evidence does not support the
view that timely, likely and sufficient entry or expansion will outweigh the
SLC we have identified.
10. The decision
We have found that the 2013 Transaction has not resulted in the creation of
a relevant merger situation.
We have found that the 2019 Transaction has resulted in the creation of a
relevant merger situation.
We have concluded that the 2019 Transaction may be expected to result in
an SLC as a result of horizontal unilateral effects in the online retail supply
of M2M blinds in the UK.
We conclude that the adverse effect arising from the identified SLC would
be that there would be the ability and incentive for the Merged Entity to
increase retail prices, lower the quality of its products or customer service,
and/or reduce the range of its products/services.
187
11. Remedies
Introduction
We have found that the 2019 Transaction may be expected to result in a
SLC.
Where the CMA finds an SLC in its final report, it must decide what, if any,
action should be taken to remedy, mitigate or prevent that SLC or any
adverse effect resulting from the SLC.
441
This chapter considers possible remedies to the SLC that we have
identified in this report and contains our decision on remedies. In reaching
our final decision on the appropriate remedy, we have considered evidence
from the Parties and third parties, including:
(a) Responses to our notice of possible remedies (Remedies Notice);
442
(b) Response hearings with both Hunter Douglas and 247;
(c) Questionnaires and calls with third parties; and
(d) The Parties’ response to our Remedies Working Paper, which set out
our provisional decision on remedies.
443
This chapter sets out:
(a) The CMA’s framework for assessing remedies;
(b) An overview of the remedy options;
(c) Our assessment of the effectiveness of each remedy option;
(d) Our assessment of the proportionality of each remedy option,
including our assessment of any associated Relevant Customer
Benefits;
(e) Our assessment of remedy implementation for each remedy option;
and
441
Enterprise Act 2002, section 35(3).
442
Remedies Notice
443
The Parties and Decora submitted responses to our Remedies Notice. We held hearings with Hunter Douglas
and 247 and calls with four third parties. We also sent a questionnaire to 20 third parties and received two
responses.
188
(f) Our final decision on remedies.
CMA framework for assessing remedies
The Enterprise Act 2002 (the ‘Act’) requires that the CMA, when
considering possible remedial actions, shall ‘in particular, have regard to
the need to achieve as comprehensive a solution as is reasonable and
practicable to the substantial lessening of competition and any adverse
effects resulting from it’.
444
To fulfil this requirement, the CMA will seek remedies that are effective in
addressing the SLC and any resulting adverse effects. The effectiveness of
a remedy is assessed by reference to its:
(a) impact on the SLC and the resulting adverse effects;
(b) duration and timing – remedies need to be capable of timely
implementation and address the SLC effectively throughout its
expected duration;
(c) practicality in terms of implementation and any subsequent
monitoring; and
(d) risk profile, relating in particular to the risk that the remedy will not
achieve its intended effect.
445
The CMA will then select the least costly and intrusive remedy that it
considers to be effective. The CMA will seek to ensure that no remedy is
disproportionate in relation to the SLC and its adverse effects. The CMA
may also have regard, in accordance with the Act,
446
to the effect of any
remedial action on any relevant customer benefits arising from the merger.
Overview of remedy options
As set out in the CMA guidance, remedies are conventionally classified as
either structural or behavioural.
447
In merger inquiries, the CMA generally
prefers structural remedies over behavioural remedies because:
448
444
Enterprise Act 2002, section 35(4).
445
Merger remedies guidance CMA87, paragraph 3.5
446
Enterprise Act 2002, section 35(5).
447
Merger remedies guidance CMA87, paragraph 3.34. Some remedies, such as those relating to access to IP
rights may have features of structural or behavioural remedies depending on their particular formulation.
448
Merger remedies guidance CMA87, paragraph 3.46.
189
(a) structural remedies are more likely to deal with an SLC and its
resulting adverse effects directly and comprehensively at source by
restoring rivalry;
(b) behavioural remedies are less likely to have an effective impact on the
SLC and its resulting adverse effects, and are more likely to create
significant costly distortions in market outcomes; and
(c) structural remedies rarely require monitoring and enforcement once
implemented.
In this chapter, we set out the following remedies options:
(a) Requiring the full divestiture of 100% of the ordinary share capital of
247; and
(b) Requiring a partial divestiture of 51% of the ordinary share capital of
247.
In the Remedies Notice, we invited views on aspects of remedy design
which might be needed to make a divestiture remedy effective and to
ensure that no new competition concerns would arise. These may include
requirements relating to the scope of any divestiture package, the process
of selecting the assets to be divested, the identification of suitable potential
purchaser(s), and the divestiture process including the timing of divestiture.
In the Remedies Notice, we also said that behavioural remedies were
unlikely to be effective in addressing the SLC that we had found. We said
that we were willing to consider any behavioural remedies that were put
forward as part of the consultation, but none were proposed. We therefore
did not consider behavioural remedies further.
A successful divestiture will effectively address at source the loss of rivalry
resulting from the merger by changing or restoring the structure of the
market.
449
There are three categories of risk that could impair the effectiveness of any
divestiture remedy; composition risk, purchaser risk and asset risk:
450
(a) composition risk arises if the scope of the divestiture package is too
narrowly constrained or not appropriately configured to attract a
449
Merger remedies guidance CMA87, paragraph 3.38.
450
Merger remedies guidance CMA87, paragraph 5.3.
190
suitable purchaser, or does not allow a purchaser to operate as an
effective competitor;
(b) purchaser risk arises if a divestiture is made to a weak or otherwise
inappropriate purchaser, or if a suitable purchaser is not available;
and
(c) asset risk arises if the competitive capability of the divestiture package
deteriorates before completion of the divestiture.
An effective divestiture remedy must give us sufficient confidence that
these practical risks can be properly addressed in its design. We therefore
consider the following design issues:
(a) the appropriate scope of the divestiture package;
(b) the identification and availability of suitable purchasers; and
(c) ensuring an effective divestiture process
We consider each of these in turn below, where we set out the views of the
Parties and third parties on each aspect of the remedy design. To be
effective in remedying the SLC, any divestiture package would need to be
appropriately configured to be attractive to potential purchasers and to
enable the purchaser to operate effectively as an independent competitor.
Independence is a key consideration in the remedy option where Hunter
Douglas retains a minority stake in 247.
In the following sections, for each remedy option we set out:
(a) A brief description of the proposed remedy;
(b) The general views from the Parties and third parties on the
effectiveness of the remedy; and
(c) Our assessment of the effectiveness of each remedy.
Full divestiture of 247
Description of remedy
Under a full divestiture remedy, Hunter Douglas would be required to divest
100% of 247’s ordinary share capital within a timeframe specified by the
CMA.
191
Remedy design issues
Scope of divestiture package
In considering the appropriate scope for a divestiture package, we should
ensure that it:
(a) is sufficiently broad in scope to address all aspects of the SLC and
resulting adverse effects;
(b) would enable the eventual purchaser to operate the divested business
as an effective competitor; and
(c) is sufficiently attractive to potential purchasers.
In the Remedies Notice, the Remedies Hearings and on calls with third
parties, we sought views on the scope of the divestiture package, including
the level of shareholding that Hunter Douglas should be required to divest
and whether 247’s international operations should be included in the
divestiture package.
Parties’ and third-party views on the scope of the divestiture package
In their response to the Remedies Notice, the Parties told us that full
divestiture would be ‘ultra vires’ and ‘disproportionate to the SLC identified
in the Provisional Conclusions’ however that a partial ‘divestment of 51% of
the ordinary share capital of 247 would be linked to the SLC… and could
be proportionate.’
451
The Parties submitted that there was no basis for anything other than
restoring the pre-merger ownership structure of 247.
452
The Parties also
submitted that ‘the aim of the remedy is not to create a situation in which
247 operates in a wholly independent [original emphasis] manner to Hunter
Douglas’
453
which would be the case in the event of a full divestiture of 247.
In the remedies hearing, Mr Peterkin, one of the Founding Shareholders of
247 who continues to run 247, commented that a full divestiture of 247 from
Hunter Douglas ‘would certainly remedy the situation, depending on who
bought it. Whether or not it would be effective in leaving 247 competitive is
obviously another hypothetical issue.’
454
451
Parties’ response to Remedies Notice, paragraph 1.5
452
Parties’ response to Remedies Notice, paragraph 1.4
453
Parties’ response to Remedies Notice, paragraph 1.4
454
Remedies hearing with 247, page 6, lines 15-17.
192
However, all four
455
third parties with whom we held calls expressed that in
their view a 100% divestiture of 247 would be the only effective remedy to
the SLC we found, noting that anything less would allow Hunter Douglas to
retain a degree of control over the strategy of 247.
[One third party] submitted that it ‘considered that the only viable remedy
to address the SLC is a 100% divestiture of the ordinary share capital of
247.’
456
[] describes that it ‘arrived at this position because Hunter
Douglas is not only dominant in the online retail supply of M2M blinds but
has a significant presence in both related horizontal and vertical
markets.’
457
[One third party] also ‘believes that a full divestiture of 247 is the most
effective remedy for addressing the provisional SLC.’
458
[Two third parties] also considered full divestiture the best option in order to
remedy the SLC.
459
[One of the two third parties] added, however, that this
was dependent on the suitability of the acquirer.’
460
Neither the Parties nor any third parties considered that, in order to
effectively remedy the SLC, it was necessary to include 247’s international
operations in the scope of the divestiture package.
CMA assessment of the scope of the divestiture package
The CMA takes divestiture of all or part of the acquired business as its
starting point because ‘restoration of the pre-merger situation in the
markets subject to an SLC will generally represent a straightforward
remedy.’
461
In defining the scope of a divestiture package that will satisfactorily address
an SLC, the CMA will normally seek to identify the smallest viable, stand-
alone business that can compete successfully on an ongoing basis and that
includes all the relevant operations pertinent to the area of competitive
overlap. This may comprise a subsidiary or a division or the whole of the
business acquired.
462
463
455
[].
456
[].
457
[].
458
[].
459
[].
460
.[].
461
Merger remedies guidance CMA87, paragraph 5.6.
462
Merger remedies guidance CMA87, paragraph 5.7.
463
We note that neither Hunter Douglas nor third parties thought that it was necessary to include 247’s overseas
operations within the scope of the divestiture package. Given the small size of these operations, we consider
193
The divestiture may comprise the sale of all relevant assets in one package
or the sale of assets grouped together in a limited number of packages.
464
We note Hunter Douglas’ comments at paragraph 11.20 that structural
remedies which re-establish the pre-merger ownership structure (ie Hunter
Douglas owning 49% of 247) should be expected to address the adverse
effects at source, and that re-establishment of this structure would not
require a full divestiture. We also note its submission in the same
paragraph that a full divestiture would be ultra vires and disproportionate.
We do not agree that a full divestiture would be ultra vires. The CMA has a
statutory duty to decide whether action should be taken to remedy, mitigate
or prevent the SLC or any adverse effect resulting from the SLC,
465
and the
CMA must have regard to the need to achieve as comprehensive a remedy
as is reasonable and practicable. The guidance says that ‘in identifying a
divestiture package, the CMA will take, as its starting point, divestiture of all
or part of the acquired business.
466
In this case we consider that the full divestiture of 247 would represent an
effective remedy which would be relatively easy to implement. We
acknowledge that the scope of a full divestiture goes beyond what would be
required to restore 247’s ownership structure to the level that would have
prevailed absent the 2019 Transaction, which would be the minimum scope
of any remedy package. However, the CMA’s guidance recognises that
merger parties may be required to add further assets to a remedy package
in order to secure divestment to a suitable purchaser
467
.
Our assessment of the proportionality of effective remedies is set out in
paragraphs
11.147 to 11.173.
Identification of a suitable purchaser
In our Remedies Notice, we invited views on whether there were any
specific factors to which the CMA should pay particular regard in assessing
purchaser suitability, and whether there were risks that a suitable
purchaser was not available.
468
The CMA will wish to satisfy itself that a prospective purchaser:
that they do not have a material impact on 247’s ability to compete in the UK. As a result, we propose to exclude
them from the divestiture package under both remedy options.
464
Merger remedies guidance CMA87, paragraph 5.8.
465
Sections 35 and 26 of the Act
466
Merger remedies guidance CMA87, paragraph 5.6.
467
Merger remedies guidance CMA87, paragraph 5.9. See for example Eurotunnel/SeaFrance.
468
Remedies Notice, paragraph 20(a).
194
(a) is independent of the merger parties;
(b) has the necessary capability to compete;
(c) is committed to competing in the relevant market;
(d) and that divestiture to the purchaser will not create further competition
concerns.
469
Parties’ and third-party views on the identification of a suitable purchaser
The Parties submitted that there should not be any difficulty in finding a
suitable purchaser, or purchasers, meeting the CMA’s normal purchaser
suitability criteria as set out in the Remedies Notice
470
for any potential
divestiture package. Hunter Douglas submitted that ‘247 is a profitable
business with growth potential, notwithstanding that its market share has
fallen in recent years, and therefore attractive for a potential purchaser.
471
However, Hunter Douglas submitted that a primary concern, whether the
purchaser be a retailer or a private equity firm, was the risk of losing the
expertise of one of the Founding Shareholders, Jason Peterkin, who
continues to manage the business.
Third party views
There was broad consensus from the parties and third parties on the need
for a suitable purchaser to meet our normal purchaser suitability criteria.
Third parties also submitted that a suitable purchaser should be
independent from Hunter Douglas. One third party, [], submitted that
‘[s]hould a purchaser come forward which has an existing corporate
relationship with these two entities it is suggested that this will likely stifle
the development of this market.’
472
All third parties who submitted evidence commented on the importance of
e-commerce experience in order to remain competitive in the online
market. One third party commented it is easy to lose money very quickly.
473
[] submitted that ‘[h]aving a background in e-commerce would be
essential, but not necessarily within the blinds space.’ Further noting that
469
Further detail of the purchaser suitability criteria are set out in Merger remedies guidance CMA87, paragraph
5.21.
470
Parties’ response to Remedies Notice, paragraph 4.11.
471
Parties’ response to Remedies Notice, paragraph 4.11.
472
[].
473
[].
195
‘[f]or example, Dunelm or Next could potentially be suitable acquirers
where they have a big home furnishing outlet already.’
474
[] said that ‘[a] potential purchaser would need either a good amount of
e-commerce experience, or to be an existing blinds manufacturer or have
substantial experience within the blinds and/or window coverings industry.
Possessing all of these attributes would be ideal but need to have at least
one or the other to be best placed.’
475
[] said that ‘full divestiture of 247 is the best option for remedying the
provisional SLC finding. However, that remedy is based upon the suitability
of the acquirer. To avoid 247 being run inefficiently and risking its viability,
the acquirer of the company must have at least a considerable amount
experience within the e-commerce sector, ideally within home
furnishings.’
476
[] also highlighted the importance of online retail experience to 247’s
ability to remain a strong competitor, commenting that ‘an acquirer without
this proven industry experience and expertise is highly likely to result in the
competitive landscape for online MTM blinds becoming worse, not
better. We reach that conclusion because in the scenario 247 blinds
market share reduces, then consistent with the CMA findings and
conclusions, the Hunter Douglas share of the market would be even
greater.’
477
Expressions of interest
[Three third parties] submitted that they would be interested in purchasing
100% of 247. [One third party] said that they would not be interested in
purchasing 247.
478
[].
479
[].
480
CMA assessment of identification and likely availability of a suitable purchaser
Hunter Douglas characterises 247 as a profitable and growing business in
an expanding market, which should prove attractive to potential
474
[].
475
[].
476
[].
477
[].
478
[].
479
[].
480
[].
196
purchasers. Evidence from both the main and third parties suggested that
there would be a wide pool of potential purchasers.
With a full divestiture, we consider that the CMA’s usual suitability criteria,
as set out in paragraph 11.36, would be appropriate. In considering the
capability of potential purchasers, we consider that any potential purchaser
would need to identify a management team with experience of online
retailing or e-commerce, preferably in blinds or similar products. This could
be achieved either from within the purchaser’s existing operations or
recruitment of an experienced management team. Based on the evidence
available to us, we consider it likely that a suitable purchaser could be
identified to acquire 100% of 247.
Conclusions on identification and availability of a suitable purchaser
Based on the information currently available, our view is that it is likely that
a suitable purchaser would be found for this remedy option. We further
consider that there are likely to be buyers that could satisfy our suitability
criteria set out in paragraph
11.36 above.
Ensuring an effective divestiture process
It is the CMA’s standard practice to provide for the appointment of a
Divestiture Trustee to dispose of the divestiture package, if Hunter Douglas
fails to achieve an effective disposal within the Initial Divestiture period, or if
the CMA has reason to be concerned that Hunter Douglas will not achieve
an effective disposal within the Initial Divestiture Period. This helps ensure
that Hunter Douglas has a sufficient incentive to implement the divestiture
promptly and effectively.
Views of the Parties and third parties on an effective divestiture process
In the event of divestiture Hunter Douglas said that:
‘[I]n the context of the global Covid-19 pandemic, a timescale of at
least 9 months is appropriate in order to allow for the divestment to
take place. Such a period will allow Hunter Douglas to identify a
suitable buyer, engage in appropriate negotiations with that buyer,
obtain CMA approvals and put in place necessary documentation.
Hunter Douglas will endeavour to identify a buyer and complete a sale
process as quickly as possible, however, given the current,
unprecedented situation resulting from the Covid-19 pandemic, it may
197
take more time than usual to do so and a modest extension from the
CMA’s usual practice of 6 months is therefore appropriate.’
481
Hunter Douglas also said that there would be no risk to the 247 business
during the divestiture period, given that:
(a) The measures put in place under the IEO would continue to ensure
that 247 is run in a profitable manner; and
(b) there is no asset risk since Hunter Douglas has no incentive to run
down a business in which it would continue to hold a significant
minority stake.
482
Furthermore, Hunter Douglas submitted that it ‘did not consider it
necessary or proportionate to mandate a divestiture trustee to oversee the
divestment process, submitting that there is no reason to believe that
Hunter Douglas will not be able to identify a suitable purchaser and that it
will not proceed with divestment. As a result, Hunter Douglas believes that
mandating a divestiture trustee would incur an unnecessary cost and would
be disproportionate.’
483
• Third party views
In response to Hunter Douglas’ request for a 9-month divestiture process,
third parties commented that it is likely only 6 months would be required
and 9 months for the divestiture would involve unnecessary delay to the
process.
[] described the request as ‘excessive’ and submitted that ‘[u]ntil a
divestiture is achieved the main parties will continue to grow at an
enhanced rate due to current market conditions and their positions – as
such it will become increasingly difficult for any new entrants to establish
themselves in the market.’
484
[] supported this view stating that it ‘does not believe nine months will be
required to sell 247, due to the profitability of the business. The online M2M
blinds market is not so niche that it would be difficult to find buyers.’
485
As regards the appointment of a divestiture trustee, [] submitted that ‘It is
suggested that the manner in which the parties attempted to implement the
transaction giving rise to the CMA's investigation showed a surprising lack
481
Parties’ response to Remedies Notice, paragraph 4.14.
482
Parties’ response to Remedies Notice, paragraph 4.15.
483
Parties’ response to Remedies Notice, paragraph 4.17.
484
[].
485
[].
198
of transparency. Therefore, it is suggested that the CMA may wish to keep
open the possibility of appointing a divestiture trustee to ensure that should
it deem that a divesture is necessary – that it is appropriately
implemented.’
486
CMA assessment of an effective divestiture process
• Timescale allowed for divestiture
While the Coronavirus (COVID-19) pandemic has had a significant impact
on a number of markets, the online M2M blinds market does not appear to
have been negatively affected. [].
487
Unlike bricks and mortar retailers, online retailers have not had to close
their stores, and do not need to demonstrate a post-lockdown trading
record to determine the value of the business to a potential purchaser. We
[] consider that a divestiture period of [] (the Initial Divestiture Period)
would be appropriate in this case.
• Provision for appointment of a divestiture trustee
We did not receive any evidence to suggest that a Divestiture Trustee
should be appointed at the outset of the divestiture process. As a result, we
do not propose to appoint a Divestiture Trustee at the outset of the
divestiture process.
However, to ensure a timely completion of this remedy, we conclude that
we should reserve the right to appoint a Divestiture Trustee if:
(a) Hunter Douglas fails to complete the divestiture process within the
Initial Divestiture Period;
(b) We believe that there is a risk that the divestiture process would be
delayed or fail to complete within the Initial Divestiture Period;
(c) We believe that Hunter Douglas is not engaging constructively with
the divestiture process; and / or
(d) There is a material deterioration in 247’s business during the
divestiture process.
486
[].
487
[].
199
In line with the CMA’s normal practice,
488
if appointed, a Divestiture Trustee
would be tasked with completing the divestiture of 247 to a potential
purchaser approved by the CMA and at no minimum price. We propose,
again in line with the CMA’s normal practice, that the Monitoring Trustee’s
appointment should cover oversight of the divestiture process and continue
until completion of the divestiture.
Assessment of the effectiveness of full divestiture
We consider that the full divestiture of 247 would represent an effective
remedy and provide a comprehensive solution to the SLC we have found.
Partial divestiture of 247
Description of remedy
This remedy option would involve Hunter Douglas divesting 51% of 247’s
ordinary share capital, whilst retaining 49% of 247’s share capital as a
significant minority shareholder. This would lead to an ownership structure
similar to the one set out in our counterfactual.
Remedy design issues
The framework for addressing design issues for divestiture remedies is set
out in paragraphs 11.12 to 11.15.
Scope of divestiture package
The general considerations that the CMA will take into account when
assessing the scope of a divestiture package are set out in paragraphs
11.28 to 11.30 above.
A specific issue with this divestiture option is that it anticipates Hunter
Douglas retaining a substantial minority shareholding in 247. A key design
consideration is which, if any, rights in 247 Hunter Douglas should also be
permitted to hold, in addition to its minority shareholding. In our Remedies
Notice, we set out a number of potential veto rights that could be held by
Hunter Douglas in the event of a partial divestiture of its stake in 247.
489
Until completion of the 2019 Transaction, Hunter Douglas held a number of
veto and other rights in 247, which were set out in the Stakeholder
488
Merger remedies guidance, CMA87 (13 December 2018), paragraph 5.43.
489
Remedies Notice
200
Agreement signed at the time of the 2013 Transaction. These are set out in
paragraph 3.4.
Under this remedy, and taking into account the treatment of minority
shareholdings in other contexts, the divestiture of 51% of 247 may include
Hunter Douglas holding some or all of the following veto rights:
(a) Appointment of additional directors;
(b) Approval of the annual budget;
(c) Acquisitions;
(d) Entering into new lines of business;
(e) Geographic expansion into new countries;
(f) Any backward integration into assembly or production of any of the
products sold by 247;
(g) Long term agreements (exceeding one year in duration);
(h) Financing arrangements with banks or other parties;
(i) Dividends in excess of 35% of profit after tax;
(j) Offers on 247’s website at less than 15% gross profit; and/or
(k) Transactions with related parties (eg companies within the same
corporate group as the purchaser or Hunter Douglas).
Consideration of the effect of each of these rights on the effectiveness of
this remedy option is set out in the assessment section below.
Parties’ and third-party views on the scope of the divestiture package
As set out above, this remedy would lead to an ownership structure similar
to that set out in our counterfactual with Hunter Douglas owning 49% of
247. However, we note the Parties’ submission that they ‘consider that this
conclusion has been reached on the basis of an incorrect assessment of
the nature of the rights acquired by Hunter Douglas in 2013 and the legal
rights and obligations created by the suite of documents entered into by
Hunter Douglas and the Founding Shareholders at the time.’
490
490
Parties’ response to Remedies Notice, paragraph 4.9.
201
In their response to the Remedies Notice, the Parties submitted that ‘a
divestment of 51% of the ordinary share capital of 247 will clearly be
effective in addressing the SLC identified by the CMA. No additional
measures are required in this respect.’
491
In the Hunter Douglas remedies hearing, Hunter Douglas stated ‘I have
never seen a no-rights participation of that size, I have never seen it in
practice.’ Hunter Douglas further submitted that ‘49% is not control’ as the
51% shareholder would be the majority shareholder and in control.
Hunter Douglas described the rights that it acquired as a result of the 2013
Transaction as ‘guardrails to prevent a defocus of the business.’
Hunter Douglas also said ‘if we were working with, let us say, a 51 per cent
other party, the first thing we would have to do is agree on a manager; that
would be the essential piece. We would want to able to veto the manager
or we would not invest, because this business is nothing without the
manager.’
Hunter Douglas said that there were some rights that it might be less
concerned about holding as a 49% investor. These included vetoes over
the annual budget, on offers at less than 15% gross profit and long-term
agreements. However, it also said that there were other veto rights that it
might require, such as the identity of a CEO, expansion into other product
or geographic markets, and related party transactions.
At its remedies hearing, 247 said that the presence of veto rights made no
difference, and that between 2012 and 2019 Hunter Douglas had never
exercised any of the veto rights. It also said that the existence of the veto
rights would not constrain 247’s ability to compete effectively.
In its response to the Remedies Working Paper, the Parties said that ‘The
CMA is not required to remove all influence in order to remedy the SLC
provisionally found in the Provisional Findings. Rather, the CMA should be
concerned only to ensure that the remedy removes control as this was the
change resulting from the 2019 Transaction.’
The Parties also said that ‘the limit of the CMA’s jurisdiction concerning the
impact of the rights is set by the relevant counterfactual.’
491
Parties’ response to Remedies Notice, paragraph 4.2
202
• Third party views
In the event that a partial divestiture of 247 was required, all four
492
third
parties, with whom we held calls, commented that this was unattractive,
with or without rights attached for Hunter Douglas’ 49% stake.
Notwithstanding that a purchaser of 51% of 247 would be a majority
shareholder, these third parties considered that Hunter Douglas would still
have significant influence over 247 given the nature of its commercial
relationship with 247 and other M2M blinds retailers. As a result, they did
not think that a partial divestiture would remedy the SLC.
[], for example ‘does not believe a partial divestment would be effective.
As Hunter Douglas supply to 247 and various other companies, Hunter
Douglas could behave as though it still has complete control of 247.’
493
[] submitted that having Hunter Douglas as a 49% shareholder may be
beneficial for certain purchasers, although this may limit the effectiveness
of the remedy:
‘If Hunter Douglas was to retain 49% of 247, this would still allow it to
exert significant influence on 247, namely through product sourcing
and pricing agreements. It would be appealing for an investor to have
Hunter Douglas on board as a partner in 247, due to Hunter Douglas’
presence in the market. For these reasons, [] doesn’t believe shared
ownership would truly remedy the SLC.’
494
The third parties with whom we held calls, by contrast, were clear that the
rights set out in the Remedies Notice make the prospect of owning 51% of
247 unattractive given the influence retained by Hunter Douglas over the
business as a result of the rights.
[] submitted that 51% would be an insufficient stake to attract a
purchaser, ‘especially if the veto/rights HD were seeking were given
as these will effectively restrict 247 Blinds so much as to make them unable
to make the changes required to grow and become competitive in
future.’
495
Therefore, a divestiture of 51% of 247 could be effective ‘only on
the basis that… Hunter Douglas do not have the vetoes and rights
suggested.’
496
492
[].
493
[].
494
[].
495
[].
496
[].
203
Similarly, [] submitted that ‘[i]n the event of shared ownership of
247, [] would not want Hunter Douglas involved in any decision making
within 247 of any sort. Information Hunter Douglas would receive
would also be the bare minimum.’
497
CMA assessment of the scope of the divestiture package
As set out in paragraph 11.3 above, this chapter sets out the inquiry
group’s decision on what remedy would be effective and proportionate in
addressing the SLC identified in the group’s findings.
The divestment of 51% of 247 would replicate the ownership structure that
we have found would exist in the absence of the Merger, with Hunter
Douglas retaining 49% of the shares.
We have assessed the scope of this option in line with CMA guidance,
which says that ‘the CMA will normally seek to identify the smallest viable,
stand-alone business that can compete successfully on an ongoing basis
and that includes all the relevant operations pertinent to the area of
competitive overlap.’
In order to comprehensively remedy the SLC, the scope of the divestiture
package should be sufficient to create market conditions no less
competitive than those set out in the counterfactual in this final report. In
this counterfactual, 247 would have more independence than it had had
prior to the 2019 Transaction, with Hunter Douglas no longer being able to
exercise the same veto and other rights it held prior to the 2019
Transaction under the Stakeholder Agreement.
We do not agree with the Parties’ submission in paragraph 11.79 that our
choice of counterfactual limits the scope of our proposed remedies. We are
required to ‘have regard to the need to achieve as comprehensive a
solution as is reasonable and practicable to the substantial lessening of
competition and any adverse effects resulting from it’.
498
This may mean
pursuing a remedy which goes beyond the counterfactual scenario, if more
limited remedies are not found to be effective e.g. because no suitable
purchaser can be found or the proposed divestment package is unviable.
The third-party evidence shows concerns that were Hunter Douglas to hold
extensive rights over 247’s operations, it would be able to exert significant
additional influence beyond the most likely counterfactual scenario set out
at paragraph 6.49 above. This would undermine the independence of 247
497
[].
498
Enterprise Act 2002, section 35(4).
204
(for example by restricting its growth prospects) and therefore the
effectiveness of the remedy in restoring competition. It would also
potentially further reduce the pool of interested purchasers, increasing the
risk profile of this remedy option.
We note Hunter Douglas’ characterisation of the veto rights as ‘guardrails’
to keep the business focussed. We are concerned that the presence of
such extensive rights might focus the business away from direct
competition with Blinds2Go. More broadly, given that these veto rights do
not exist in the counterfactual that we have found and the fact that Hunter
Douglas and 247 are important rivals to each other, we consider that
inclusion of extensive rights in the divestiture package would undermine the
effectiveness of the remedy.
We consider that a limited set of veto rights could, in theory, be designed
so as to ensure that Hunter Douglas would have limited influence over the
operational independence and competitive strategies of 247, and so
present a significantly lower, and more manageable, risk to the
effectiveness of this remedy option.
Therefore, we consider that it is conceivable that Hunter Douglas could
hold a limited set of veto rights whilst still providing for the divestiture
package to represent an effective remedy.
Should this remedy be implemented, these rights are likely to form part of
the divestiture negotiations between Hunter Douglas and potential
purchasers. As a result, we do not think that it is appropriate at this stage to
specify precisely what form these rights might take. In line with normal
practice, the final divestiture (including any additional rights held by Hunter
Douglas) is subject to approval from a CMA Remedies Group.
However, we would expect any rights held by Hunter Douglas to be
consistent with the following principles:
(a) The right(s) must not have a material impact on the ability of the
purchaser to pursue an independent competitive strategy with 247
from Hunter Douglas’ other business interests, including Blinds2Go;
and
(b) The rights may be consistent with protecting the value in Hunter
Douglas’s 49% equity stake in 247.
Should this remedy option be implemented, we will consider whether these
principles need to be further refined as part of the remedies implementation
process. In addition, during the divestiture process it is usual practice for
205
the Sale and Purchase Agreement to be approved by the CMA. This gives
us further assurance that any rights held by Hunter Douglas would not
undermine the effectiveness of the remedy.
We conclude that partial divestiture of 51% of the ordinary share capital in
247 could be sufficient in scope to remedy the SLC we have found,
provided any limited rights held by Hunter Douglas post divestiture adhere
to the principles set out in paragraph 11.96. Such a divestiture would
replicate the ownership structure of 247 absent the merger, enabling 247 to
operate with sufficient independence of Hunter Douglas to remedy the SLC
we have found. We consider the costs of this remedy and its proportionality
in paragraphs 11.147 to 11.173 below.
Identification of a suitable purchaser
We set out the CMA’s criteria for the identification of a suitable purchaser in
paragraphs 11.35 and 11.36.
Parties’ and third-party views on the identification of a suitable purchaser
Hunter Douglas said that whether or not it holds the rights set out in the
Annex to the Remedies Notice, it should not be difficult to find a suitable
purchaser for a 51% shareholding in 247. Hunter Douglas submitted that
the ‘rights do not significantly go beyond rights which would typically be
granted to a minority investor for the protection of its investment. Such
rights should not, therefore dissuade a potential purchaser or purchasers
from acquiring 51% of the ordinary shares of 247. Indeed, many private
equity and conglomerate businesses have partial stakes in their portfolio
companies.’
499
In their response to the Remedies Working Paper, the
Parties said that they did not think that identifying a suitable purchaser for
51% of 247 would be any more difficult than for 100%.
At the remedies hearing, [].
[].
With regards to the rights held by Hunter Douglas, at the 247 remedies
hearing, Mr Peterkin said ‘none of the veto rights were ever exercised, and
I was never really put under any pressure to change the things that I was
doing.’
499
Parties’ response to Remedies Notice, paragraph 4.12.
206
[], Mr Peterkin also said ‘In terms of Hunter Douglas's veto rights and
shareholder rights, I personally did not have any problems with them.’ [].
In their response to the Remedies Working Paper, the Parties said that
247 had a track record of competing strongly against other Hunter Douglas
entities, [].
• Third party views
There were a variety of views from third parties as to a suitable
purchaser of 51% of 247. As described in paragraphs 11.39 to 11.44
above, third parties generally deemed it essential that any purchaser had e-
commerce experience.
However, in the scenario of owning 51%, third parties commented on
their concerns regarding the operational relationship with Hunter Douglas
as a significant shareholder of 247 and therefore impacting the
attractiveness of owning 51% of 247 as opposed to 100%.
[], commented that ‘Assuming the current leadership at 247 retain
their roles, a less experienced acquirer of the 51% share could still make it
work. However, if a new Managing Director came in, that person would
need e-commerce experience.’ Adding that, ‘[i]f the acquirer didn’t have a
clear plan for online marketing, they could lose a lot of money very quickly.’
500
[] explained that they would not wish to own 51% of 247, ‘[t]his is due
to the control and insight shared ownership would give to Hunter Douglas.
Knowledge of []’ marketing techniques and popular products could be
gained and exploited by Hunter Douglas.’ Furthermore, ‘[e]ven if Hunter
Douglas were to operate as a silent partner in 247, Hunter Douglas would
still be able to apply backdoor pressure through Arena (one of its
subsidiaries).’
501
Hunter Douglas responded to this evidence, saying that
we should place no weight on this statement since Arena does not supply
247, and it was not explained how this pressure would be asserted given
the purchaser’s majority stake.
[] submitted that, ‘[e]xpertise can be bought provided sufficient funds
are available’. [] further submitted that:
‘If the purchaser is not a blinds manufacturer who can be
competitive regarding product sourcing and pricing, then it is likely
500
[].
501
[].
207
that purchaser would want Hunter Douglas to retain a 49% share in
247. This allows the purchaser to tap into Hunter Douglas’ expertise
and product sourcing. 49% is a large portion of 247, so Hunter Douglas
would use every resource available to it to achieve success. If the
purchaser was a company like [], the purchaser would likely want
100% of 247.’
502
[] stated that:
‘The only scenario in which [it] would consider a 51% ownership of 247
would be if Hunter Douglas operated as a passive partner with no
control rights, no veto rights, no influence whatsoever. Hunter Douglas
would be treated like a minority investor and its interest to be purely
financial return. [] would report numbers to Hunter Douglas on an
annual or quarterly basis. Anything more than that invites a level of
insight or involvement that is not consistent with the role of a passive or
minority shareholder.’
503
[] commented that ‘Whether or not a 51% stake vs. a 100% stake
would be desirable to a potential buyer would depend on their individual
circumstances and the attributes that they would need… having some
experience in the market would of course be advantageous and probably
necessary to balance out Hunter Douglas’ influence over 247.’
504
CMA assessment of identification and availability of a suitable purchaser
As set out in paragraph 11.48, the evidence suggests that 247 is an
attractive business to potential purchasers wishing to own assets in this
sector. We would also expect a majority owner to have, or be able to
procure, the expertise set out in paragraph 11.49. We would note our
finding that we consider this expertise to be sufficiently uncommon as to
represent a barrier to entry or expansion, and so may also limit the number
of suitable purchasers.
505
For the reasons set out at paragraph 6.46 above, we found that, in the
most likely counterfactual to the 2019 Transaction, the interest in 247 held
by the 247 Founding Shareholders would have been sold to a single
purchaser. As such, to ensure that Hunter Douglas is not able to exercise
more influence over 247 than in the most likely counterfactual, our strong
preference is a requirement that the 51% shareholding to be divested to a
502
[].
503
[].
504
[].
505
This is discussed further in the chapter on countervailing factors, paragraphs [9.12 to 9.15]
208
single purchaser. If the divestiture is to be made to more than one suitable
purchaser, we would need to be satisfied that this would not lead to
additional effectiveness risk.
We note the mixed evidence from third parties as to whether a potential
purchaser would want to own a business where Hunter Douglas was a
significant minority shareholder, having direct material influence over 247
through an ability to, at least, block special resolutions, or indirect influence
through the commercial relationships of its other group companies. We also
note the evidence from [] that Hunter Douglas might be able to exert
‘backdoor pressure’. While we accept Hunter Douglas’ response that Arena
does not supply 247, we would note that Hunter Douglas is a significant
and growing player in the wholesale supply of M2M blinds, which may
mean it will have an increasing ability to influence any owners of 51% of
247.
506
As a result of this evidence we would expect the pool of potential
purchasers to be significantly smaller than that for a full divestiture, [].
This significantly smaller pool would materially increase the risk that a
suitable purchaser might not be found to effectively remedy the SLC.
The CMA’s guidance says that a prospective purchaser should:
(a) be independent of the merger parties;
(b) have the necessary capability to compete; and
(c) be committed to competing in the relevant market.
The guidance further states that divestiture to the purchaser will not create
further competition concerns.
507
Given Hunter Douglas’ majority ownership of and potentially increasing
stake
508
in the largest supplier in the market, Blinds2Go, we have particular
concerns with the ability, incentives and commitment of a potential
purchaser to ensure that 247 has the necessary capability to compete
independently with Blinds2Go. We consider that a financial purchaser
would generally have less ability to mitigate the potential influence of
Hunter Douglas on 247’s competitive strategy than would a trade purchaser
with its own expertise and perspective on competitive conditions.
506
For example, in July 2020, Hillarys Blinds Holdings Limited purchased TLC Shutters S.R.L., a Romanian
assembler of shutters which supplies Blinds2Go and other UK retailers outside the Hunter Douglas group.
507
Merger remedies guidance CMA87, paragraph 5.21.
508
Hunter Douglas currently owns 65% of Blinds2Go, with options to acquire a further 15% in 2021 and the
remaining 20% in 2026.
209
We also have concerns that Hunter Douglas and a potential purchaser
may enter into further arrangements (for example, such as those Hunter
Douglas made with the 247 Founding Shareholders) that may limit 247’s
future competitive capability and incentives. While this risk will be reduced
by the fact that the entry into such arrangements between competitors is
restricted by competition law, we still consider this to be a real risk. This
risk can be mitigated by making provision in any final undertakings or Order
to prohibit Hunter Douglas from entering into any such arrangements that
limit 247’s competitive capabilities. However, effective mitigation relies on
the CMA being able to successfully monitor and enforce this provision.
Purchaser suitability is usually assessed by a CMA Remedy Group as
part of the purchaser approval process during the implementation phase of
a divestiture remedy, using evidence such as detailed business plans. As
we are not likely to receive such evidence before publication of our Final
Report, we are unable to assess at this stage whether any proposed
standalone purchaser, [] meets the criteria to be considered a suitable
purchaser.
[]:
(a) [].
(b) [].
(c) [].
(d) [].
In summary, we consider that a small pool of potential purchasers may
exist for the divestiture package specified in this remedy option. It is
possible that a suitable purchaser could be found from this pool of potential
purchasers, in which case an effective divestiture could take place.
However, the evidence available to us on the existence of an interested
suitable purchaser for 51% of 247 is incomplete at this point as such a
package has not been widely marketed, we have not yet conducted a full
suitability assessment of potential purchasers, and we can only judge
levels of interest on the basis of what we have been told by third parties.
509
On the basis of the information currently available to us we consider
that, in principle, there is sufficient likelihood of one or more potential
purchasers meeting our suitability requirements, and associated restrictions
being put in place to prevent Hunter Douglas from subsequently seeking to
509
This is not uncommon at this stage of a merger inquiry.
210
limit 247’s future competitiveness through separate agreements with any
such purchaser. However, we also consider that there remains a material
risk that no purchaser will meet the requirements needed for an effective
divestiture.
CMA guidance states that ‘substantial uncertainty as to whether a
suitable purchaser will emerge will generally not be sufficient for the CMA
to conclude that any form of divestiture remedy is not feasible. The CMA
has found that it is normally possible to implement divestiture remedies,
despite such uncertainties, given flexibility in the disposal price.’
510
However, we note the evidence from several third parties, who
suggested that a potential purchaser might not want to own a business
where Hunter Douglas was the minority shareholder.
As noted in paragraphs 11.115 to 11.116 above, the more extensive
the rights that Hunter Douglas holds in 247, the less likely that the remedy
package will be effective, or that a suitable purchaser will be available.
On the basis of the evidence we have seen, we conclude that a
majority ownership stake in 247 would be less attractive and would lead to
a smaller pool of potential purchasers in a situation where Hunter Douglas
holds more than a limited set of minority rights.
In the circumstances of this case, and consistent with CMA guidance,
we consider that the level of uncertainty around finding an interested
suitable purchaser of 51% of the shares in 247 falls within acceptable
bounds, and so does not in itself prevent the remedy from being effective.
Ensuring an effective divestiture process
Our assessment of an effective divestiture process for a Full Divestiture
is set out above at paragraphs 11.51 to 11.63.
In our view, the steps needed to ensure an effective divestiture process
are similar for partial divestiture. We received no evidence from main or
third parties to suggest otherwise.
[]. We also do not propose to appoint a divestiture trustee at the
outset of the divestiture process for this option, but reserve the right to do
so if any of the situations set out in paragraph 11.62 come to pass. We also
510
Merger remedies guidance CMA87, paragraph 3.51
211
propose to continue the Monitoring Trustee’s appointment until completion
of the divestiture.
[].
Assessment of the effectiveness of partial divestiture
We consider that the partial divestiture of 51% of the ordinary share
capital of 247 with a limited set of veto rights presents a certain level of risk
in terms of effectiveness, principally the uncertainty around the existence of
an interested and suitable purchaser. However, consistent with CMA
guidance and given flexibility in the disposal price, we consider that these
risks are within tolerable bounds. Therefore, this remedy option would
represent an effective remedy and provide a comprehensive solution to the
SLC we have found, provided that a suitable purchaser can be found.
Conclusion on remedy effectiveness
We conclude that the following remedies would be effective in
addressing the SLC and resulting adverse effects that we have found:
(a) Full divestiture of 247; and
(b) Partial divestiture of 247, subject to a suitable purchaser being found.
Relevant customer benefits (RCBs)
When deciding on remedies, the CMA may have regard to the effects
of remedial action on any RCBs. In this sub-section, we consider whether
there are any RCBs (within the meaning of section 30 of the Act) that
should be taken into account in our remedy assessment.
An effective remedy to an SLC could be considered disproportionate if
it prevents customers from securing substantial benefits arising from the
Merger, where these benefits outweigh the SLC and any resulting adverse
effects. Insofar as these benefits constitute RCBs for the purposes of the
Act, the statutory framework allows us to take them into account when we
decide whether any remedy is appropriate.
RCBs that will be foregone due to the implementation of a particular
remedy may be considered as costs of that remedy. The CMA may modify
a remedy to ensure retention of an RCB or it may change its remedy
selection. For instance, it may decide to implement an alternative effective
remedy, or it may decide that no remedy is appropriate.
212
Framework for assessing RCBs
The burden of proof of whether RCBs arise from a merger is on the
merging parties. CMA guidance states that the 'merger parties will be
expected to provide convincing evidence regarding the nature and scale of
RCBs that they claim to result from the merger and to demonstrate that
these fall within the Act's definition of such benefits.'
The Act defines RCBs as a benefit to relevant customers in the form of
lower prices, higher quality, or greater choice of goods or services in any
market in the UK, or greater innovation in relation to those goods or
services. For these purposes, relevant customers are direct and indirect
customers (including future customers) of the merger parties at any point in
the chain of production and distribution - they are not limited to final
consumers.
In addition, in the case of completed mergers, to be properly
considered as an RCB under the statutory definition, the CMA must believe
that:
(a) the benefit has accrued as a result of the creation of the relevant
merger situation concerned or may be expected to accrue within a
reasonable period as a result of the creation of that situation; and
(b) the benefit was, or is, unlikely to accrue without the creation of that
situation or a similar lessening of competition.
When assessing the merger parties' evidence on the claimed benefits,
the CMA must therefore ask itself whether each claimed benefit has or may
be expected to accrue as a result of the merger ((a) above), and, whether
that benefit was, or is, unlikely to accrue without the merger or a similar
lessening of competition ((b) above). With regard to the latter, in practice
the CMA will consider whether the merger parties' evidence is sufficient to
demonstrate that the claimed benefit could not be achieved by plausible
less anti-competitive alternatives to the merger.
In previous cases where RCBs have been accepted, the type of
evidence accepted included implementation plans which have been found
more persuasive were detailed and advanced. The merging parties'
incentives to implement and pass on the benefits post-merger will also be
relevant to the likelihood of RCBs being realised in practice.
213
Parties’ and third-party views on RCBs
In our Remedies Notice, we invited views on the nature of any RCBs
and on the scale and likelihood of such benefits and the extent (if any) to
which these were affected by different remedy options.
We received no submissions on RCBs from the Parties or third parties.
Assessment of RCBs
Since we have received no evidence on the nature or scale of RCBs,
we conclude that there are no RCBs associated with the remedy options
we have found to be effective.
The proportionality of effective remedies
In paragraph 11.135, we summarised our conclusions on which
remedies would be effective in addressing the SLC and the resulting
adverse effects. We set out below our assessment of, and conclusions on,
which of these would constitute a proportionate remedy.
Framework for assessment of proportionality of remedies
In order to be reasonable and proportionate, the CMA will seek to
select the least costly remedy, or package of remedies, that it considers will
be effective. If the CMA is choosing between two remedies which it
considers will be equally effective, it will select the remedy that imposes the
least cost or that is least restrictive (we call this the ‘least onerous effective
remedy’). In addition, the CMA will seek to ensure that no remedy is more
onerous than necessary or disproportionate in relation to the SLC and its
adverse effects.
511
To fulfil this, we first consider whether there are any relevant costs
associated with each effective remedy option. When considering relevant
costs, the CMA's considerations may include (but are not limited to):
512
(a) distortions in market outcomes;
(b) compliance and monitoring costs incurred by the Parties, third parties,
or the CMA; and
511
Merger remedies guidance, CMA87 (13 December 2018), paragraph 3.6.
512
Merger remedies guidance, CMA87 (13 December 2018), paragraph 3.10.
214
(c) the loss of any RCBs that may arise from the Merger which are
foregone as a result of the remedy.
However, CMA guidance states that ‘[as] the merger parties have the
choice of whether or not to proceed with the merger, the CMA will generally
attribute less significance to the costs of a remedy that will be incurred by
the merger parties than the costs that will be imposed by a remedy on third
parties.
513
In particular, for completed mergers, the CMA will not normally
take account of costs or losses that will be incurred by the merger parties
as a result of a divestiture remedy’, as it is ‘for the merger parties to assess
whether there is a risk that a completed merger would be subject to an SLC
finding, and the CMA would expect this risk to be reflected in the agreed
acquisition price’.
514
Having identified the least onerous effective remedy, we then consider
whether this remedy would be disproportionate to the SLC and its resulting
adverse effects. In doing so, we compare the extent of harm associated
with the SLC with the relevant costs of the proposed remedy.
515
Views of the Parties
The Parties consider the only remedy proportionate to the SLC to be
that where Hunter Douglas divests itself of 51% of 247’s ordinary share
capital and goes no further than this. Hunter Douglas submitted that any
further restrictions would exceed the scope of the SLC found by the 2019
Transaction and would both impose unnecessary costs and be
disproportionate.
516
As noted at paragraph 11.20 above, in their response to the Remedies
Working Paper, the Parties said that a full divestiture remedy would be ultra
vires because there was an alternative, less costly, less intrusive and
complete remedy – a partial divestiture. We do not agree that a full
divestiture is ultra vires, for the reasons set out at paragraphs 11.31 and
11.32 above. [].
Assessment of proportionality
In our assessment of proportionality, we first identified those remedies
that are likely to be effective and selected the remedy with the lowest cost,
or that was least restrictive (‘the least onerous effective remedy’). We then
513
Merger remedies guidance CMA87, paragraph 3.8.
514
Merger remedies guidance, CMA87 (13 December 2018), paragraph 3.9.
515
Merger remedies guidance, CMA87 (13 December 2018), paragraph 3.6.
516
Hunter Douglas response to the Remedies Notice, paragraph 4.2.
215
considered whether this remedy was disproportionate to the SLC and its
adverse effects.
Hunter Douglas acquired its 100% shareholding as a result of two
connected transactions. The 2019 Transaction, the SLC arising from which
we are looking to remedy, took its stake from an effective 49% to 100%.
The 2019 Transaction arose from the exercise of call options by the 247
Founding Shareholders. These call options had been granted at the time of
and as part of the 2013 transaction, when Hunter Douglas acquired loan
notes which gave it an effective 49% shareholding in 247.
The CMA’s guidance says:
‘In particular, for completed mergers, the CMA will not normally take
account of costs or losses that will be incurred by the merger parties as a
result of a divestiture remedy, as it is open to the merger parties to make
merger proposals conditional on the approval of the relevant competition
authorities. It is for the merger parties to assess whether there is a risk
that a completed merger would be subject to an SLC finding, and the
CMA would expect this risk to be reflected in the agreed acquisition price.
Since the cost of divestiture is, in essence, avoidable, the CMA will not, in
the absence of exceptional circumstances, accept that the cost of
divestiture should be considered when selecting remedies.’
517
While we note the circumstances surrounding the two transactions, we
still consider that the risk of an SLC finding was foreseeable and so have
not considered the cost of divestiture in choosing our preferred remedy.
Full divestiture
A full divestiture would not lead to any market distortions, and the CMA
would not face any monitoring costs. As set out in paragraph 11.146, we
have not identified any RCBs.
In line with CMA guidance set out at paragraph 11.150 above, we have
not taken into account any additional cost or loss to Hunter Douglas arising
from a full divestiture compared to a partial divestiture.
As a result, we do not consider there to be any material costs
associated with full divestiture. However, we do consider that a full
divestiture may be considered to be more intrusive than a partial divestiture
517
Merger remedies guidance, CMA87, paragraph 3.9.
216
in this case, since Hunter Douglas held the equivalent of 49% of 247’s
shares before the 2019 Transaction.
Partial divestiture
A partial divestiture would not lead to any market distortions, and as set
out in paragraph 11.146, we have not identified any RCBs. Given that the
rights granted to Hunter Douglas are limited to those necessary to protect
the underlying value of its minority stake, we do not consider that the
existence of these rights would lead to market distortions.
However, as set out in our assessment of effectiveness in paragraph
11.119 above, we consider that under this remedy option it would be
necessary to monitor the relationship between Hunter Douglas and the
purchaser of 247’s shares to ensure that there were no agreements
between them that might undermine the ongoing effectiveness of the
remedy.
As a result, we consider that partial divestiture would incur some
additional monitoring costs relative to full divestiture. However, as
discussed in paragraph 11.160, weighed against this, partial divestiture
would be less intrusive than full divestiture.
Identification of the least onerous effective remedy
We identified the following remedies as being effective solutions to the
SLC that we have found:
(a) Full divestiture of 247; and
(b) Partial Divestiture of 51% of 247, subject to a suitable purchaser
being found.
We have found that both remedies are, in principle, effective, provided
a suitable purchaser could be found.
We have found that full divestiture has a lower cost associated with it
than the second remedy option as it would not require any ongoing
monitoring by the CMA. However, this option is also significantly more
intrusive as it would leave Hunter Douglas with no shareholding in 247.
This compares to the pre-Merger conditions of competition that we found,
where it would still hold a 49% stake.
Set against these costs is the intrusiveness of Hunter Douglas
divesting its entire shareholding in 247. Balancing this cost and
217
intrusiveness, we find that full divestiture, despite avoiding monitoring costs
associated with partial divestiture, is not the least onerous effective
remedy.
We therefore conclude that, [], partial divestiture represents the least
onerous effective remedy. It is therefore our preferred remedy option.
[].
Are the remedies disproportionate to the SLC and / or adverse effects?
As set out in paragraph 10.4, we have found that the 2019 Transaction
has resulted in, or may be expected to result in, an SLC as a result of
horizontal unilateral effects in the online retail supply of M2M blinds in the
UK. In particular, we find that the 2019 Transaction has removed a direct
competitor from this market, resulting in an ability and incentive for the
Merged Entity to increase retail prices, lower the quality of its products or
customer service, and/or reduce the range of its products/services.
As discussed in paragraphs 11.154 to 11.169, the costs of our
preferred remedies are low in comparison to the harm that has arisen, or
may be expected to arise, from the SLC.
We note that, by its nature, divestiture is an intrusive intervention.
However, given the expected level of consumer harm which is expected to
arise from the SLC resulting from the Merger, we consider that this
intrusion is justified.
Conclusion on proportionality
We have found that each of the options set out in paragraph 11.164
would be an effective and proportionate remedy to the SLC and its resulting
adverse effects we have identified.
Remedy implementation
Having identified our preferred remedy, we now consider how it should
be implemented. We start by describing the risks associated with remedy
options, how these might materialise during the implementation process
and how they could be effectively managed. We then consider issues
relating to purchaser suitability.
218
Divestiture risks
The incentives of merger parties may serve to increase the risks of
divestiture. Although merger parties will normally have an incentive to
maximise the disposal proceeds of a divestiture, they will also have
incentives to limit the future competitive impact of a divestiture on
themselves. Parties may therefore have, on balance, an incentive to make
divestitures to weaker competitors of less competitive assets and may also
allow the competitiveness of divestiture packages to decline during the
divestiture process.
518
Composition risk
Having specified the scope of the effective divestiture package, this
should remedy the SLC and its adverse effects. However, partial
divestiture package has some residual composition risk due to the
uncertainties around any rights being granted to Hunter Douglas. We
propose to manage this through close oversight of the divesture process,
for example the specification of principles that would govern such rights
and by retaining the right to approve the Sale and Purchase Agreement
accompanying any divestiture.
Asset risk
In the context of the Merger, asset risk would be most likely to
materialise through the degradation of the business (intentional or
otherwise) via:
(a) Loss of customer contracts, or reducing quality of service;
(b) Insufficient maintenance of physical assets; and/or
(c) Key staff leaving;
Since the start of the CMA investigation, the Parties have been subject
to an Initial Enforcement Order (IEO) aimed at preventing pre-emptive
action through any potential asset degradation. Since November 2019,
compliance with the IEO has been overseen by a Monitoring Trustee. The
CMA will therefore maintain these arrangements in force, by introducing
similar provisions in the final undertakings or Order. This will ensure that
effective asset maintenance obligations are put in place during the
518
Merger remedies guidance, CMA87 (13 December 2018), paragraph 5.4
219
divestment process to minimise the risk of any degradation of the
divestment package.
Purchaser risk
Purchaser risk would apply if a suitable purchaser is not available or if,
following the divestment, the purchaser acted in such a way that the
competition that is expected to be substantially lessened as a result of the
merger is not restored. The effect of purchaser risk on the effectiveness of
each remedy option is discussed in the ‘identification of a suitable
purchaser’ sub-sections of the effectiveness assessment above.
[]. Evidence from third parties suggested that the pool of potential
purchasers might be limited under the options where Hunter Douglas was
to retain a 49% shareholding in 247, and possibly further limited if Hunter
Douglas were to hold limited additional rights.
We consider that divestiture risk in general, and these purchaser risks
in particular, can be partially mitigated by specifying the characteristics of
an effective divestiture process and setting out the criteria that we will apply
to establish purchaser suitability in the divestiture package. In particular, we
propose to require:
(a) CMA oversight of the divestment process, including the continued use
of a monitoring trustee;
(b) the suitability of potential purchasers to be approved by the CMA;
(c) the final divestiture proposed by the Parties, including the identity of
the purchaser, be subject to approval by the CMA;
(d) the final divestment be completed in accordance with any order
issued or undertakings accepted; and
(e) the option for the CMA to appoint a Divestiture Trustee.
[].
[].
[].
[].
220
Decision on remedies
We have decided that each of the remedy options set out in paragraph
11.9 would be an effective and proportionate remedy to address the SLC
and its resulting adverse effects we have found.
Partial divestiture of 51% of the ordinary shares of 247, with Hunter
Douglas potentially holding certain limited rights (in accordance with the
principles set out in paragraph 11.96) represents the least onerous remedy
of the effective options. [].
We have therefore decided that the following will represent an effective
and proportionate remedies package, and so require the Parties to Divest
51% of the ordinary shares of 247, with Hunter Douglas being permitted to
hold limited certain rights, in accordance with the principles set out in
paragraph 11.96 [].
[].
Furthermore, we have decided that the following is required:
(a) a restriction on agreements between Hunter Douglas and the
purchaser;
(b) CMA oversight of the divestment process, including the continued use
of a monitoring trustee;
(c) the suitability of potential purchasers to be approved by the CMA;
(d) the final divestiture proposed by the Parties, including the identity of
the purchaser, be subject to approval by the CMA;
(e) the final divestment be completed in accordance with any order
issued or undertakings accepted; and
(f) the option for the CMA to appoint a Divestiture Trustee.
The CMA has the choice of implementing any final remedy decision
either by accepting final undertakings pursuant to Section 82 of the Act if
the Parties wish to offer them, or by making a final order under Section 84
of the Act. Either the final undertakings or the final order must be
implemented within 12 weeks of publication of our final report (or extended
once by up to 6 weeks under exceptional circumstances),
519
including the
period for any formal public consultation on the draft undertakings or order
519
Section 82 (final undertakings) and Section 84 (final order) of the Act.
221
as specified in Schedule 10 of the Act. We expect to implement the
structural remedy by seeking suitable undertakings from the Parties. We
will make an Order if we are unable to obtain suitable undertakings from
the Parties.
In line with CMA guidance once this remedy has been fully
implemented in line with the conclusions set out in this decision, we have
decided that Hunter Douglas should be prohibited from subsequently
acquiring additional assets or shares of 247 or acquiring any additional
material influence or control over 247. CMA guidance states that the CMA
will normally limit this prohibition to a period of 10 years.
520
We find no
compelling reason to depart from the Guidance in this case by seeking a
shorter or longer prohibition period.
520
Merger remedies guidance, CMA87 (13 December 2018), paragraph 5.10.