Commission, January 9, 2014, No M.7023

Dear Sir/Madam,

Subject:  Case No COMP/M.7023 – PUBLICIS / OMNICOM

Commission decision pursuant to Article 6(1)(b) of Council Regulation No 139/20041

1.          On 25 November 2013, the European Commission received a notification of a proposed concentration pursuant to Article 4 of the Merger Regulation by which Publicis Groupe

S.A. ("Publicis", France) and Omnicom Group, Inc. ("Omnicom", the United States of America) enter into a full merger within the meaning of Article 3(1)(a) of the Merger Regulation by way of a stock-for-stock exchange (the "Transaction").

2.          Publicis and Omnicom are together designated hereinafter as the "Parties".

I.           THE PARTIES

3.          Publicis is a French international communications and advertising group. Publicis' agencies are organised in separate networks of offices present throughout the world and collectively provide a broad range of advertising services, including digital advertising, creative services, public affairs, corporate communications and events, media strategy, planning and buying, and specialty communications.2

4.          Omnicom is a U.S.-based global advertising, marketing and corporate communications company. Omnicom's agencies are also organised in separate networks of offices present throughout the world and collectively offer a range of advertising, marketing, media and other related services.3

II.           THE TRANSACTION

5.          By way of a Business Combination Agreement dated 27 July 2013 between Omnicom and Publicis, HoldCo, a newly-formed Dutch holding company, will successively acquire Publicis and Omnicom. First, Publicis will merge directly with HoldCo, with HoldCo continuing as the surviving legal entity. Then, Omnicom will merge with a newly formed wholly owned subsidiary of HoldCo, Merger Sub, with Omnicom continuing as the surviving legal entity and a wholly owned subsidiary of HoldCo. The Transaction has been structured so that the shareholders of Publicis and Omnicom will each hold approximately 50% of the fully diluted equity of the merged group.

6.          The Transaction therefore constitutes a concentration within the meaning of Article 3(1)(a) of the Merger Regulation.

III.        EU DIMENSION

7.          The undertakings concerned had a combined aggregate worldwide turnover of more than EUR 5 000 million4 in 2012 (Publicis: EUR 6 610 million; Omnicom: EUR 11 067 million). They both had a combined aggregate EU-wide turnover of more than EUR 250 million in 2012 (Publicis: EUR […]; Omnicom: […]) and each did not achieve more than two-thirds of its aggregate EU-wide turnover in 2012 within one and the same Member State.

8.          The Transaction therefore has an EU dimension pursuant to Article 1(2) of the Merger Regulation.

IV.        COMPETITIVE  ASSESSMENT

9.          Publicis and Omnicom are both active, through their various subsidiaries, in the advertising, marketing, communications ("AMC") services sector, including marketing communications services ("MCS") and media buying services ("MBS") in several Contracting Parties to the EEA Agreement (referred to hereinafter as "EEA countries").

10.       Publicis is active in the sale of advertising space in France and in the Netherlands. These markets are vertically related to the MBS markets, and some of their possible segments. Publicis’ activity is downstream from the activity of both Parties’ media agencies on the procurement side of MBS, in the sense that Publicis sells on behalf of cinema owners advertising space to advertisers either directly or through media agencies.5 The effects of this vertical relationship will be assessed in section IV.6 of the Decision.

11.       Regarding data analytics services, Publicis and Omnicom have marketing data analytics capabilities but they are essentially used in-house as a mean of targeting and optimising MCS campaigns or MBS in support of the provision of core MCS or MBS services.6 However, Publicis has one subsidiary that provides data analytics services to third parties on a stand-alone basis, named Ninah. Within the EEA, Ninah is active in the United Kingdom only, with revenues of around EUR […].7 Omnicom has a subsidiary called Analect, with world-wide turnover of around EUR […], of which only less than EUR […] is achieved in the United Kingdom.8 On a hypothetical data analytics market, or any of its segments, in the United Kingdom, the Parties' 2012 combined market shares would therefore have been small, and no material increment will occur as a consequence of the Transaction. Therefore, data analytics services are not assessed any further in this Decision.

IV.1.         RELEVANT  MARKETS

IV.1.1       MEDIA BUYING SERVICES

IV.1.1.1.   PRODUCT MARKET

12.       MBS include purchasing of advertising time and/or space in various types of media, such as broadcast and cable TV, newspapers and magazines, radio, billboards and the Internet, for clients running advertising campaigns.9 According to the Parties, media buying agencies will also usually provide media planning and strategic advice, including research into target audiences, which media to use, and the monitoring/tracking of the success of a campaign.

13.       In previous decisions in the advertising sector,10 the Commission has identified separate product markets for MBS and MCS. The Parties support this segmentation and their submissions are based on the distinction between MBS and MCS.

14.       A majority of respondents to the market investigation indicated that there is a distinction between MBS and MCS.11 The strategies, necessary skills, tools and overall activities are different between MBS and MCS. Advertisers use MCS to create a message and MBS to deliver that message to the target group. While a key competitive parameter in MCS is creativity, MBS depend on negotiation power and skills as well as know-how about media planning and monitoring, characteristics which are of rather technical nature. Likewise, the market investigation suggested that a majority of advertisers use different agencies for MCS and for MBS.12

15.       In WPP/Grey, the Commission considered a further segmentation of the market for MBS between: (i) the sales market, in which media buying agencies act as suppliers of MBS to final customers (advertisers); and (ii) the procurement market, in which media buying agencies buy (usually on behalf of their clients) advertising time or space in the media from media owners (for example TV broadcasters, publishing houses, radio stations, etc.).13 The Parties do not dispute this additional segmentation. Equally, none of the respondents to the market investigation opposed such a further distinction.

16.       The competitive assessment for the present Transaction will therefore be based on separate markets for MBS and MCS, on the one hand, and within the MBS market, the Commission will further distinguish between the sales and procurement of MBS, on the other hand.

IV.1.1.1.1.   Sale of media buying services Segmentation by type of service

17.       The Commission has investigated whether on the sales-side of the MBS market, an additional segmentation should be made, in particular based on the type of MBS provided by media buying agencies, such as the development of media placement plans, strategic advice surrounding media placement plans, media buying, etc. The Parties do not consider as relevant any further segmentation on the sales-side of the MBS market.

18.       In previous Commission decisions,14 no such segmentation by type of MBS was considered. Respondents to the market investigation did not provide any evidence that would justify a further distinction on the sales-side market of MBS. Even though some competitors mentioned that certain advertisers may want to keep separate the strategic planning from the buying and split them between different agencies, the majority of competitors indicated that the most common offered service includes all MBS.15 Likewise, the majority of customers confirmed that they generally purchase these services together as the combined scope of services is more efficient and allows for a more coherent media strategy as well as for optimised costs.16

19.       In light of the above, the Commission considers that no additional segmentation by type of service within the sales-side of the MBS market should be made for the purpose of assessing the present Transaction.

Segmentation by type of media

20.       The Commission considered in previous decisions that the market for the sale of MBS should not be further segmented according to the media in question, because media buying agencies are not specialised in one media segment, but normally carry out MBS for all media channels. Additionally, most advertisers buy multiple types of media and want central coordination and oversight of their media planning.17

21.       The Parties submit that this conclusion is still valid and that the sales-side of the MBS market should not be further segmented by type of media, in particular for the following reasons. First, from a demand-side perspective, most advertisers buy multiple media types as they need to interact on a cross-channel basis with consumers whose media consumption is increasingly diverse and fragmented. This feature of MBS is also evidenced by the fact that an overwhelming majority of MBS pitches require media agencies to buy on all types of media channels. Second, from a supply-side perspective, with the exception of the agencies that only buy digital advertising space, media buying agencies are not compartmentalised across media channels. Third, the Parties consider that the growth of the Internet has neither changed the nature of demand for MBS, nor the manner in which these services are supplied and, thus, does not warrant segmentation by type of media in the sales-side of MBS or for digital services in particular.

22.       Respondents to the market investigation indicated that it would be appropriate not to segment by type of media within the sales-side of the MBS market. While some MBS may focus on a specific media channel such as the Internet, the majority of media buying agencies do not usually specialise in selling MBS to advertisers for specific types of media.18 A majority of customers also suggested that they usually purchase a range of MBS covering more than one media channel,19 which allow them to reach their final consumer target, and that they usually look for MBS suppliers that are able to provide services with respect to all types of media.20 This is a separate question to whether for an advertiser the placement of an online advertisement is substitutable with the placement of an advertisement on other offline media channels. This question does not need to be answered in the present Decision since advertisers will typically seek to buy from an MBS agency a full package of complementary services covering several (online and offline) media channels and including advice on how best to distribute advertisement space over these different media channels. MBS agencies offer such packaged solutions.

23.       In light of the above, the Commission considers that no additional segmentation by type of media within the sales-side of the MBS market should be made for the purpose of assessing the present Transaction.

Segmentation by type of sector

24.       In previous Commission decisions, no further segmentation by type of sector within the sales-side of the MBS market was considered. The Parties do not consider relevant any further segmentation on the sales-side of the MBS market.

25.       Respondents to the market investigation suggested that no such segmentation by type of sector should be made. Even though certain sectors are governed by rules that may require specific knowledge by the MBS supplier, such as healthcare or alcoholic beverages, the majority of respondents to the market investigation did not consider that such sectors have specific requirements that only certain media buying agencies can address.21

26.       In light of the above, the Commission considers that no additional segmentation by type of sector within the sales-side of the MBS market should be made for the purpose of assessing the present Transaction.

Segmentation by size of account

27.       The Commission has previously considered whether the sales-side of the MBS market, MBS should be further segmented into: (i) large-scale MBS for large advertisers, that is to say advertisers that pursue mostly extensive and costly nation-wide advertising campaigns ("large-scale media buying") and (ii) small-scale MBS for small advertisers pursuing predominantly local/regional advertising campaigns ("small-scale media buying").22 However, the Commission did not retain this possible segmentation except in Germany, where only large agencies belonging to one of the international networks were viewed as being capable of providing MBS to large-scale advertisers.23

28.       The Parties contest any possible segmentation, including in Germany, on the basis that the same group of media buying agencies compete for "large-scale" and "small-scale" accounts. The Parties explain that, on the one hand, independent agencies compete with international groups for large accounts and, on the other hand, international groups do not only compete for large accounts, but also for smaller ones.

29.       Responses to the market investigation differed with regard to a possible segmentation by size of account within the sales-side of the MBS market. On the one hand, a majority of competitors considered that not every media buying agency active in MBS can be selected by large-scale advertisers, as larger MBS agencies may be able to offer larger scale services that smaller agencies may not.24 On the other hand, competitors indicated, in the majority of cases, that their respective media buying agencies can be selected by any advertiser, no matter its size.25 Furthermore, a majority of customers pointed to the fact that large-scale advertisers require MBS agencies that have large-scale capabilities, a broad geographic presence, and are able to build up the skills and knowledge, and manage the complexities of the required services.26

30.       In any event, for the purposes of the present Decision, it is not necessary to conclude on whether the market for the sales of MBS should be further segmented between small- scale and large-scale media buying as the Transaction does not raise serious doubts even considering that large-scale advertisers may have specific requirements.

Segmentation depending on the inclusion/exclusion of direct sales

31.       The Commission also investigated whether direct sales are part of the same market as MBS and whether an additional segmentation should be considered. The direct sales represent the media purchased directly by advertisers from media owners rather than through a media buying agency

32.       The Parties see no basis to exclude direct sales from the market size and disagree with any such exclusion because direct purchasing is a competitive constraint on media buying agencies. The higher the level of direct buying in a given country, the more media agencies are constrained by the threat of advertisers bypassing their services and dealing directly with media owners, and, in all events, major advertisers especially have as good an opportunity as media buying agencies to acquire the human talent necessary to satisfy their MBS needs

33.       Responses to the market investigation differed regarding such an additional segmentation. In particular, the replies of customers were mixed regarding whether they would switch to buying directly from media owners if prices of MBS agencies were to increase by 5-10% post-Transaction.28 However, a majority of media owners explained that they sell advertising time and/or space directly to advertisers.29

34.       Moreover, the Commission considers that in two EEA countries where the Parties' activities overlap (Sweden and Denmark), direct sales represent a high percentage of total media buying sales of media owners (on average 50-55% and 55-60% respectively), and would likely exercise a significant competition constraint on advertisers. In the other countries where the Parties' activities overlap, the ratio of direct sales against total media buying sales is significantly lower.

35.       In any event, for the purposes of the present Decision, it is not necessary to conclude on whether direct sales belong to the same market for MBS in some or all EEA countries where the Parties' activities overlap, except in Sweden and Denmark, as the Transaction does not raise serious doubts under any alternative product market definition. The extent to which the threat of actual or prospective advertisers opting to bypass the services of a media buying agency may act as a competitive constraint on the media buying agencies will be assessed in the competitive analysis. For Sweden and Denmark however, direct sales, given their magnitude, have to be included in the MBS markets.

Conclusion

36.       In light of the above, the Commission’s competitive assessment will be based on a broader market for the sale of MBS encompassing all types of MBS services, all media channels, all types of sectors, and will also take into consideration the specific requirements of small-scale versus large-scale media buying, as well as including and excluding direct sales depending on the countries.

IV.1.1.1.2.   Procurement of media buying services Segmentation by type of media

37.       In past decisions, the Commission has considered30 a segmentation of MBS on the procurement-side of the MBS market based on different types of media and concluded that different procurement markets for TV and print should be defined, due to the general lack of demand substitutability between different types of media and the fact that media owners could not switch from one type of media to the other without substantial investment and know-how. Additionally, the Commission left open the question whether a further segmentation is necessary within the procurement market for a specific type of media.31

38.       While the Parties generally concur with the Commission's previous findings, they consider, however, with respect to the Internet channel which has considerably developed since WPP/Grey, that from the point of view of advertisers, there is a certain degree of substitutability between the procurement of advertising space online and in other media (in particular print). Moreover, the Parties argue that as newspaper and magazine publishers can easily launch digital editions of their publications, there is also a certain degree of supply-side substitution between the two. In any event, the Parties submit that for the purpose of assessing the present Transaction it can be left open whether the procurement of online advertising space constitutes a separate market. Finally, the Parties reject any further segmentation within the procurement markets between TV, print and online/mobile.

39.       Responses to the market investigation differed with regard to a possible segmentation by type of media within the procurement-side of the MBS market. On the one hand, a majority of media owners considered that the traditional delineation between various types of media is becoming increasingly blurred, because media owners compete with other media owners in order to provide the best product for the advertisers' campaigns and because media channels are to some extent substitutable.32

40.       On the other hand, a majority of customers indicated that the different media segments are in general not substitutable with respect to the type of advertising they offer.33 The degree of flexibility across media channels depends to a large extent on the individual advertiser and the specific target groups to be reached. All types of media have specific characteristics in relation to the effects of advertisement. In that regard, a majority of customers and competitors that responded to the market investigation indicated that depending on the needs and media strategy of advertisers, there may be one or several "must-have" media channels to promote advertisers' products or services, which are not substitutable with other media channels.34

41.       In any event, for the purposes of the present Decision, it is not necessary to conclude on whether the procurement side of the MBS market should be further segmented by type of media as the Transaction does not raise serious doubts under any alternative product market definition.

Segmentation by size of account

42.       The Commission has previously considered a further segmentation of the different MBS procurement media segments into small-scale and large-scale clients, but concluded that such distinction would be artificial.35 The Parties agree with the Commission’s past approach and support the lack of segmentation by size of client.

43.       In the present case, respondents to the market investigation did not bring forward any conclusive evidence indicating that such segmentation should be made.

44.       In light of the above, the Commission considers that no additional segmentation by size of account within the procurement-side of the MBS market should be made in order to assess the present Transaction.

Segmentation depending on the inclusion/exclusion of direct sales

45.       The Commission has also investigated whether direct sales should be included on the procurement side of the MBS market and whether an additional segmentation should be considered.36

46.       The Parties disagree to any exclusion of direct sales. They consider that direct purchasing by advertisers represents a competitive constraint on media buying agencies.

47.       Responses to the market investigation differed regarding such an additional segmentation. In particular, the replies of customers were mixed regarding whether they would switch to buying directly from media owners if prices of MBS agencies were to increase by 5-10% post-Transaction.37 However, a majority of media owners explained that they sell advertising time and/or space directly to advertisers.38

48.       For the purposes of the present Decision, it is not necessary to conclude on whether direct sales belong to the same market for MBS on all EEA countries where the Parties' activities overlap (except for Sweden and Denmark, given their important size of direct sales in these two countries), as the Transaction does not raise serious doubts under any alternative product market definition. The extent to which the threat of actual or prospective advertisers opting to bypass the services of a MBS agency may act as a competitive constraint on the media buying agencies will be assessed in the competitive analysis.

49.       For Sweden and Denmark, direct sales, given their magnitude, have to be included in the MBS markets for all media channels, except for TV and outdoor in Sweden, and TV in Denmark. This is because, different from the average percentage at country level,   the level of direct sales is low for TV (6%) and outdoor (16%) in Sweden, and for TV (1%) in Denmark.39

Conclusion

50.       In light of the above, the Commission’s assessment of the Transaction will be based on the market for procurement of MBS segmented along different types of media, as well as including and excluding direct sales depending on the countries and media channels.

IV.1.1.2.   GEOGRAPHIC MARKET

IV.1.1.2.1.   Sale of media buying services

51.       The Commission considered in previous cases40 that the market for the sale of MBS were national in scope. This conclusion was based inter alia on the fact that media buying agencies need to gather locally the necessary knowledge of customer patterns with regard to different media channels and of the differing national regulatory frameworks. In addition, the Commission also noted the tendency of the national markets for the sale of MBS to become wider in scope, at least with regard to multinational firms advertising in different countries. However, it was concluded at that time that the tendency was not developed enough to justify wider markets, in particular due to the limited number of multinational advertisers having a global strategy for media buying as well as acknowledgement by those multinational advertisers of the necessity for local presence.41

52.       The Parties concur with the Commission's previous conclusions and submit that the necessity of a local presence remains a key element for the sale of MBS. First, it is essential to master the national language when negotiating with media buyers, who are mostly organised on a national basis. Second, local knowledge is needed to adapt an advertiser’s media strategy to differing media consumption trends and habits. Third, even though only high-level strategies and ideas may be defined on an international basis, the most important part of the services offered by MBS agencies is still performed on a national basis.

53.       In addition, the Parties reject any additional segmentation of MBS into markets that would be narrower than national, as most campaigns are national in scope, and most media channels cover the whole territory of a country

54.       Responses to the market investigation differed with regard to the geographic scope of the sales-side of the market for MBS. On the one hand, a majority of customers indicated that, in many cases, their contracts with MBS suppliers have in a large proportion a national dimension.42 On the other hand, a majority of competitors considered that many advertisers have a global media strategy and customise the experience for geographic and national considerations. Moreover, a number of competitors noted the rise of digital media and the appearance of large global digital groups, which allows global or regional negotiations and buying.43

55.       In any event, for the purposes of the present Decision, it is not necessary to conclude on whether the sales side of the market for MBS should be national or wider than national (EEA-wide) because the Transaction does not raise serious doubts under any alternative geographic market definition.

IV.1.1.2.2.   Procurement of media buying services

56.       In WPP/Grey, the Commission considered that the procurement market for MBS was national in scope, notably due to the fact that as most advertisers and media buying agencies need to advertise in the national media, they require their media buying agencies to purchase time or space with national TV broadcasters or publishers.44 Furthermore, the Commission considered in Google/DoubleClick that the market for online advertising space could be EEA-wide in scope from a technical point of view, but concluded that this market is to be considered as segmented alongside national or linguistic borders within the EEA.45

57.       The Parties do not contest such an approach and submit that advertising is directed to the area where media owners have their main audience. As media advertising is intrinsically linked to media broadcasting, and media broadcasting is national, negotiations for media owners’ advertising inventory take place at national level.

58.       The view of the Parties was supported by the market investigation, as the majority of respondents indicated that the market for procurement of advertising time and/or space is national in scope.46

59.       In light of the above, for the purposes of the present Decision, the Commission considers that the procurement side of the MBS market, segmented along different types of media, in each of the EEA countries where the Parties operate, is national.

IV.1.2       MARKETING AND COMMUNICATION SERVICES

IV.1.2.1.   PRODUCT MARKET

60.       In WPP/Grey, the Commission considered that the relevant product market for the provision of MCS encompasses an array of disciplines including advertising, insight and consultancy, public relations, consumer relationship management/direct marketing/event management, brand identity and design and other areas of specialist communications.47

Segmentation by type of service

61.       The Commission considered in past decisions that all MCS belonged to the same relevant market and that a further segmentation would not have reflected the way in which these disciplines were demanded and offered.48 The Commission also considered that the types of MCS represented different methods for a business to communicate with a group of people, be they consumers/customers, the press, industry, other companies, government and other regulatory bodies and that all these disciplines were substitutable to a sufficient extent and that most agencies were able to offer all type of disciplines.

62.       The Parties submit that the Commission's previous conclusions should be upheld.49 First, both large international network and country-specific independent agencies tend to offer the full range of disciplines to clients to create and deliver a coherent and consistent brand message across different media channels. Few agencies choose to specialise in one particular MCS discipline and even those that do are likely in practice to provide more than one MCS discipline.

63.       Second, although customers may purchase single disciplines from different providers for specific projects, they expect and demand that agencies provide services across all or most types of MCS disciplines.

64.       Third, segmentation by type of MCS discipline is meaningless in light of the significant blurring of the lines between different disciplines brought about in recent years by the increased use of digital resources.50

65.       In general, respondents to the market investigation indicated that the different MCS disciplines are substitutable both from the supply and demand side perspective.51

66.       Although small independent agencies may specialise in a given MCS discipline, a majority of respondents to the market investigation pointed out that a majority of agencies aim to provide an integrated offer to customers. Likewise, a majority of customers indicated that they purchase a variety of MCS disciplines to achieve the same goal of communicating with a group of people.

67.       The Commission is of the view that MCS disciplines may be purchased on a stand-alone basis for a given project or as part of a wider strategy or campaign. Furthermore, customers may instruct either individual agencies or a number of agencies. This indicates that different MCS discipline and types of agencies could be used to design a creative message for a campaign and that fully-fledged and more specialised agencies compete with each other for the provision of these disciplines.

68.       Even if one competitor that replied to the market investigation referred to a possible separation of public relation ("PR") from other MCS disciplines, no one else raised the issue regarding the need to consider a segmentation for PR, as such disciplines are not usually provided on a stand-alone basis, but rather are offered by large agencies alongside a wide range of other MCS disciplines.52

69.       In light of the above, the Commission concludes that all MCS disciplines are part of the same market, since the different types of MCS are substitutable to a sufficient extent, and most agencies offer all or most MCS disciplines to their customers.

Segmentation by type of media

70.       The Commission has previously left open whether it would be appropriate to segment MCS by media channel.53

71.       The Parties consider that a further segmentation by media channel (including the Internet) would be artificial. First, from a demand-side perspective, advertisers purchase MCS irrespective of the media channel where the message or the advertising campaign will be placed, as they are interested in targeting a broad audience and in conveying a coherent and consistent message across all media.

72.       Second, from a supply-side point of view, agencies create messages and insights with a multi-channel approach, without focusing on a specific type of media.

73.       Third, traditional advertising can easily be adapted for on-line publication. In that respect, there are no barriers to entry for a traditional agency to create advertising for online distribution. Pure digital MCS agencies compete thus with traditional players.

74.       A majority of respondents to the market investigation indicated that MCS are provided through all types of media, including the Internet. In particular, several respondents indicated that agencies are expected to deliver consistency and recognise interconnectivity between different channels and thus are not requested to provide aservice tailored according to media type. In addition, media channels are seen as different ways to convey a message and to reach different targeted audiences.54 This question of whether MCS should be segmented by media channel is a different one to whether for an advertiser the placement of an online advertisement is substitutable with the placement of an advertisement on an offline media channel. This latter question does not need to be answered in the present Decision.

75.       In light of the above, the Commission concludes that no additional segmentation of the MCS market by type of media is necessary when assessing the present Transaction.

Segmentation by type of sector

76.       In past decisions55, the Commission has investigated whether a separate market for the provision of specialist communications advice, such as to the healthcare sector, exists.

77.       The Parties acknowledge that while certain sectors present specificities, these are insufficient to justify a further segmentation of the market. The Parties refer in support of their submission to the healthcare sector. On the one hand, the communication of messages about drugs and medical products is strictly limited as regards the content and the audience that can be targeted, with the result that agencies prefer to organise their MCS activities in the healthcare sector separately. On the other hand, the MCS skills and tools in this sector are the same as those necessary to carry out MCS in non- regulated sectors.

78.       While a minority of respondents to the market investigation pointed out that certain sectors, such as healthcare, requires specialised MCS agencies, a majority of respondents considered that their sectors of activity do not require specific knowledge, expertise, skills and tools that only certain MCS agencies can address.56

79.       In light of the above, the Commission concludes that no additional segmentation of the MCS market by type of sector is necessary when assessing the present Transaction.

Segmentation by size of account

80.       In past decisions57, the Commission left open whether large-scale customers that develop international budget and campaigns constitute a distinct market. On the one hand, the Commission noted that the supply of international MCS may require specific implementation needs that are not necessary for the provision of MCS at national level and that can be successfully offered only by large-scale MCS agencies. On the other hand, the evidence the Commission gathered was insufficient to support the definition of a distinct market for the provision of international MCS.

81.       The issue was not meaningfully addressed by the Parties in their submissions.

82.       Even though certain respondents pointed out that only large-scale agencies may be appointed by multi-national customers to deliver global campaigns, as they are the only ones able to offer the required scale of resources and network to ensure implementation of the campaign across different countries, a majority of respondents to the market investigation indicated that the supply of international MCS should not be considered as a distinct market. In particular, these respondents indicated that any agency can be selected by large-scale customers as advertising services mostly depend on creativity which can also be provided by a small agency.58

83.       In light of the above, the Commission concludes that no additional segmentation of the MCS market by size of account is necessary when assessing the present Transaction. The specific needs of large, multinational, customers will, however, be taken into account in the assessment of the Transaction.

IV.1.2.2.   GEOGRAPHIC MARKET

84.       In WPP/ Grey, the Commission considered that the market for MCS was national in scope, even with regard to large, multinational, customers, because of language differences, different media conditions, pricing differences between countries and the need to inform the public, government or other institutions and therefore plan a campaign on a national basis.59

85.       The Parties submit that the Commission's conclusions in WPP/Grey are still valid, even with reference to multinational customers, as campaigns continue to be designed mostly on a national basis.

86.       Even if a majority of the respondents to the market investigation did not express a view on whether the conclusions drawn in WPP/Grey are still valid today, a majority of the customers that expressed their opinion replied affirmatively, whereas answers were more nuanced among competitors. In particular, competitors acknowledged that MCS contracts are mostly national in scope, while referring to a tendency towards globalisation, especially in relation to customers with an international footprint. However, a majority of the respondents to the market investigation stated that MCS contracts have mostly national coverage.60

87.       Given the still high percentage of national coverage of MCS contracts, the Commission takes the view that the conclusion drawn in WPP/Grey are still valid, and the MCS markets are national in scope.

Conclusion

88.       In light of the above, the competitive assessment of the Transaction will be based on the provision of MCS in each EEA country where the Parties are present.

IV.1.3       SALE OF ADVERTISING SPACE IN CINEMAS

IV.1.3.1.   PRODUCT MARKET

89.       The Commission has so far never considered whether the sale of cinema advertising constitutes a separate market or should be considered as part of the global market for the sale of advertising space in all media.

90.       However, in past decisions relating to other media channels, the Commission has considered that the supply of advertising space and time in certain media channels, such as TV, outdoor and radio, could be considered as a separate product market. In particular, the Commission has considered as a separate market the market of advertising space on TV, while it has left open the exact market definition in relation to the provision of advertising space for outdoor and radio.61

91.       The Parties consider that the exact scope of the relevant product market can be left open in this case, as the Transaction does not raise any serious doubts under any possible market definition.

92.       For the purpose of the present Decision, the Commission agrees that the exact scope of the relevant product market can be left open because the Transaction does not raise serious doubts under any possible market definition.

IV.1.3.2.   GEOGRAPHIC MARKET

93.       The Commission has considered in previous decisions62 that the procurement side of the MBS market, which is a downstream market of the market for the sale of advertising space or time, is national in scope, as most advertisers and media buying agencies need to advertise in the national media.

94.       The Parties have not expressed any specific view on the possible geographic scope of the relevant market beyond the statements it has made regarding the national dimension of procurement markets, where negotiations for the advertising inventory of media owners take place at national level (see Section IV.1.1.2.2).

95.       In light of the above, the Commission concludes that the markets for the sale of advertising space in cinema are also national in scope.

IV.2.     METHODOLOGIES FOR ESTIMATING MARKET SHARES FOR MEDIA BUYING SERVICES

96.       The Commission will first present the methodologies proposed by the Parties on of the sales and procurement sides for estimating the total MBS market size and the Parties' and their competitor's market shares (Section IV.2.1). The Commission will then conduct its own assessment of the Parties' proposed methodologies (Section IV.2.2).

IV.2.1      METHODOLOGIES PROPOSED BY THE PARTIES

IV.2.1.1.   SALES SIDE

IV.2.1.1.1.   Estimates of market size and market shares of the Parties

97.       The Parties have provided estimated market sizes for the MBS markets in the EEA countries where the activities of the Parties overlap from industry reports that cover these countries, such as the ZenithOptimedia Advertising Expenditure Forecasts for 2012 released in June (the "ZenithOptimedia report")63, the World Advertising Research Center 2012 Reports released on 9 September 2013 (the "WARC report")64, the Group M report "This Year, Next Year"65 (the "Group M report") and other national reports.

98.       As to the calculation of their own market shares, the Parties have primarily used their own revenues and the market size estimated in the ZenithOptimedia and the WARC reports. According to the Parties, these two reports are reliable benchmarks for total market size estimates, and the market shares estimated using either source are broadly consistent – see sub-section (a) below. In addition, the other sources which the Parties have used will be described in sub-section (b) below.

(a)   The ZenithOptimedia report and the WARC report

99.       The ZenithOptimedia report includes annual reports estimating advertising expenditure by country and is produced by ZenithOptimedia, which belongs to the Publicis Group. The ZenithOptimedia report excludes agency commissions and seeks to estimate the net cost of media purchasing.66 It relies both on surveys of advertisers, agencies and media owners, and on data containing advertising volumes and advertising rate cards.

100.   The Parties have used the WARC report as an alternative source for estimating the total market size. WARC collects data on advertising expenditure in the same way as ZenithOptimedia, primarily through an annual survey.

101.   The main difference between the two methodologies used for producing the two reports is that the WARC report seeks to incorporate agency commissions and fees whereas the ZenithOptimedia report seeks to exclude them.

102.   For the purpose of estimating their own market shares, the Parties have used their own revenues in two different manners for consistency with the methodologies used by the ZenithOptimedia report and the WARC report respectively: (i) when using the ZenithOptimedia market size, they divided their media purchase costs by the ZenithOptimedia report market size (that is to say they excluded their own fees and commissions from their total revenues); and (ii) when using the WARC report market size, they divided their total revenues by the WARC report market size (that is to   say they included their agency fees and commissions).67 The Parties did so in order to allow the Commission to verify whether the inclusion of agency commissions and fees would materially change the Parties’ market share estimates.

103.   The Parties submit that their own market shares using media purchase costs and the total market size from the ZenithOptimedia report are close to their shares using total revenues and the total market size from the WARC report. The market size estimates in the WARC report are similar to the market size estimates in the ZenithOptimedia report, although typically slightly higher, which is likely because of the inclusion of agency commissions and fees. The Parties further submit that no adjustment to the WARC report estimated market size is necessary as was done at the time of the WPP/Grey decision68  which mentioned that the WARC report’s estimates of the total size of   the market were based on rate cards, and did not take into account discounts granted by media vendors. This implied that, at that time, the WARC report data over-estimated the total size of the market and thus needed to be scaled down to account for discounts granted by media vendors. The Parties explain that today, the WARC report as used by the Parties indicates that, for all EEA countries, data taking into account the discounts. Therefore, the WARC report is sufficiently reliable as such and there is no need to“scale down” WARC-produced  data.69

(b)        Other sources for estimating the market size and the Parties' market shares

104.   The Parties have also provided other sources for estimating the MBS market size in the overlap EEA countries

105.   In particular, the Parties have provided the Group M report which estimates media expenditures per country. Group M is WPP's media buying organisation. In addition, the Parties have provided advertising expenditure estimations by Nielsen in all EEA countries where, to the Parties’ best knowledge, such estimations, exist, albeit even partially.70 In certain EEA countries, the Parties have provided the total market size as estimated by national local sources such as IREP Communication and Kantar in France or Infoadex in Spain.

106.   The Parties have provided the market size as estimated in the reports of the Research Company Evaluating the Media Agency Industry (RECMA), namely the Overall Activity Ranking reports (the "RECMA reports").71

107.   As regards the calculation of their own market shares, the Parties have provided three additional methodologies: (i) the Parties have divided their media purchase costs by the total market size calculated in the Group M report72; (ii) they have calculated their own market shares by using their actual revenues and the total market size as estimated in the RECMA reports; and (iii) they have provided the market shares as estimated in the RECMA reports (that is to say without calculating their shares on the basis of their real revenues).

108.   However, according to the Parties, the RECMA reports suffer from several important shortcomings. First, the RECMA reports focus only on larger agencies and do not take into account direct sales, thus excluding part of demand, which introduces a downwards bias in the RECMA reports' total demand. Second, as of its 2013 reports, RECMA no longer reports buying billings; instead, it reports a measure it refers to as "Overall Activity". This measure includes several categories of services in addition to media buying and planning, including digital services, diversified services and international account coordination. This introduces an upwards bias into the Overall Activity when considered as a measure of MBS demand size. Third, the Parties know from their actual data that the RECMA reports considerably overstate the Parties’ own activity in many EEA countries. Fourth, in Bulgaria, Estonia, Finland, Ireland and Slovakia, RECMA allocates billings to the Parties from entities that are not controlled by them. Such overestimation is likely to affect all agencies covered by the RECMA reports but the Parties do not know the degree of the overestimation of their competitors’ activity. Fifth, the RECMA reports are only used by the industry to assess relative rankings of agencies rather than demand size or market shares.

109.   The Parties believe that all these shortcomings prevent the RECMA reports from being a reliable estimate of total market size and of their own market shares.

IV.2.1.1.2.   Competitors' market share estimates

110.   The Parties submit that neither the ZenithOptimedia report, nor the WARC report provide data broken down by competitor, so these sources cannot be used to estimate market shares of their competitors

111.   The Parties submit that the RECMA reports are the only data source that provide billings estimates for large international groups, and provide the Parties with at least some means to estimate the shares of competitors. While some information on competitor rankings can be deduced from the RECMA reports73, the Parties explain that there are several important shortcomings with the RECMA reports. First, the RECMA reports focus on large global groups and ignore most independent media buyers. Second, RECMA limits research to large accounts (above a certain threshold which varies by country, typically EUR 1 000 000). Third, the RECMA reports do not include direct media purchasing performed in-house by advertisers. Fourth, RECMA relies on data and assumptions which, according to the Parties, are not robust – for example, it assumes that purchasers pay full advertising rate card rates to build up gross billings activity and does not adequately adjust for the discounts which are achieved in reality. As a result, RECMA grossly overestimates the billings and shares of the major global networks.74 Such overestimation of global network activity comes at the expense of the many independent media buying agencies that are not accounted for in the RECMA reports and which compete vigorously with the Parties for major MBS accounts.

112.   To adjust for the fact that the RECMA reports overstate the gross billing activities of international groups, the Parties attempted to estimate the size of media buying activity that was not accounted for in the RECMA reports, which includes independent media buyers and direct media purchases. The remainder of the ZenithOptimedia-estimated market size was allocated among the competitors tracked by the RECMA reports in the same proportion to one another as in the RECMA reports.

IV.2.1.2.   PROCUREMENT SIDE

113.   The Parties have provided estimated shares of media purchases by media channel following the same methodology as for MBS sales, using their own media purchases by channel and the market size estimated in the ZenithOptimedia report, which is also available by media channel. In addition, the Parties have also provided estimates of the share of the Internet channel within the 2012 total advertising expenditures at the EEA- level as provided by the ZenithOptimedia report.

114.   The Parties have not estimated their combined market shares on each media channel excluding direct sales, except for Denmark and Sweden, due to the specifics of these countries (see Section IV.2.2.2.).

115.   The Parties have been unable to provide market shares of competitors on each media channel. They explained that the only relevant data source that provides billing estimates for competitors is the RECMA reports which however, do not provide any information by media channel. The Parties have further explained that they do not know the actual MBS revenues, media purchase costs and billings of their competitors and that therefore they have no data to split these by media channel.

IV.2.2      COMMISSION'S ASSESSMENT

IV.2.2.1.   SALES   SIDE:  MARKET   SIZE   AND   MARKET   SHARES   (FOR THE

PARTIES AND THEIR COMPETITORS)

116.   The Commission has conducted an in-depth assessment of the methodologies used by the Parties to assess their own market shares and to the extent possible the market shares of their competitors. The Commission engaged with the main MBS competitors of the Parties in order to test the methodologies relied on by the Parties. In addition, the Commission has sought to obtain MBS revenue information from the Parties' competitors.

(a)   Market size and market shares of the Parties

117.   The main competitors of the Parties were first asked whether the methodologies used by the Parties to calculate the Parties' combined market shares (using the ZenithOptimedia report and the WARC report) was appropriate and, if not, whether they could suggest alternative methodologies to the ones proposed by the Parties.

118.   The main competitors of the Parties considered that the Parties' proposed methodologies for estimating the total market size were generally reasonable. Competitors also cited other sources for estimating the total market size, which the Parties provided to the Commission (see Section IV.2.1 above).

119.   The Commission notes that when comparing the total market sizes estimated in the various sources provided by the Parties75, in most countries, the smallest market size is estimated in the ZenithOptimedia report, the WARC report or the Group M report. For the EEA countries in which this is not the case, the difference in the estimated market size does not lead to any material differences in the combined market shares of the Parties.76

120.   As a result, the Commission considers that a comparison of the Parties' combined market shares on the basis of these three sources appears reasonable. Table 1 below provides the overview of the Parties' combined market shares on the MBS sales market based on the market size as estimated in each of the ZenithOptimedia, WARC, and Group M reports 

Table 1: Overview of Parties’ combined share on the MBS market

Source: Form CO, Annex 38 (updated for the United Kingdom). All MBS share are based on the latest reports available (ZenithOptimedia, WARC and Group M). The Parties are not aware of any events that would materially alter their estimates after the publication of these reports.

121.   As can be seen in Table 1, Austria, Croatia, Denmark, Germany, Greece, Ireland, the Netherlands and Sweden are not affected markets under any of the three sources on the sales side of the MBS market.

122.   Table 1 further shows that if estimated by using the market size in the ZenithOptimedia report in all EEA countries where the activities of the Parties overlap except for Slovenia, the estimated combined market shares of the Parties are either at their highest or not significantly different from the other sources. While the difference in the total market size estimations is also rather high in Croatia, the Transaction does not lead to an affected market under any of the three sources. The Commission therefore considers that it is appropriate to conduct its competitive assessment using market shares of the Parties calculated on the basis of the ZenithOptimedia report in all EEA countries except for Slovenia.

123.   In Slovenia, the ZenithOptimedia report estimates of the market size are significantly higher than the other two reports, WARC and Group M, leading to an affected market under one methodology. As the Group M report excludes certain types of media and therefore underestimates the total market size, the Commission will consider the WARC report for the purpose of assessing the competitive effects on the sales side market for media buying services in Slovenia.

124.   The Parties also provided information regarding their combined market shares if direct sales were excluded from the total market size (that is to say media sold directly by media vendors to advertisers, rather than through a media buying agency). Table 2a below shows the Parties' estimates of the level of direct buying in each EEA   country. ;Using these estimated shares of direct sales, Table 2b shows the Parties market shares excluding direct sales from the total market size80:

Table 2a: Parties’ estimates of level of direct sales per EEA country

             Source: Paragraph 346 of the Form CO.

Table 2b: Parties’ MBS share estimates excluding direct sales

Source: ZenithOptimedia market size - Annex 14/R30 for all EEA countries except for Croatia and Slovenia for which WARC market size- Annex R25.

125.   As can be seen from Table 2b, Austria, Croatia, Germany, Greece, Ireland, the Netherlands and Slovenia are not affected markets even if direct sales are excluded from the MBS market.

126.   According to the Parties, excluding direct sales would artificially increase the Parties’ market shares. However, as explained in paragraphs 48 and 49 above, it is not necessary to decide whether direct sales belong to the same market for MBS in all EEA countries where the activities of the Parties overlap (except for Sweden and Denmark, where they have to be included for reasons specific to these two countries), as the Transaction does not raise any serious doubts under any alternative product market definition.

127.   Therefore, the Commission will conduct its competitive assessment on a market including and excluding direct sales and will consider the competitive constraint exercised by direct sales in each country.

(b)  Market shares of competitors

128.   The Parties' main competitors were also asked whether the Parties' methodology for estimating the market shares of their competitors (using the relative rankings in the RECMA reports) was appropriate, and, if not whether they could suggest an alternative methodology to the one proposed by the Parties.

129.   The Parties' main competitors consulted on the methodology issue responded that they considered the approach proposed by the Parties to be a suitable starting point for an analysis of market shares on the sale side for the MBS market. Certain of these competitors also stated that they use the market shares estimated in the RECMA reports in the course of their business to gain a better understanding of the rankings of the various agencies on the markets. However, certain of these competitors also raised certain shortcomings in the RECMA reports as described by the Parties

130.   Table 3 below provides an overview of the Parties' estimates of the market shares of competitors:

Table 3: MBS shares of Parties and competitors, 2012 (%)81

Source: Table 43 of the Form CO, as amended for the United Kingdom in Annex R30. For Croatia and Slovenia, the Parties' combined market shares are calculated on the basis of the WARC report

131.   The Commission obtained revenue information during the market investigation from certain of the Parties’ competitors (including from certain independent agencies). The Commission compared the actual revenues of the Parties and of these competitors on the overlap  EEA  countries  with  the  revenues  estimated  in  the  RECMA  reports. The Commission found that the RECMA reports may heavily over- or underestimate the revenues of both the Parties and their competitors. Therefore, the Commission does not consider that the market size or the market shares as estimated in the RECMA reports are a sufficiently reliable source for the purpose of the present Decision in all EEA countries where the activities of the Parties overlap.

132.   Furthermore, the Commission compared the market shares of the competitors as estimated by the Parties with the market shares these competitors would have if considering the market size estimated in the ZenithOptimedia report. The Commission found that for some EEA countries, the relative rankings of the Parties versus their competitors is different from the Parties' estimates in Table 3 above in that the Parties' underestimate their own shares compared to their competitors or overestimate their competitors' shares. This can be explained by the fact that the Parties used the relative rankings in the RECMA reports for estimating their competitors' relative positions, which, as explained above, is not sufficiently reliable. However, despite the existence of such differences in the relative rankings, this does not change the Commission's conclusions with regard to the competitive assessment that it has conducted on all relevant markets (see Section IV.4.1).

133.   Considering all the above, and the other available evidence, after having analysed the revenue data of certain of the Parties' competitors insofar as such information was available and of the Parties independently, the Commission considers that the methodology used by the Parties, relying on the ZenithOptimedia report and using actual media cost for calculating their own individual and combined market shares, is sufficiently reliable82 as a starting point for the purpose of assessing the Transaction.

134.   As to the market shares of the Parties’ competitors, the Commission notes that there are some discrepancies in the relative rankings; however despite these discrepancies, the Commission's conclusions as set out in Section IV.4.1 remain materially the same.

IV.2.2.2.   PROCUREMENT SIDE

135.   The Commission considers that the methodology provided by the Parties for estimating their combined market shares on the procurement side on each media channel is sufficiently reliable. The Parties relied on the ZenithOptimedia report for providing the split by media channel and calculated their own market shares by using their actual media purchase cost data. As explained in Section IV.2.2, the Commission considers that the ZenithOptimedia report is sufficiently reliable for the purpose of the present Decision.

136.   Table 4 presents the Parties’ combined market shares by media channel on the procurement side of the MBS markets:

Table 4: Parties’ combined MBS share estimates by media channel

Source: Table 46 to the Form CO as amended in Annex R27 for all EEA countries except the United Kingdom and Slovenia. Annex R30 for the United Kingdom and the WARC report for Croatia and Slovenia.

137.   As mentioned above, for Slovenia, the Commission considers that the market size estimated by the ZenithOptimedia report is not sufficiently reliable. The Commission therefore uses the alternative WARC report to estimate the market shares split by type of media in Slovenia.83

138.   As can be seen in Table 4 above, the Transaction will not lead to any affected markets on any type of media in Austria and Croatia. These two countries will therefore not be considered further for the purpose of the present Decision.

139.   In the same way as for the sales side for MBS, the Commission has collected revenue information split by media type from the Parties' competitors during the market investigation. Having analysed the Parties' revenues and some of their main competitors' revenues, insofar as such information was available, the Commission is satisfied that the Parties' estimations are sufficiently reliable for the purpose of this Decision.

140.   As regards the market shares on the media channels excluding direct sales, the Parties have not been able to provide the estimated share of direct sales by type of media in each country. Based on their experience, they estimate that the level of direct buying is generally lower in the TV channel than in the Internet and print channels.84 In particular in Denmark and Sweden, the two EEA countries where the activities of the Parties overlap and in which the direct sales represent on average a significant proportion of the total market, the Parties estimated that the share of direct sales would be 6% for TV and 16% for outdoor in Sweden, and 1% for TV in Denmark85.

141.   In the light of the above and on the basis of all available evidence, the Commission considers that calculating the Parties' market shares on the basis of their actual media cost in each media type and the market size per media type as estimated in the Zenith Optimedia report is a reliable starting point for the purpose of assessing the Transaction.

IV.2.2.3.   CONCLUSION

142.   In light of all the above, the Commission considers that affected markets exist for the sales MBS markets when at least one of the three methodologies (ZenithOptimedia, WARC, Group M) or a country-relevant source, leads to combined market shares of the Parties of more than 15%, and regardless of whether direct sales are excluded. The following geographic markets, where both Parties are active in MBS, are therefore affected: Belgium, Czech Republic, Denmark, France, Hungary, Italy, Latvia, Lithuania, Poland, Portugal, Romania, Slovenia, Spain, Sweden, the United Kingdom, and Norway.

143.   Furthermore, when considering market shares by media channel on the procurement side of the MBS markets, further geographic markets are affected (in addition to the EEA countries already mentioned above). These countries are: Germany, Greece, Ireland and the Netherlands.

IV.3.          METHODOLOGIES  FOR  ESTIMATED    MARKET  SHARES    FOR MARKETING  AND  COMMUNICATIONS SERVICES

IV.3.1       VIEWS OF THE PARTIES

144.   The Parties state that due to the high fragmentation of the MCS market, there are few sources that estimate total market size by EEA country. To estimate their combined shares of supply for the MCS market, the Parties have used their own MCS revenues and the following sources: (i) Eurostat; (ii) Barnes 2013 Worldwide Advertising Agencies Industry report (the "Barnes Report"); (iii) and some local sources of information86

145.   Eurostat compiles annual revenue data for advertising agencies and public relations and communications companies, by country. However, these revenues include media buying revenues that accrue within these companies. The most recently published data is for 2010. The Parties made a number of adjustments to this data in order to use it as a source of demand for their share of supply estimates

146.   First, the Parties made an adjustment for growth to obtain an estimate of demand for 2011 and 2012. They assumed that the demand for MCS grew from 2010 to 2011 and 2012 at the same rate as the demand for MBS has grown. They therefore applied a growth rate based on the MBS growth in advertising expenditure (from ZenithOptimedia). Given that the purchase of media placements is closely linked to the supply of advertising itself (and the related MCS work that goes into producing that supply), MBS growth rate should be a relatively good proxy for MCS growth rate.

147.   Second, the Parties subtracted MBS revenues from the Eurostat figures. To do this, they estimated the maximum amount of MBS revenues that could be included. ZenithOptimedia reports that MBS revenues are at most 15% of advertising expenditures. The Parties therefore reduced the Eurostat total revenues by that amount.

148.   The Barnes Report calculates total revenues for advertising agencies by country, for 18 EEA countries.87 The demand estimates in the Barnes Report are slightly smaller, on average, than the adjusted Eurostat estimates (and about half of the size of the raw Eurostat data).

149.   When available, the Parties also cross-checked the above estimates against the few country-specific sources of information with estimates of total agency advertising revenues.

150.   According to the Parties, the combined entity estimated market shares from adjusted Eurostat figures and from Barnes do not materially differs and this conclusion is further supported by country-specific sources of information.

151.   By discipline, the Parties have been able to calculate their share of supply only for PR as Eurostat provides only an estimation of the total revenues derived from PR services. This share is below [5-10]% in each EEA country. The Parties consider that as they are not relatively stronger in any particular MCS discipline, their shares in any one of these other disciplines would be of the same order of magnitude.

IV.3.2       COMMISSION'S ASSESSMENT

152.   The Commission has assessed the methodologies used by the Parties to estimate their combined market shares. The Commission has also engaged with the Parties' four main MCS competitors in order to understand whether the methodology identified is appropriate to reflect the positioning of the Parties on the MCS market.

153.   On the one hand, one of the four main competitors referred to a number of shortcomings regarding the methodologies used by the Parties. In particular, it mentioned that the sources proposed by the Parties are not recognised in the industry, that the data is out-of- date and that the adjustments and assumptions of the Parties are not sufficiently reliable.

154.   On the other hand, another of the four main competitors indicated that examining revenues is an appropriate basis to estimate MCS market share but suggested to complement this data by revenue ranking from an industry publication such as Advertising Age. Another of the four main competitors acknowledged the evolution of the sector since the WPP/Grey decision and highlighted the importance of conducting an analysis of the current situation.

155.   In WPP/Grey, the Commission estimated the MCS demand in EEA countries on the basis of the 2002 report issued by the U.S. organization Advertising Age. The Commission used this report with adjustments made to update data for MCS growth to 2002 and 2003.

156.   However, the last Advertising Age report for individual European Member States ("Top Agencies in 146 Countries") dates back to 2002 and refers to estimates for 2001.

157.   Therefore, the Commission considers that this report is not a good proxy to estimate MCS demand for 2012.

158.   After having conducted robustness checks, the Commission considers both methodologies applied by the Parties as adequate to estimate the MCS share of demand. Data provided by Eurostat and Barnes is more recent and thus more reliable than the Advertising Age report, the only alternative source mentioned by the four main competitors. Eurostat in particular is entrusted by the European Union for the provision of statistics in Europe.88 Moreover, the two data sources corroborate each as other as Barnes' share estimates, while generally higher, do not deviate substantially from the share estimates based on Eurostat data.

159.   Regarding the Parties’ assumption that MCS has been growing at the same rate as MBS, on which is based the adjustment mechanism applied to Eurostat revenues, the Commission considers that the volume of MCS drives to a sufficient extent the volume of MBS so that the MBS growth rate may be taken as a reasonable proxy for MCS growth. In any case, even if the Commission were to depart from this approach and assume that the MCS market grew at the same rate as GDP, or that the MCS market did not grow between 2010 and 2012, this would not lead to a materially different assessment of the Transaction. Equally, on the conservative approach that MCS demand declined in size by 20% a year in 2011 and again in 2012, no affected market will result.89

160.   Table 5 gives an overview of the Parties' market shares for the supply of marketing communication services estimated on the basis of the sources and the methodologies applied by the Parties:

Table 5: 2012 MCS shares

Note: this table combine calculations based on the Parties' actual revenues and Eurostat market size adjusted for MBS revenues and calculations based on the Parties' actual revenues and Barnes report market size.

161.   Table 5 shows that there is no overlap in Bulgaria, Cyprus, Estonia, Ireland, Latvia, Lithuania and Slovakia. The Parties' activities overlap in the following EEA countries: Austria, Belgium, Croatia, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Italy, Netherlands, Poland, Portugal, Romania, Slovenia, Spain, Sweden, United Kingdom and Norway. However, under any of the Parties' estimates, their combined share of MCS is below 15% in all the overlap EEA countries.

IV.3.3       CONCLUSION

162.   In light of the above, the Commission considers that no affected market for the supply of MCS exist at both national and EEA level.

163.   However, given the difficulties in estimating market shares in the industry, the Commission has nonetheless conducted a competitive assessment for MCS.

IV.4.           ANALYSIS      OF      NON-COORDINATED      EFFECTS      OF      THE TRANSACTION

IV.4.1       MEDIA BUYING SERVICES

164.   In this section, the Commission will first set out the Parties’ description of the media buying negotiation process between, on the one hand advertisers and MBS agencies, and on the other hand, MBS agencies and media vendors.

165.   The Commission will then present the views of the Parties on the impact of the transaction on competition in the MBS market, before finally, assessing the impact of the Transaction on competition on the sales side and procurement side of the MBS market. The Commission will first make a general assessment of arguments that cut across the different EEA countries affected by the Transaction, before undertaking a more detailed analysis on a country per country basis.

IV.4.1.1.   Parties’ general overview of the media buying negotiation process

166.   Timeframe. According to the Parties, when an advertiser has a need for media space, it will typically work with a media buying agency to establish a media buying plan, considering all possible channels. The advertiser and the MBS agency will negotiate the qualitative and quantitative conditions at which the MBS agency will seek to obtain media space. The media agency will be evaluated based on its ability to execute the plan.

167.   The MBS agency will suggest to their clients different scenarios depending on media channels, volumes and targeted audience. The contract may be between the MBS agency and the media owner or between the media owner and the client.

168.   The Parties further submit that while the precise negotiation process between an MBS agency and media owners differs for each media channel, it is broadly similar across the various EEA countries. In most EEA countries and for most media channels, media buying is part of an annual process. The negotiation usually occurs towards the end of a given year for the following year. The negotiations will typically result in one-year framework agreements that usually govern placement, volume, price and rebates. Although the majority of advertising is traded during the annual deal round, it can also be purchased during the course of the year through a process of in-year trading.

169.   According to the Parties, MBS agencies do not usually purchase advertising time/space in advance. This practice, called "brokering", is rare.

170.   Pricing. The Parties explain that MBS pricing is based on: (i) fees for the MBS agency; and (ii) the media cost. Fees are typically calculated by the MBS agency on the basis of a cost approach. The MBS agency will consider the internal resources required to staff a particular project, develop a staffing plan, and calculate its labour costs under that staffing plan, and some contribution to overhead and a margin.

171.   Regarding TV time and space, three methods are individually or in combination used to sell airtime:

-        Rate cards and discounts. Rate cards are published by the TV channels. The MBS agency and its client will negotiate an annual/volume/seasonal discount that applies to each booking. The discount is secured on the basis of share or volume commitments. The MBS agency then buys airtime using the rate card rates less the negotiated discounts. Advertisers are often unwilling to guarantee annual expenditure in advance. For this reason, agreements can include a sliding scale that links volume, or share of volume, to price. Germany, France, the Netherlands and Italy are typical rate card markets.

-        Package: The MBS agency and its client negotiate a cost per Gross Rating Point (GRP) with TV channels, including controls on the quality of airtime to be provided (for example primetime, off-peak).91 Media buyers endeavour to deliver a maximum number of GRPs at a minimum cost. The media channel compensates for any shortfall in GRP delivery. The rebates to which the customer will be entitled in such a scenario can take different forms. For example, the advertiser can provide additional slots free of charge or it can commit to better placing of the advertisement. This practice is not developed in all markets, for example no GRP or CPR92 is guaranteed in Poland. The United Kingdom, Ireland and Spain are typical package markets.

-        Upfront: the client will agree to buy an advertising sport in each episode of a TV series. The contract is agreed in advance for a period of 6-9 months and it often not cancellable. Good or poor rating delivery is generally at the client’s risk, although it may be possible to negotiate compensation if the TV series does not reach the expected GRP. This is hardly ever seen in the EEA, although it is the main mechanism used in the United States.

172.   Print advertising space is traditionally negotiated off a rate card that distinguishes pricing based on a range of format parameters, including size and location of the advertisement within a particular publication. MBS agencies and publishers negotiate intensively over discounts off the rate card, with advertisers trading volume and share, usually over a calendar year, for discount levels.

173.   Outdoor advertising space is typically sold on a per billboard/time formula. Sometimes it includes a focus on a particular city or area. Agencies and media owners negotiate discounts off billboard owners' rate card prices depending on location as well as the prevailing conditions of supply and demand.

174.   Cinema space is typically sold on a per-viewing (number of admissions) basis.

175.   For the Internet, all key word search buys take place through online auctions. Display buys are predominantly made on a CPM93 basis. However, an increasing proportion of display buys are being made on online auctions.

176.   Client discounts/rebates. The Parties explain that the terms and conditions of all media owners contain a detailed description of pricing, including the discounts and rebates that a media buyer’s clients can benefit from. These discounts can be dependent on volume/quantity, on the client obtaining a certain percentage of its advertising requirements from a specific media channel (the so called share commitment) or on the progression of the customer’s spending from one year to the next. Discounts can also exist for new customers or certain types of customers.

177.   Although the terms of agreement of media owners are typically public, the Parties submit that discounts are client specific: the amount of the rebate is not solely dependent on volume, but also on other parameters such as trend, sector of activity and season. Media owners also gratify growth, or sometimes even stability, of the customer’s media budget. The rebates are cumulative: a customer can benefit from the application of different rebates.

178.   According to the Parties, volume bonuses, rebates, discounts, free spots and any other incentives which are earned directly in connection to the specific spending level of a client are typically passed on to the client.

179.   Discounts on prices are negotiated by the media buying agency with the media owner during annual negotiations. In countries using the rate card94 , the discounts are negotiated off the rate card by the media agency before any media time or space is bought. In such a case, discounts are automatically passed-on to clients.

180.   In certain EEA countries that use the so-called “package mechanism”, the media vendor commits to provide discounts if either the actual GRP negotiated by the media  buying agency for its customer during annual deal rounds or the CPR is weaker than agreed. Usually, these discounts will take the form of additional free slots, or better placement of the advertisement at later stages of the campaign. In such cases, the discounts will be awarded during the campaign. In such a case, the discounts are also automatically passed-on to clients.

181.   Agency Volume Bonifications. The Parties explain that Agency Volume Bonifications (AVBs) typically take the form of cash incentives that are offered to MBS agencies by media owners.

182.   AVBs exist in most EEA countries. [Description of Parties' AVB practice in certain EEA countries]. In France, media owners may grant rebates to clients depending on the overall volume of the clients’ purchase. These rebates, as all other rebates under the Loi Sapin, are automatically invoiced to the client who benefits from such rebates in full. The client is then free to share part of these rebates with the MBS agency. In most other EEA countries, AVBs are received by MBS agencies from media vendors in consideration for the overall volume of media space or time placed with a particular media supplier for all agency clients.

183.   In particular, with respect to TV, [Description of Parties' practice regarding AVBs negotiations]. TV broadcasters historically have provided incentives to MBS agencies to place more volume, or a greater overall share, of client ad placements, over the course of a year, and in general routinely do so now in some, but not all, EEA countries. [Description of Parties' practice regarding AVBs negotiations]. The Parties, however, explain that the rebates received by MBS agencies from TV broadcasters are not significant in comparison to the total discounts negotiated off the rate card for the direct benefit of customers.

184.   [Description of Parties' practice regarding the amount of AVBs passed on to clients]

IV.4.1.2.   VIEWS OF THE PARTIES ON COMPETITIVE ASSESSMENT

185.   On the sales-side, the Parties submit that the Transaction will not lead to unilateral anticompetitive effects95 as contracts with customers are awarded through competitive bidding processes, which enable advertisers to play MBS agencies against each other to obtain the best overall value proposition.

186.   In addition, according to the Parties, switching costs are low and clients frequently switch between MBS agencies. In support of their views, the Parties have submitted an analytical document assessing MBS tenders organised by advertisers in a large number of EEA countries.96 Contracts are generally negotiated on a project basis with relatively short terms and broad client termination rights. Furthermore, discounts received from media owners are often portable when an advertiser switches to another media buying agency, which further facilitates clients' freedom of movement.

187.   On the procurement-side, the Parties submit that even post-Transaction, they will be far from matching the leverage most media owners can exert in negotiating the sale of media. They consider that the media vendors control an asset that is unique, not easily replicable and a "must-have" for agencies providing MBS to their clients. Moreover, in most EEA countries, media ownership tends to be highly concentrated in the hands of a few operators, which gives them significant countervailing power when negotiating with the MBS agencies.

188.   First, as regards online advertising, the Parties submit that online publishers have a broad range of channels through which they can sell inventory to advertisers. Those channels include individually negotiated direct sales and automated (or programmatic) sales through intermediaries such as ad networks, ad exchanges and supply-side platforms. Increasingly, online publishers open their inventory to auctions and real time bidding. Omnicom estimates that in the EEA approximately 5-15% of display advertising, and 100% of search advertising, is bought through programmatic bidding (auctions). These numbers do not vary in any appreciable way by EEA country. Many such auctions are run either: (i) by an online publisher’s supply-side platform, where the publisher chooses which advertiser, agency, or demand-side platform may participate; or  (ii) in general auctions by Google, Yahoo! and others. In both instances, buyers or their agents submit bids for individual impressions, which are then sold to the highest bidder. The Parties submit that whatever fraction of demand the Parties account for in the online space does not confer on them any market power, because media owners will continue to have the choice between dealing with advertisers directly, dealing with competing MBS agencies, dealing with independent demand-side platforms, or selling at programmatic auctions. In the latter case, the Parties further submit that any buyer power that the merged entity may have is completely negated as each online advertising comes up for bid individually and anyone can bid for it regardless of its size.

189.   Second, MBS agencies are constrained by the need to fulfil a client’s media plan, which specifies the volume, media outlet and quality parameters and they must contend with client conflicts (that is to say a media purchasing client not wanting its media agency to purchase media for its client’s rivals, or to allow aggregation with rival purchases). Any influence an MBS agency has on specific purchase decisions of an advertiser is constrained by the agencies’ obligation to deliver on the key performance indicators agreed with the client. The advertiser (or an auditor hired by the advertiser) assesses an agency’s performance against such indicators, and, in turn, the agency’s continued engagement. If, post-merger, a media owner refused to accept the terms of the merged entity, that demand would not disappear, but would be channelled through competing MBS agencies, to the detriment of the merged entity, which would lose its commission and not be able to serve its clients

190.   Third, according to the Parties, the presence of multiple MBS agencies in most EEA countries and the threat of disintermediation will shield media vendors from any effort by the merged entity to force ad space pricing below competitive levels. Media vendors could retaliate against any attempt by the merged entity to exercise buyer power post- merger in several ways. For example, they could discriminate against the merged entity by refusing to sell through it the most precious ad slots (for example half time breaks in Champions League games), placing it at a critical competitive disadvantage when winning business from its clients. They could also by-pass the merged entity altogether, absorbing its commissions and directly selling to the advertisers, a practice that is prevalent in some EEA countries and in digital media.

191.   Fourth, as regards the specific question whether certain TV broadcasters may try to balance losses due to rebates granted to one media buyer against other media buyers, the Parties submit that they are not aware of TV broadcasters suffering losses due to rebating, and do not believe that this in fact occurs. According to them, the TV broadcasting industry in most, if not all, EEA countries is highly concentrated and TV broadcasters enjoy significant negotiating leverage in the sale of advertising spots. TV broadcasters can also sell to multiple alternative buyers. The Parties are not aware of situations where a TV broadcaster engaged in a "balancing-of-rebates" approach by offering a lower rebate or discount to one MBS agency’s clients because it had already offered a higher rebate or net discount to another agency’s clients; nor have they ever faced a situation where a TV broadcaster communicated that it was denying a rebate or discount level to one agency’s clients because it had offered a greater rebate or discount to another agency’s clients. The Parties do not believe TV broadcasters put themselves in a loss position through rebates with any agency or advertiser.

192.   Finally, the Parties submit that even if, as a result of the Transaction, the merged entity would be able to negotiate better rates with certain media vendors by committing to higher volumes, competition between MBS agencies and the transparency governing media agencies-advertisers relationships in most EEA countries would ensure that they are passed on to advertisers.

193.   The Parties submit that the features described in the present section, which prevent the merged entity from acquiring market power in the purchase of media, are present throughout the EEA.

IV.4.1.3.   COMMISSION'S ASSESSMENT

IV.4.1.3.1.   Introduction

194.   In assessing the competitive effects of the Transaction on the markets for MBS, the Commission has, consistent with its decisional practice, distinguished between; (i) the sales side of MBS and (ii) the procurement side of MBS.

195.   According to the Commission Guidelines on the assessment of horizontal mergers, large market shares - 50 % or more - may in themselves be evidence of the existence of a dominant market position. However, smaller competitors may act as a sufficient constraining influence if, for example, they have the ability and incentive to increase their supplies. A merger involving a firm whose market share will remain below 50 % after the merger may also raise competition concerns in view of other factors than market shares such as the strength and number of competitors, the presence of capacity constraints or the extent to which the products of the merging parties are close substitutes.97

196.   In this context, and given the presence of other large and smaller competitors in each overlap EEA market, the Commission performed a more detailed analysis of the markets where the combined market shares of the Parties is over 30% on the sales side of the MBS market, that is to say: Italy, Poland and Romania. A more succinct analysis of the other markets has also been performed. On the procurement side, the Commission also looked in particular at markets by media channel where the combined market shares of the Parties were above 30%, that is to say: the Czech Republic, Denmark, Lithuania, Norway, the United Kingdom and Spain, in addition to Italy, Poland and Romania (see Section IV.4.1.4 about country-by-country assessment)

197.   Preliminarily, the Commission notes that issues about capacity constraints98 for MBS agencies offering their services to advertisers have in general not been mentioned by respondents to the market investigation, subject to the existence of possible difficulties for the small agencies to provide some specific, large-scale, services to the largest advertisers, considering for instance the time-constraints imposed by the calendar aspects of defining and negotiating campaigns with media vendors, or the need for a global reach.99 As a result, the Commission considers that issues relating to capacity constraints are generally not present on the sales side of the MBS market and MBS agencies can generally provide their services to all advertisers.

198.   In addition, as was indicated in WPP/Grey100, an advertiser may be concerned that its MBS agency does not in parallel provide services to its main competitors. Indeed, advertisers are concerned about their communication strategy being disclosed to competitors; the agency might, in addition, be faced with choosing between two advertisers in a critical time-slot on TV (for instance a match during the FIFA World- Cup) or with respect to an in-demand magazine page. A number of customers that responded to the market investigation mentioned that the Parties run some risk of losing actual or potential customers due to conflict of interest issues. Notably, concerns relating to the handling of sensitive customer information on pricing, or timing of campaigns were mentioned.101 As a consequence, the Commission considers that the merged entity may have to maintain a number of separate MBS agencies that will compete against each other for advertisers on the sales side of the MBS market, and some advertisers may even decide to stop using the services of the merged entity and use the MBS agencies of their competitors post-Transaction.

199.   Conversely, a small number of media vendors explained that issues of conflict of interests may reduce the number of possible agencies that large advertisers may choose from102. However, the Commission considers that this argument is not inconsistent with the one outlined on the previous paragraph, that the Transaction may result in some customers leaving the merged entity to use the MBS agencies of their competitors. RECMA report Compitches 2012 notes in this regard “every year shows big changes in the hierarchy partly because of client conflicts preventing agencies to participate”.103

200.   On the sales-side, in assessing the effects of the Transaction, the Commission analysed a number of factors.

201.   In addition to the combined market shares of the Parties104, the Commission first assessed what competitive constraints other competing MBS agencies have exerted, and will continue to exert on the Parties post-Transaction.

202.   Second, the Commission assessed to what extent contracts with customers are awarded through competitive bidding processes so that advertisers are, and will remain post- Transaction, in a position to play a sufficient number of MBS agencies against each other to obtain the best overall value proposition. In particular, the Commission assessed whether this would hold true not only for smaller customers, but also for larger customers that may have specific needs that could only be fulfilled by MBS agencies with  wider-than-national footprint.

203.   To that effect, the Parties submitted an analytical document, prepared by the economic consultancy, CompassLexecon, which assesses MBS tenders organised by advertisers in 22 EEA countries.105

204.   The Commission asked the Parties to conduct a consistency check of the underlying RECMA Compitches data for 2010, 2011, and 2012 with their own tender's data.106 Overall, despite a few identified corrections to the RECMA data, the Commission takes the view that the RECMA Compitches reports are generally accurate for the purpose of the present assessment.107  The Commission therefore considers that it can base its analysis on the report produced by Compass Lexecon and the underlying data provided. This analysis, as well as others, such as the results of the market investigation, is outlined in more detail below in Section IV.4.1.3.2.

205.   Third, the Commission assessed to what extent switching costs are low and customers will be able to switch to other MBS agencies than the merged entity if it starts to raise prices post-Transaction.

206.   Fourth, the Commission assessed to what extent the Parties were close competitors prior to the Transaction, so as to determine whether the Transaction will remove an important competitive force in the market. The above-mentioned CompassLexecon analysis provided useful information in this regard, which was complemented by the responses to the market investigation.

207.   On the procurement-side, the Commission considers that the results of the competitive assessment on the procurement side of the MBS market should mirror the results of the competitive assessment on the sales side of the MBS market. This is because, notably, the combined market shares of the Parties in any given EEA country are similar on the sales side and on the procurement side of the MBS market

208.   As a result, even though the combined market share of the Parties may vary per type of media on the procurement side of the MBS market, the global competition assessment on the sales side has to be considered when analysing the procurement side of any MBS market. By way of example, if the combined market share of the merged entity is 25% on the sales side of a national MBS market and other competing MBS agencies are present on that market and able to exercise a significant competitive constraint on the merged entity so that no competition concern can arise from the Transaction on the sales side of the MBS market, even if the merged entity has a combined market share of for instance 40% on the procurement of TV advertising (or other media channels), proper account should be taken in the competitive assessment of the presence of competing MBS agencies, to which media vendors could turn to if the merged entity were to be in a position to dictate its price conditions on them.

209.   For the countries where a more detailed analysis has been performed, in addition to the combined market shares of the Parties108, the Commission assessed in particular whether the Transaction would lead in the next three years to the merged entity exerting buyer power on media owners because of its increased size, and thereby obtaining lower prices or better conditions for space/time from media owners (or conversely whether media owners would remain in a position to exercise significant countervailing power when negotiating with the merged entity).

IV.4.1.3.2.   Competitive assessment – sales side Competitive constraint from other MBS agencies

210.   Based on the available evidence outlined in this section of the Decision, the Commission considers that there will remain significant competitive constraints on the merged entity post-Transaction.

211.   First, the RECMA report Compitches 2012 states that “the index per year demonstrates that most of the networks are not successful every year therefore that the competition is not dominated by one or a few players.”109 (emphasis added). In the EEA countries where the activities of the Parties overlap, this statement is generally confirmed by the combined 2012 market shares of the merged entity, which remained generally below 30%, and never reached 40%.

212.   Second the Commission notes that when an advertiser is in need of MBS services, and wishes to sign a new contract with an MBS agency, it typically either negotiates directly with an MBS agency, or puts in place a tender, where different MBS agencies compete against each other, before one is eventually selected by the advertiser as the winner of the tender.

213.   Even if the responses of competitors that replied to the market investigation were mixed with regard to the extent to which MBS contracts are concluded by tender or direct negotiation, a large majority of customers indicated that their common practice is to award contracts through a bidding process.110

214.   RECMA report Compitches 2012 notes in this regard that “Pitches and reviews are more than ever a major focus for media agencies.”111 The Commission considers that this further confirms the use of tenders as a means for advertisers to select their MBS agencies.

215.   The tender procedure encompasses essentially three phases: the request for information (RFI) stage, the request for proposal (RFP) stage, and the “chemistry” meeting stage. Customers responding to the market investigation explained that they invite on average 4.7 contenders at the RFI stage, and 3.2 contenders at the RFP stage of their tenders.112 These two stages are the most important one for the competition analysis.113

216.   Based on the data provided in the CompassLexecon analysis and the responses to the market investigation, the Commission considers that overall, across the EEA, the Parties face competition from at least the other four international groups, namely WPP, IPG, Dentsu-Aegis and Havas, when they participate in tenders. Furthermore, when the Parties overlapped in tenders, they often faced competition from at least two other competitors, including but not limited to the other four international groups.

217.   In a more detailed fashion, the data analysed by CompassLexecon show that there were four or more participants in 68% of the tenders in which both Parties were present, compared to 28% of tenders when the Parties do not overlap. Furthermore, there was at least one competitor in 92% of the Parties’ overlap tenders. The other international groups’ participation in tenders was also somewhat similar for overlap tenders and for non-overlap tenders: (i) WPP participated in 66% of overlap tenders and 68% of non- overlap tenders; (ii) Dentsu-Aegis participated in 44% of overlap tenders and 43% of non-overlap tenders; (iii) Havas participated in 21% of overlap tenders and 21% of non- overlap tenders. The only exception is IPG, which participated in 31% of overlap tenders compared to 42% of non-overlap tenders. Independent MBS agencies also participated less often in overlap tenders compared to non-overlap tenders, but are still present in 18% of the tenders involving both Omnicom and Publicis.

218.   Third, the analysis does not show significantly different results on a country level analysis than when analysing data at EEA-wide level.

219.   The average number of participants in a tender is higher for overlap tenders than for non-overlap tenders in all analysed EEA countries except for Ireland. There is at least one other participant in more than 78% of the tenders in which the Parties overlap in all EEA countries except Ireland, Norway and the United Kingdom (67% for the 3 countries). There are 4 or more participants in at least 50% of tenders (and at least 70% of tenders in 8 countries) in which both the Parties are present in all EEA countries except Ireland, where there are 4 or more participants in 33% of the tenders.

220.   Furthermore, WPP is the most prominent participant in Omnicom tenders in all EEA countries except Croatia and Greece (where it is IPG). WPP is also the most prominent participant in Publicis tenders in all EEA countries except, Greece (where it is Omnicom), Hungary (where it is IPG), and Portugal and Spain (where it is Havas Media). Finally, there are 9 or more independent agencies participating in tenders in France, Germany, Ireland, the Netherlands, Romania, Spain, and the United Kingdom, and there are 2 or more independent MBS agencies participating in tenders in all countries analysed.

221.   As regards Latvia, Lithuania and Slovenia, which are not covered by RECMA Compitches data, the Parties have provided internal data concerning their participation in 2012 tenders. [Summary of Parties' win, loss and incumbency data in Latvia, Lithuania and Slovenia].114

222.   Fourth, some form of intra-group competition will continue to exist post-Transaction. A majority of competitors that responded to the market investigation indicated that MBS agencies belonging to the same group currently compete on tenders, and that the Transaction will not change this practice, which is ultimately decided by clients who can decide to pitch agencies of a same group against each other.115 Certain customers confirmed that agencies belonging to the same group currently compete as hard against each other as against other agencies116, and the Transaction will have no effect on the current practice of intra-group competition.

223.   Fifth, respondents to the market investigation indicated that it is likely that there will remain a sufficient number of MBS agencies post-Transaction to prevent the merged entity from raising prices. Indeed, even if some competitors responded that the merged entity may be able to obtain better terms from media vendors because of its increased size, which in turn could drive some smaller MBS agencies out of the market as they would not be able to compete with the merged entity, a majority of competitors considered that there will remain a sufficient number of MBS agencies post-merger to prevent the merged entity from raising its prices. A majority of competitors also believed that this would hold true both for local MBS agencies and for larger international ones.117

224.   Sixth, a large majority of customers believed that there will remain a sufficient number of MBS agencies post-merger to prevent the merged entity from raising its prices.118 These customers believed that this would hold true both for local MBS agencies and for larger international ones.119

225.   Finally, in all the EEA countries where the Parties activities overlap, the Commission found that post-Transaction, the Parties will continue to face competition from at least three large advertising groups from the largest competitors of Publicis and Omnicom, that is to say WPP, IPG, Dentsu-Aegis, and Havas. The Commission considers that the presence of such large competitors, with global footprints, will meet any specific needs that larger advertisers may have, post-Transaction

226.   In light of all the above, the Commission concludes that there will remain significant competitive constraints on the merged entity post-Transaction on the sale side of the MBS market in all overlap EEA countries.

Customers' ability to switch, incumbent advantage, and barriers to entry

227.   Should the merged entity raise its prices post-Transaction, a majority of competitors considered that it would be “very easy” or “somewhat easy” for an advertiser to switch to another MBS agency. A number of competitors explained that it would take in general 3 to 6 months to switch to another MBS agency, and that no significant costs would be incurred, apart from the cost of re-tendering. One competitor explained that the time may be longer for large international advertisers (between 6 and 9 months).120

228.   The responses to the market investigation were more mixed from the point of view of customers. While a number believed that it would be either very easy or somewhat easy to switch to another MBS agency, a number of other customers responded that it would be either somewhat difficult or very difficult. The Commission notes, however, that the customers which considered it rather difficult or rather easy to switch to another MBS agency, agreed to a large extent that it would take between 3-6 months to switch (and up to 9 months for larger international groups), and that the main costs associated with switching would relate to training, development and staff-related costs.121

229.   In addition, a large majority of customers replied that there are advantages in purchasing services from incumbent MBS agencies. The most commonly-mentioned advantage of using incumbent agencies is the knowledge of the clients' industry and specific needs and requirements. A number of customers, among them large ones, qualified their answer by explaining that even though there are some advantages to purchasing services from incumbent MBS agencies, they can decide to switch agencies when they believe that an alternative agency could better serve their business needs.122

230.   Regarding the question whether it would be easy for them to start providing MBS services in another EEA country where they are not already present, a number of competitors explained that they could do this by for instance (i) launching a start-up and growing organically by prospecting new clients; or (ii) acquiring an MBS agency that is already present on the market; or (iii) signing a commercial agreement or joint venture with a local MBS agency; or (iv) starting up with a specialised agency (for instance in digital media, where one competitor explained that it is easy to enter any EEA country and buy digital media), and use this as a foundation to buy traditional media afterwards.123 According to these respondents, the time required to do so would range from immediately for digital advertising services, to between 6 months and 2 years in the case of other media.124

231.   Regarding tenders, based on the 2012 data provided in the CompassLexecon analysis (with complete information on the winner, incumbent and contenders), Omnicom was the incumbent in 118 tenders over the sample period. Out of these, 38 tenders were retained125 or switched between Omnicom agencies (32%) and 80 tenders were lost (68%). It lost most accounts to WPP (29 accounts, 36% of the loss), followed by IPG (14 accounts, 18% of the loss), Dentsu-Aegis (12 accounts, 15% of the loss), Publicis (9 tenders, 11% of the loss) and Havas (3 accounts, 4% of the loss); 13 accounts were lost to independent agencies.126

232.   In turn, Publicis was the incumbent in 125 tenders. Out of these, 35 tenders were retained or switched between Publicis agencies (28%) and 90 tenders were lost (72%). It lost most accounts to WPP (32 accounts, 36% of the loss), followed by Dentsu-Aegis and Omnicom (17 accounts each, 19% of the loss each), IPG (14 accounts, 16% of the loss), and Havas (7 accounts, 8% of the loss). 3 accounts were lost to independent agencies.127

233.   In light of the above, the Commission takes the view that customers have the ability, and actually do switch regularly between MBS agencies. The Commission also considers that there does not seem to be any significant advantage in being the incumbent MBS agency in a new tender. These conclusions apply generally to all overlap EEA countries.

Direct media buying

234.   The Commission also assessed whether, and in what proportion, customers currently purchase advertising space and time directly from media vendors, hence by-passing MBS agencies such as the merged entity.

235.   A majority of competitors that replied to the market investigation indicated that some customers buy advertising directly from media vendors. They, however, estimated that the proportion of direct purchase from advertisers to media vendors was relatively small, and was the exception rather than the general trend, which is to use MBS agencies that are able to provide price benchmarks, dedicated tools and expert teams. The proportion of direct sales varies across country and type of media. Generally, competitors estimated that the percentage of direct sales ranges between 5% to around 20%, with some significant exceptions, in particular in Sweden and Denmark.128

236.   A large majority of media owners also replied that they sell advertising time or space directly to advertisers. Again, however, they estimated that the proportion of such sales is limited. It varies across country and type of media, but generally, media owners estimated that it is less than 10%.129

237.   Half of the customers replied that they have purchased advertising time or space directly from media vendors in the past 3 years, while the other half replied that they had not.130 Out of those who did, a majority confirmed that their amount of direct purchase has been relatively minor (up to 10%).

238.   In light of the above, the Commission considers that direct media buying, even though its current proportion of total media purchase seems limited, could constitute an additional possible competitive constraint on the merged entity post-Transaction, in particular when shares of direct sales become significant (as of 30% in the context of the present Decision). Country-specific situations will be examined in Section IV.4.1.4.

Closeness of competition between the Parties

239.   First, based on the data provided in the CompassLexecon analysis, it appears that in the EEA, the Parties do not compete frequently against each other.

240.   Out of the 3,029 tenders analysed during the period 2010-2012, the Parties only overlapped in 169 of them (5%). While this may understate the extent to which the Parties overlapped in tenders because many of these tenders have incomplete information, even if the analysis is restricted to tenders where full information on winner, incumbent and contenders is available, this percentage of overlap tenders increases only to 16%. In other words, even in the worst-case scenario, the Parties did not compete in a large majority (84%) of the tenders sampled. Furthermore, as will be outlined below, this trend is confirmed on a country per country basis. The Parties overlapped in a small proportion of tenders in each country (less than 20% in all countries, excepted in Czech Republic (22%), Greece (38%), Italy (31%), Poland (29%), and Spain (21%).131

241.   Second, the Parties tend to lose more business to other groups than to each other during the period 2010-2012. As will be outlined in more below, this trend is mostly confirmed on a country per country basis. Publicis loses more accounts to WPP, Dentsu-Aegis, IPG and Havas than to Omnicom in all EEA countries except in the Czech Republic, Denmark, Greece and Norway. In turn, Omnicom loses more accounts to WPP, Dentsu- Aegis, IPG and Havas in each EEA country except in Greece and Poland, where it loses an equal number of accounts to WPP and Publicis.132

242.   Third, market participants that replied to the market investigation indicated that Publicis and Omnicom are in general not each other's closest competitors.133

243.   In light of the above, the Commission considers that Publicis and Omnicom generally are not each other's closest competitors, so that the Transaction will not remove an important source of competition in the MBS market. Country-specific situations will, however, be assessed in Section IV.4.1.4 below.

IV.4.1.3.3.   Competitive assessment – procurement side Negotiating power and ability to set prices between media vendors and MBS agencies.

244.   When assessing the purchasing power on the procurement market, the Commission has examined, as it did in WPP/Grey, whether a strong MBS agency will be able to put pressure on its media vendors with the effect of virtually controlling their business. In an extreme case, this customer could even force dependent suppliers out of the market by withdrawing demand from him.134

245.   Responses to the market investigation were mixed as regards the question of who holds negotiating power between media vendors and MBS agencies. On one hand, a number of competitors explained that generally, MBS agencies have negotiating power vis-à-vis small and medium-size media vendors. On the other hand, several competitors explained that the distribution of negotiating power between media owners and media buying agencies varies between media channels135; for instance, they considered that in particular in digital media, a vendor such as Google has equal to superior negotiation power vis-à-vis MBS agencies. Other competitors explained that major groups owning TV channels also enjoy strong bargaining power versus MBS agencies.136

246.   However, the Commission's assessment on the sales side indicates that in all EEA countries where the activities of the Parties overlap, the merged entity will continue competing to buy ad time/space with several competitors. Should the merged entity try to exert leverage on media owners, the latter will continue to have the possibility to turn to competitors. This in turn will have consequences for the merged entity on the sales side, as customers may decide to move to competing MBS agencies which are able to purchase advertising space or time from the media vendors which will work with the competitors of the Parties rather than with them.

247.   The Commission therefore considers that even though the Parties may represent a relatively important share of the media owners' sale in some markets, the countervailing power on the media owners' side will remain significant enough to offset the increase in negotiation power of the merged entity. The degree of consolidation at the level of media vendors will play a significant role in the assessment at country and media level.137

248.   Furthermore, the ability to negotiate prices and rebates does not seem to be the only factor influencing the choice of advertisers when selecting an MBS agency, other factors such as the proposed media plan and the expertise are also important. When asked how important the ability of a MBS agency to achieve significant rebates or other conditions such as free TV spot or free advertising pages from media owners is, a number of respondents to the market investigation explained that: "It is important, but not the only factor"138, Likewise another competitor explained, "In some MBS countries there may be some specific examples of higher market share which might influence negotiating power with media owners. However, purchasing conditions may be affected by a number of different factors (e.g. the specific client, historical and potential nature of client spend, media owner interests, overall agency spend, seasonality, agency trading expertise) which would not necessarily allow one party to dictate conditions".139

249.   Moreover, there is no convincing evidence to support the proposition that the merged entity's market power may in the longer term prompt media owners to raise their prices or lower the conditions offered to other smaller MBS agencies in order to compensate for turnover loss, and put these MBS agencies at a competitive disadvantage to the extent that they would be forced out of the market, eventually leading to price increase for advertisers.

250.   In particular, under the (undemonstrated) hypothesis that the Transaction would increase the Parties’ bargaining power in their negotiations with media vendors, the Commission shares the view that these media vendors would not have the incentive to increase their prices to other smaller media agencies. First, any attempt to raise prices to smaller agencies would likely reduce demand from these agencies. The reduced demand of smaller agencies would leave vendors with advertising time they would have to sell to the larger agencies at lower prices and the merged entity and media vendors’ profits would thus be lower than what they would otherwise obtain without raising prices charged to smaller agencies. Second, media vendors prefer to have as many credible media buying alternatives to the merged entity as possible post-Transaction, as this will help them counter any hypothetical increase in the merged entity’s bargaining power. Third, media vendors are unlikely to price discriminate against the smaller media buying agencies.

251.   A trade association that replied to the market investigation raised the concern that the Transaction may increase the bargaining power of the merged entity, due to its increased size, especially in the digital environment. In particular, it expressed the concern that “a single entity controlling the spending of many leading brands would be in a strong position to actively block publishers from working with other agencies. This would lead to less competition for advertising inventory, which could also have a negative impact on pricing”.140 The Commission considers, however, that the arguments used in the previous paragraphs such as the remaining presence of competing MBS agencies and the degree of concentration of media vendors also apply to the sale of digital advertising inventory. In addition, Google, as an online advertising intermediary is a strong actor in the sale of digital advertising inventory and is likely able to withstand any possible increase in the negotiating power of the merged entity post-Transaction.

252.   Finally, a limited number of media vendors stated that the merged entity will become so large that when they will extract better terms against media vendors, they will also not have the incentive anymore to pass on their rebates to their customers. The Commission considers, however, that this concern is unfounded. First, evidence collected during the market investigation supports the view [Evidence about pass-on]141. In addition, a large majority of customers explained that they are aware of the levels of rebates obtained by their MBS suppliers, and usually contractually require that rebates relating to their activity are passed on to them, and have put in place either internally or externally through media auditors a system of checks to ensure that rebates obtained by MBS agencies are actually passed on to them.142 Therefore, even if the merged entity could hypothetically (quod non) consistently extract better prices and rebates than their competitors from media vendors, they will have to pass these rebates to advertisers. If they did not, advertisers would switch to other competing MBS agencies that will remain present on the market.

253.   While the elements analysed in the present section are not on balance indicative of any serious doubts being raised by the Transaction in all overlap EEA countries in the MBS markets, the Commission will refine its assessment on a country-by-country basis, and draw conclusions in the following section.

IV.4.1.4.   COUNTRY ANALYSIS

IV.4.1.4.1.   Belgium

a.      Sales side

254.   In 2012, the Parties’ combined market share in MBS in Belgium amounted to [5-10]% including direct sales, and [10-20]-[10-20]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

255.   The merged entity will continue to face competition from the four other large advertising networks143, WPP, Dentsu-Aegis, IPG and Havas, as well as from independent players144. Many of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach

256.   The analysis of RECMA Compitches data indicates that the number of participants in Belgium was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.145 In 75% of overlap tenders there were four or more participants compared to 32% for the rest of the tenders.146

257.   Therefore, the merged entity will continue to face competition from a number of competitors in the tenders in which it will take part post-Transaction.

258.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data for Belgium shows that, on average, in tenders where it was incumbent, Omnicom won in 33% of the cases and in tenders where it was incumbent, Publicis won in 44% of the cases.147

259.   According to the analysis of the Parties' participation in 2010-2012 tenders in Belgium in the RECMA Compitches data, Publicis and Omnicom overlapped in 15% of tenders. Out of the tenders in which Omnicom participated, Publicis was also present in 44% of the cases. Out of the tenders in which Publicis participated, Omnicom was also present in 33% of the cases.148 The most prominent competitors in all tenders in which Omnicom participated in Belgium were WPP first and IPG second. Likewise, in all tenders in which Publicis participated in Belgium, WPP was first followed by Dentsu- Aegis. 149 In tenders in which Publicis was the incumbent, it lost to Havas, and its three other large competitors (WPP, IPG and Dentsu-Aegis). Omnicom lost contracts in favour of WPP and Dentsu-Aegis.150

260.   Moreover, a majority of respondents to the market investigation consider that Publicis' closest competitor in Belgium is WPP, followed by Dentsu-Aegis. Likewise, Omnicom's closest competitor is WPP, followed by Dentsu-Aegis.151

261.   In view of the above, the Commission considers that Publicis and Omnicom are not each other's closest competitors in Belgium.

262.   Finally, a majority of customers that responded to the market investigation did not raise any country-specific issue as regards Belgium that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.152

263.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Belgium.153

b.      Procurement side

264.   Table 6 below presents market shares of the Parties on the procurement side of the MBS market in Belgium154:

Table 6: Belgium - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time155)

265.   As can be seen in Table 6, none of the different types of media channels are affected by the Transaction

266.   The Transaction does not therefore raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Belgium.

IV.4.1.4.2.   Czech Republic

a.      Sales side

267.   The Parties’ combined 2012 market share in MBS in the Czech Republic amounted to [20-30]% including direct sales, and [20-30]-[20-30]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

268.   In the Czech Republic, the merged entity will face competition from the four other large advertising networks156, WPP, Dentsu-Aegis, IPG, Havas, as well as from a number of other independent players.157 Several of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA- wide or global) reach.

269.   The analysis of RECMA Compitches data for the Czech Republic indicates that the number of participants is higher in tenders where Publicis and Omnicom are both present compared to the rest of the tenders.158 In 56% of the overlap tenders there were four or more participants while this ratio was only 19% for the rest of the tenders.159

270.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. Indeed, the analysis of RECMA Compitches data for the Czech Republic shows that, on average, in tenders where it was the incumbent, Omnicom won in 33% of the cases.160 No conclusions can be drawn with regard to Publicis as the complete dataset includes only one tender in which Publicis was the incumbent.

271.   According to the analysis of the Parties' participation in 2010-2012 tenders in the Czech Republic in the RECMA Compitches data, Publicis and Omnicom overlapped only in 22% of all tenders. Omnicom was present in around half of the tenders in which Publicis was present (53%) while Publicis was present in less than half of the tenders in which Omnicom was present (35%).161 The most prominent competitor in all tenders in which each of Omnicom or Publicis participated in the Czech Republic was WPP, while Publicis and Omnicom were the second most prominent competitors respectively.162 In tenders in which Omnicom was the incumbent, it lost mainly to WPP.163 When Omnicom was not the incumbent, it mainly won from WPP. Likewise, Publicis mainly won from WPP.164

272.   A majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Omnicom. Omnicom's closest competitor is also WPP, followed by Publicis.165

273.   In view of the above, Publicis and Omnicom are not each other's closest competitors in the Czech Republic.

274.   Finally, a majority of customers that responded to the market investigation did not raise any country-specific issue as regards the Czech Republic that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.166

275.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in the Czech Republic.

b.      Procurement side

276.   Table 7 below presents market shares of the Parties on the procurement side of the MBS market in the Czech Republic167:

Table 7: Czech Republic - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales – Shares of the Parties per cost of media time168)

277.   As can be seen in Table 7, the following media channels are affected by the proposed Transaction in the Czech Republic: TV and outdoor.

278.   TV and Outdoor media owners are concentrated in the Czech Republic. Indeed, in the TV sector, three players represent 74% of the market (Nova TV, Prima TV and Czech Television). Furthermore, the market for TV advertising is led by the private broadcaster groups Central European Media Enterprises, Mediaclub and AT Media as supported by the market investigation.169 Likewise, in the outdoor advertising market in the Czech Republic, there are three key media owners: JCDecaux, Euro AWK and Bigmedia170 and in general, outdoor media owners are constrained by the limited availability of advertising space and by the regulatory limitations as regards the choice of space for the outdoor media owners.

279.   The merged entity will therefore face strong media owners on the outdoor and TV media channels in the Czech Republic. Even though the Parties will represent a relatively important share of the sales of media owners, the countervailing power on the   media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

280.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in the Czech Republic.

IV.4.1.4.3.   Denmark

a.      Sales side

281.   In 2012, the Parties’ combined market share in MBS in Denmark amounted to [10-20]% including direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1 above)

282.   The Commission considers that in the particular case of Denmark, the assessment of the competitive position of the merged entity on the MBS market excluding direct sales is not appropriate, considering the high level of direct sales in this country (except for TV– see paragraph 34 above). The Commission therefore conducts its assessment on a market including direct sales only.171

283.   The Commission further notes that even if direct sales were to be considered as being outside of the relevant market, their high level in Denmark would likely exercise a further significant competitive constraint on the merged entity post-Transaction.

284.   Post-Transaction, the merged entity will continue to face competition from the four other large advertising networks172, WPP, Dentsu-Aegis, IPG and Havas, as well as from a number of independent players.173 Several of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

285.   The analysis of RECMA Compitches data indicates that the number of participants in Denmark was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.174 In 60% of overlap tenders there were four or more participants compared to 29% for the rest of the tenders in the country.175

286.   Therefore, the merged entity will continue to face competition from a number of competitors in the tenders in which it will take part post-Transaction.

287.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data for Denmark shows that, on average, in tenders where it was incumbent, Omnicom won in 23% of the cases; and in tenders where it was incumbent, Publicis did not win any tenders.176

288.   According to the analysis of the Parties' participation in 2010-2012 tenders in Denmark in the RECMA Compitches data, Publicis and Omnicom overlapped in 10% of tenders. Out of the tenders in which Omnicom participated, Publicis was also present in 17% of the cases. Out of the tenders in which Publicis participated, Omnicom was present in 56% of the cases.177 The most prominent competitors in all tenders in which Omnicom participated in Denmark were WPP first and Dentsu-Aegis second. Likewise, in all tenders in which Publicis participated in Denmark, WPP was first, followed by Omnicom. 178 In tenders in which Publicis was the incumbent, it lost to Omnicom, WPP and Dentsu-Aegis. In tenders in which Omnicom was the incumbent, it lost to Dentsu- Aegis first, followed by IPG, and WPP. Omnicom won its tenders from Publicis, but Publicis won its tenders from WPP, IPG and Havas.179

289.   Moreover, a majority of market participants that responded to the market investigation consider that Publicis' closest competitor is WPP, followed by Dentsu-Aegis. Likewise, Omnicom's closest competitor is considered to be WPP, followed by Dentsu-Aegis.180 Therefore, Publicis and Omnicom are not each other's closest competitors in Denmark

290.   Finally, a majority of customers that responded to the market investigation did not raise any country-specific issue as regards Denmark that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.181

291.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Denmark.

b.      Procurement side

292.   Table 8 below presents market shares of the Parties on the procurement side of the MBS market in Denmark:

Table 8 - Denmark - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time182)

293.   As can be seen in Table 8, the TV, outdoor and cinema sectors are affected by the Transaction. The other sectors, namely print, radio and online/mobile advertising are not affected by the Transaction. They are therefore not further discussed in this section.

294.   Outdoor. The outdoor advertising segment is affected by the Transaction. This combined market share is, however, small ([10-30]%). The Danish outdoor advertising market is also led by two players: AFA JCDecaux, and Clear Channel Denmark, which together have more than [90-100]% of the market.183

295.   Television. The TV advertising segment is affected by the Transaction when considering market shares including direct sales.184 The merged entity has a combined market share close to [30-50]%. However, the Danish TV advertising segment is led by two public broadcasting groups, TV-2-Danmark and the Danish Broadcasting Company (DR), with a combined TV audience share of nearly 70% in 2011. The main private broadcasting groups are Modern Time Group (9.6%), SBS (7.0%), formerly owned by ProSiebenSat1 Media AG, and The Walt Disney Company (2.7%). TV-2-Danmark is the clear leader in advertising sales with at least 67% of television ads.

296.   Cinema. Likewise, the cinema segment is also concentrated in Denmark, with two main players: (i) Nordisk Film, which accounts for 45% of the market share in terms of tickets sold;185 and (ii) Cinemaxx.

297.   The merged entity will face countervailing power by media owners on the outdoor, TV and cinema media channels in Denmark. Even though, in certain of these channels, the Parties will represent a relatively important share of the sales of media owners, the countervailing power of the media owners will be significant enough to offset any increase in negotiation power of the merged entity.

298.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Denmark.

IV.4.1.4.4.   France

a.      Sales side

299.   In 2012, the Parties’ combined market share in MBS in France amounted to [20-30]% including direct sales, and [20-30]-[20-30]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

300.   In France, the merged entity will face competition from the four other large advertising networks186, Dentsu-Aegis, Havas, WPP and IPG, as well as from a number of independent players.187 Several of these independent players have a number of large- scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

301.   The analysis of RECMA Compitches data indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.188 In 71% of the overlap tenders there were four or more participants compared to 26% for the other tenders in the country.189

302.   Therefore, the merged entity will continue to face competition from a number of competitors in the tenders in which it will take part post-Transaction.

303.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data shows that on average, Omnicom won tenders in 17% of the cases where it was the incumbent and Publicis in 57% of the cases.190 In addition to this, the Loi Sapin enables customers to know prices and rebates awarded to their MBS agency, and prohibits MBS agencies from aggregating the demand of several advertisers in their negotiations. This gives less weight to the negotiation power of the agency in the decision-making process of customers.

304.   According to the analysis of the Parties' participation in 2010-2012 tenders in France in the RECMA Compitches data, Publicis and Omnicom were both present in 13% of all tenders and in less than half of the tenders in which the other one was present: Publicis was present in 39% of all tenders in which Omnicom participated, and Omnicom in 25% of Publicis tenders.191 In addition, the Parties face other groups in tenders more often than each other. The most prominent competitors in the tenders in which Omnicom participated were WPP first and Dentsu-Aegis second. Likewise, in tenders in which Publicis participated, WPP and Dentsu-Aegis appeared the most often, followed by Havas.192 Omnicom's customers that switched agency, mainly switched to WPP and Publicis. Moreover, the data does not show that any client switched from Publicis to Omnicom. Omnicom won tenders from Dentsu-Aegis and independent competitor MyMedia, while Publicis won tenders first from Dentsu-Aegis, and second from Omnicom.193

305.   Moreover, a majority of market participants that replied to the market investigation consider that Publicis' closest competitors are WPP and Havas, and Omnicom's closest competitors are WPP and Havas.194

306.   In view of the above, Publicis and Omnicom are not each other's closest competitors in France.

307.   Finally, a majority of customers that responded to the market investigation did not raise any country-specific issue as regards France that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.195

308.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in France.

b.      Procurement side

309.   Table 9 below presents the market shares of the Parties on the procurement side of the MBS market in France by media channel:19

Table 9: France - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time197)

310.   As can be seen in Table 9, the print, online/mobile and cinema sectors are not affected by the Transaction. They are therefore not further discussed in this section.

311.   The TV sector in France is concentrated with three players representing approximately 60% of the market. The state-owned group France Télévisions include two of these main TV channels (France 2 with 14.9%, France 3 with 9.7%). TF1 remains the most popular channel with a daily audience share of 22.7% in 2012; M6 follows with 11.2%.198 TF1 concentrates more than 50% of advertising expenditure on TV, and M6 obtains approximately 25% of advertising expenditure.199 Since 2009, advertising is banned after 20h00 on France 2 and France 3 and more generally on all state-owned television channels. This is the peak audience slot for advertisers, making the ad time available on the private channels even more scarce and valuable.

312.   One respondent to the market investigation underlined the strong position of TF1 as compared to the other media owners on the French TV market. The respondent considered that the merged entity may be prevented from using its strong position to dictate prices to TF1 but will be able to do so for the other TV owners, strengthening in this way the leading position of TF1. The Commission considers, however, that even though the Parties will represent a relatively important share of the sales of media owners such as TF1 and the other TV broadcasters, the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

313.   The radio and outdoor advertising channels are also concentrated, since four players active in the radio sector together represent 80% of advertising sales200 while three main operators, JCDecaux, Clear Channel and CBS outdoor control most of the outdoor advertising space in France.201 The market investigation also indicated that media buying agencies cannot aggregate media time/space for each advertiser individually because of the Loi Sapin.202

314.   In light of paragraph 313 above, the Commission therefore considers that the merged entity will face strong media owners on the radio, and outdoor channels in France. The countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

315.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in France.

IV.4.1.4.5.   Germany

a.      Sales side

316.   In 2012, the Parties’ combined market share in MBS in Germany amounted to [10-20]% including direct sales, and [10-20]-[10-20]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

317.   In Germany, the merged entity will face competition from the four other large advertising networks203, WPP, Dentsu-Aegis, IPG, Havas, as well as from a number of independent players such as Mediaplus, Pilot Media, Crossmedia and Moccamedia. Several of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

318.   The analysis of RECMA Compitches data for Germany indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.204 Indeed, in 86% of overlap tenders, there was four or more participants compared to 25% for the other tenders in the country.205

319.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data for Germany shows that on average, in tenders where they were incumbents, Publicis won in only 20% of the cases while Omnicom won in only 33% of the cases.206

320.   According to the analysis of the Parties' participation in 2010-2012 tenders in Germany in the RECMA Compitches data, Publicis and Omnicom overlapped in tenders to a limited extent (18% of all tenders). Publicis and Omnicom were present in less than half of the tenders in which the other was present (44% and 45% respectively)207. In addition, the most prominent competitors in the tenders in which Omnicom and Publicis participated in Germany were WPP first and Dentsu-Aegis second.208 In tenders in which they were the incumbents, Publicis mainly lost tenders to other international groups such as WPP followed by IPG, while Omnicom lost tenders to independents, WPP, Dentsu-Aegis and IPG. When Omnicom was not the incumbent, it mainly won from WPP followed by Dentsu-Aegis, Publicis and other independent competitors. When Publicis was not the incumbent, it mainly won tenders from WPP.209

321.   A majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Omnicom. Omnicom's closest competitor is also WPP, followed by Dentsu-Aegis.

322.   In view of the above, Publicis and Omnicom are not each other's closest competitors in Germany.

323.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Germany that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction in particular also for large-scale advertisers.210

324.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Germany.

b.      Procurement side

325.   Table 10 below presents market shares of the Parties on the procurement side of the MBS market in Germany211:

Table 10: Germany - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time212)

326.   As can be seen in Table 10, the following media channels are affected by the Transaction in Germany: TV, outdoor and radio. Considering the level of the combined market shares and the fact that other competing MBS agencies that could service these segments are present in Germany, it is unlikely that competition concerns will arise as a result of the Transaction.

327.   Some respondents to the Commission's market investigation mentioned that the Transaction will lead to a substantial increase in bargaining power of the merged entity vis-à-vis media owners and that it is doubtful whether the merged entity will eventually pass on the benefits of their increased power (in the form of better prices) to their customers (the advertisers). These respondents mentioned in particular the TV, radio and print segments. In particular they stated that a significant share of advertisement is being purchased by the top three/five agencies (for TV, more than [80-90]% and for print more than [50-60]%). These respondents also indicated that the merged entity will have a combined market share of [20-30]-[30-40]% on the media procurement market in Germany (on the basis of RECMA estimations). They also mentioned that the Transaction will widen the gap between the agencies at the top of the market and the rest of the agencies.

328.   The Commission considers that even though post-Transaction, the Parties will represent a higher share of the sales of media owner, the countervailing power on the media owners' side will be significant enough to offset any increase in negotiation power of the merged entity

329.   The market for TV advertisement in Germany is characterised by a duopoly of the two private broadcasters groups RTL and Pro7Sat.1213 which together have a combined market share of between 80-90%.214

330.   As regards radio advertisement in Germany, advertisers and their agencies do not source radio advertising time from a multitude of radio stations but rather buy “bundled” radio advertising time that covers various regions or the entire country.215 The two most important players offering this kind of coverage in Germany are RMS Radio Marketing Service GmbH & Co. KG (RMS) and ARD Werbung Sales & Services GmbH (AS&S) whose joint share in the radio advertising segment is estimated to be approximately 90%.216

331.   The outdoor advertising market is likewise led by two main operators: Stroer and JCDecaux, which control most of the outdoor advertising space sold in Germany.217

332.   Lastly, the print segment is rather concentrated in Germany with four main players: Gruner+Jahr, Axel Springer, Bauer and Burda218

333.   The Commission therefore considers that it is unlikely that the merged entity will have significant buyer power on the market for buying advertisement in these media channels in Germany. In addition, the four other international players, WPP, Dentsu-Aegis, IPG and Havas, will continue to be present in Germany in addition to the merged entity so that advertisers will continue to have a sufficient number of agencies to choose from.

334.   Furthermore, the Parties explained that in Germany, the negotiation for TV space and radio time is negotiated during annual deal rounds between November and January as annual commitments must be finalised by the end of January. Time is sold through a rate card and discount mechanism, the rate cards being published by the TV and radio channels themselves. The discount off the rate card is secured on the basis of volume commitment and other qualitative parameters. The agency then buys airtime using the rate card less the negotiated discounts. In 2007, the Bundeskartellamt sanctioned the two leading providers of TV advertising (IP Deutschland GmbH active for RTL and SevenOne Media GmbH active for Pro7Sat.1) for anticompetitive discount agreements in the form of proportional volume discounts based on advertising budget and retroactive sliding scale discount agreements.219 Since then, the Parties explain that these types of discounts are no longer in place, as IP Deutschland GmbH and SevenOne’s discount systems are based on level of spend rather than share of volume. In  addition, following the Bundeskartellamt’s decision, the two sales houses no longer offer sliding scale arrangements and only negotiate on a fixed basis of volume.220

335.   Negotiations in Germany are done on a client per client basis and most contracts are customer specific, meaning they include terms and rebates that are specific to a particular advertiser.221 Even if in Germany individual agencies may aggregate their agency’s overall volume in order to negotiate free spots and AVBs222, a large proportion of AVBs is passed on to clients due to growing client awareness and pressure. Most client contracts ensure transparency regarding the level of AVBs to be returned to the client, in particular through the client’s right to have the agency regularly audited by independent auditors. As a consequence, […].223 The other discounts achieved by media buying agencies are also passed on to their clients.224 In addition, when an agreement is reached to carry out joint negotiations (that is to say cross-agencies for some clients), it is because the clients concerned and their MBS agency consider that such a negotiation process allows for the purchase of media spaces at a better price and/or of a better quality. However, these business decisions are specific to certain clients and these types of negotiations represent a rather limited share of the Parties' media buying in Germany (below [10-20]%).225 The Commission considers that the Transaction is not likely to bring about any changes in this respect; customers (advertisers) are therefore likely to continue to benefit from any discounts that the agencies would achieve.

336.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Germany.

IV.4.1.4.6.   Greece

a.      Sales side

337.   In 2012, the Parties’ combined market share on the sales side of the MBS in Greece amounted to [5-10]% including direct sales, and [10-20]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

338.   The merged entity will continue to face competition from  the  four  other  large  advertising  networks226,  WPP227,  IPG,  Dentsu-Aegis  and  Havas,  as  well  as  from  a number of independent players.228 Several of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA- wide or global) reach.

339.   The analysis of RECMA Compitches data indicates that the number of participants in Greece was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.229 In 56% of overlap tenders there were four or more participants compared to 27% for the rest of the tenders in the country. There was also at least one other participant in 78% of overlap tenders. 230

340.   Therefore, the merged entity will continue to face competition from a number of competitors in the tenders in which it will take part post-Transaction.

341.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data for Greece shows that, on average, in tenders where it was incumbent, Omnicom won in only 33% of the cases and in tenders where it was incumbent, Publicis did not win any tenders.231

342.   According to the analysis of the Parties' participation in 2010-2012 tenders in Greece in the RECMA Compitches data, Publicis and Omnicom overlapped in 38% of tenders. In the tenders in which Omnicom participated, Publicis was also present in 42% of the cases. In the tenders in which Publicis participated, Omnicom was present in 100% of the cases.232 The most prominent competitors in all tenders in which Omnicom participated in Greece were IPG first and WPP second. In all tenders in which Publicis participated in Greece, Omnicom was first followed by IPG. 233 In tenders in which Publicis was the incumbent, it lost to Omnicom, IPG and Dentsu-Aegis. In tenders in which Omnicom was the incumbent, it lost to WPP and Publicis. Omnicom won its tenders against Publicis, WPP and independents.234

343.   Moreover, market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Omnicom. Omnicom's closest competitor is also WPP, followed by Dentsu-Aegis.235

344.   In view of the above, Publicis and Omnicom are somewhat closer competitors in Greece than in other countries. Other competitors are, however, also present in tenders in which Publicis and Omnicom participate.

345.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Greece that contradicts any of the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.236.

346.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Greece.

b.      Procurement side

347.   Table 11 below presents the market shares of the Parties on the procurement side of the MBS market in Greece by media channel237:

Table 11: Greece - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time238)

348.   As can be seen in Table 11, the print, online / mobile, outdoor, radio and cinema sectors are not affected by the Transaction when direct sales are included.

349.   Regarding TV, the combined market share of the Parties is small, at [10-30]%.

350.   Furthermore, the TV sector in Greece is concentrated with 4 players representing more than 60% of the market. MEGA is the most popular channel with a daily audience share of 20% in 2012; ANT1 follows with 17%, and Alpha TV and Star with 13% and 10% respectively. The total share of the public service channels in 2012 was 13.6%.239 The merged entity will therefore face strong media owners on the TV channel segment in Greece. One media owner expressed concerns as regards the Transaction in Greece, but these concerns were not substantiated and are not in line with the market share data of the Parties. The Commission considers in addition that the countervailing power on the media owners' side will be significant enough to offset any increase in negotiation power of the merged entity.

351.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Greece.

IV.4.1.4.7.   Hungary

a.      Sales side

352.   In 2012, the Parties’ combined market share in MBS in Hungary amounted to [10-20]% including direct sales, and [20-30]-[20-30]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

353.   The merged entity will face competition from the four other large advertising networks240, WPP, Dentsu-Aegis, IPG, Havas, and other independent players.241 Some of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

354.   The analysis of RECMA Compitches data for Hungary indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.242 In 50% of the overlap tenders there were four or more participants while this ratio was only 30% for the rest of the tenders.243

355.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data for Hungary shows that on average, in tenders where they were incumbents, Publicis won in 60% of the cases while Omnicom won in 50% of the cases.244

356.   According to the analysis of the Parties' participation in 2010-2012 tenders in Hungary in the RECMA Compitches data, Publicis and Omnicom overlapped in tenders to a limited extent (17% of all tenders). Publicis and Omnicom were present in less than half of the tenders in which the other was present (32% and 40% respectively).245 The most prominent competitors in all tenders in which Omnicom participated in Hungary were IPG and WPP. In the tenders in which Publicis participated in Hungary, IPG was first, followed by Omnicom246. In tenders in which they were the incumbents Publicis mainly lost tenders to WPP and IPG while Omnicom lost tenders to WPP, IPG and Havas. When Omnicom was not the incumbent, it mainly won from IPG, WPP and independents. Likewise when Publicis was not the incumbent, it mainly won from IPG.247

357.   A majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Dentsu-Aegis. Omnicom's closest competitor is also WPP, followed by IPG.248

358.   In view of the above, Publicis and Omnicom are not each other's closest competitors in Hungary.

359.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Hungary that contradicts any of the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.249

360.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Hungary.

b.      Procurement side

361.   Table 12 below presents market shares of the Parties on the procurement side of the MBS market in Hungary250:

Table 12: Hungary - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales – Shares of the Parties per cost of media time251)

362.   As can be seen in Table 12, the following media channels are affected by the Transaction in Hungary: TV and outdoor.

363.   The TV and outdoor segments are concentrated in Hungary.252 The TV advertising segment is led by the two private broadcasting groups, RTL Group (RTL Klub) and ProSiebenSat1 (TV2), followed by the public broadcaster, MTVA. RTL Klub remained the market leader in 2011 with 27.9% while TV2 had 19.5%. The public channels M1 and M2 had a combined market share in 2011 of 11.6%.253 Likewise, two main operators are present on the outdoor advertising segment, Publimont Kft and Clear Channel. JCDecaux re-entered the Hungarian outdoor market through the acquisition of Epamedia Hungary in 2012.254

364.   The merged entity will therefore face strong media owners on the TV and outdoor channels in Hungary. Even though the Parties will represent a relatively important share of the sales of media owners, the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

365.   Considering the level of the combined market shares and the fact that other competing MBS agencies that can service these segments are present in Hungary, it is unlikely that competition concerns will arise as a result of the Transaction.

366.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Hungary.

IV.4.1.4.8.   Ireland

a.      Sales side

367.   In 2012, the Parties’ combined market share in MBS in Ireland amounted to [10-20]% including direct sales, and [10-20]-[10-20]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

368.   The merged entity will face competition from the four other large advertising networks255, namely WPP, Dentsu-Aegis IPG and Havas256, as well as from a number of independent players257. Several of these independent players have a number of large- scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

369.   The analysis of RECMA Compitches data indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.258 In 33% of overlap tenders there were four or more participants compared to 21% of the rest of the tenders in the country.259

370.   Therefore, the merged entity will continue to face competition from a number of competitors in the tenders in which it will take part post-Transaction.

371.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data shows that Omnicom won no tenders where it was the incumbent and so did also Publicis260.

372.   According to the analysis of the Parties' participation in 2010-2012 tenders in Ireland in the RECMA Compitches data, Publicis and Omnicom were present in 9% of all tenders and in less than half of the tenders in which the other was present: Publicis was present in 21% of all tenders in which Omnicom participated, and Omnicom in 50% of Publicis tenders261. In addition, the Parties faced other groups in tenders more often than each other. The most prominent competitors in the tenders in which Omnicom participated was WPP first and Dentsu-Aegis second. In tenders in which Publicis participated, WPP and Omnicom appeared the most often262. Omnicom's customers that switched agency, mainly switched to WPP and IPG, Publicis' customers mainly switched to WPP, followed by IPG and Omnicom. Omnicom won tenders from WPP first followed by Publicis, while Publicis won tenders from WPP and from Dentsu-Aegis263.

373.   Moreover, a majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Dentsu-Aegis. Omnicom's closest competitor is also WPP, followed by Dentsu-Aegis.264

374.   In view of the above, Publicis and Omnicom are not each other's closest competitors in Ireland.

375.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Ireland that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction in particular also for large-scale advertisers.265

376.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Ireland.

b.  Procurement side

377.   Table 13 below presents the market shares of the Parties on the procurement side of the MBS market in Ireland by media channel266:

Table 13: Ireland - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time267)

378.   As can be seen in Table 13, only the outdoor channel is affected by the Transaction. TV, print, online / mobile, radio and cinema sectors are therefore not further discussed in this section.

379.   Ireland’s outdoor advertising segment is concentrated. Clear Channel Ireland, JCDecaux Ireland, CBS Outdoor and Bravo Outdoor together represent 98% of the outdoor advertising media owners throughout Ireland.268 They offer advertising space on buses, trains, bus shelters, billboards and other outdoor platforms.269 In view of the relatively low combined market share of the Parties and the presence of strong players among the media vendors, the Commission considers that the countervailing power on the media owners' side will be significant enough to offset any increase in negotiation power of the merged entity.

380.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Ireland.

IV.4.1.4.9.   Italy

a.      Sales side

381.   In 2012, the Parties’ combined market share in MBS in Italy amounted to [20-30]% including direct sales, and [20-30]-[30-40]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).270

382.   The merged entity will face competition from the four other large advertising networks271, WPP, Dentsu-Aegis, Havas, and IPG, plus from Media Italia and a number of independent players272. Some of these independent players have a number of large- scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

383.   The analysis of the RECMA Compitches data for Italy indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.273 In 93% of overlap tenders there were four or more participants while this ratio was only 47% for the rest of the tenders.274

384.   Moreover, customers have, and still have post-Transaction, the ability to switch to other MBS agencies. Indeed, the analysis of RECMA Compitches data for Italy shows that, on average, in tenders where it was incumbent, Publicis won in only 29% of the cases.275 No conclusions can be drawn with regard to Omnicom as the complete dataset included only two tenders in which Omnicom was the incumbent.276

385.   According to the analysis of the Parties' participation in 2010-2012 tenders in Italy in the RECMA Compitches data, Publicis and Omnicom overlapped in 31% of tenders. Publicis participated in more than half of the tenders in which Omnicom participated while Omnicom participated in around half of the tenders in which Publicis participated277. However, the most prominent competitors in all tenders in which Omnicom participated in Italy was WPP first and Dentsu-Aegis second. Likewise, in all tenders in which Publicis participated in Italy, WPP was first, followed by Dentsu-Aegis that came second.278 In tenders in which Publicis was the incumbent, it lost to WPP and Dentsu-Aegis.279 Publicis won its tenders mainly from WPP, and Omnicom won its tenders mainly from Dentsu Aegis.280

386.   A majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Dentsu-Aegis. Likewise, Omnicom's closest competitor is WPP, followed by Dentsu-Aegis.281

387.   In view of the above, Publicis and Omnicom are not each other's closest competitors in Italy.

388.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Italy that contradicts the conclusions outlined in   the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.282

389.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Italy.

b.      Procurement side

390.   Table 14 below presents market shares of the Parties on the procurement side of the MBS market in Italy283:

Table 14: Italy - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time285)

391.   As can be seen in Table 14, all media channels in Italy are affected by the Transaction.

392.   Cinema. Cinema owners are concentrated in Italy with two main cinema groups, UCI Cinemas and the Space, which together hold more than 40% of the cinema market286. The sale of advertising space is carried out by companies operating on behalf of the cinema owners. The market investigation indicated that there are currently three  main media suppliers in the cinema channel in Italy, namely MovieMedia, Rai Pubblicità and the Space.287

393.   Second, cinema owners or companies operating on behalf of cinema owners in Italy indicated in response to the market investigation that the merged entity will not be able to achieve buyer power to a level that it will be able to dictate its prices and other purchase conditions to media owners post-Transaction with the exception of one respondent which reasoned that such buyer power will lead to a decrease in prices.288 The majority of customers that require MBS services in Italy did not consider that the merged entity will significantly increase its buyer power289.

394.   The merged entity will therefore face strong media owners on the cinema channel in Italy. Even though the Parties will represent an important share of the sales of media owners, the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

395.   Other types of media: TV, print, online/mobile, outdoor and radio. The Parties' combined market shares on the procurement markets on the print, online/mobile and radio channels are below [10-30]%. The print and radio segments are characterised by a few leading players while the online/mobile segment is led by Google.290

396.   On the TV and outdoor channels, the market shares are [10-30]% and [10-30]% respectively. However, media owners are concentrated in Italy.291 Indeed, the TV advertising segment is led by the private broadcasting group Mediaset and the public service broadcaster RAI. In the last three years, these two groups have had a combined share of over 78% of total TV advertising sales in Italy, with the four main players (RAI

and  Mediaset  plus  Sky  and  Telecom  Italia)  holding  a  total  share  of  above    87%.292

Likewise, the outdoor advertising segment in Italy is led by four main operators: IGPDecaux (27%), Clear Channel (16%), IPAS (10%) and CBS Outdoor (7%).293

397.   The merged entity will therefore face strong media owners on the TV, print, online/mobile, outdoor and radio channels in Italy. The countervailing power on the media owners' side will be significant enough to offset any increase in negotiation power of the merged entity.

398.   Considering the level of the combined market shares, the fact that other competing MBS agencies that can service these segments are present in Italy and the level of concentration of media owners, it is unlikely that competition concerns will arise as a result of the Transaction.

399.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Italy.

IV.4.1.4.10.      Latvia

a.      Sales side

400.   In 2012, the Parties’ combined market share in MBS in Latvia amounted to [20-30]% including direct sales, and [20-30]-[20-30]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

401.   The merged entity will face competition from three out of the four other large advertising networks294, Dentsu-Aegis, Havas and IPG, as well as from other independent players such as Creative Media Services295. Some of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

402.   Customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of the Parties' internal data for Latvia indicates that, [Summary of Publicis and Omnicom's win, loss and incumbency data].296

403.   A majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by IPG while Omnicom's closest competitor is WPP, followed by IPG.297

404.   In view of the above, Omnicom and Publicis are not each other's closest competitors in Latvia.

405.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Latvia that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction in particular also for large-scale advertisers.298

406.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Latvia.

b.      Procurement side

407.   Table 15 below presents market shares of the Parties on the procurement side of the MBS market in Latvia299:

Table 15: Latvia - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time300)

408.   As can be seen in Table 15, the following media channels are affected by the proposed Transaction in Latvia: TV, online/mobile, outdoor, radio and cinema.

409.   The Parties' combined market share on the procurement markets on the radio channel was [10-30]%. Considering the small market shares of the Parties in this affected market, the Transaction is unlikely to raise any serious doubts as to its compatibility with the internal market on this segment.

410.   TV. On the TV, online/mobile, outdoor and cinema channels, the market shares were [10-30]%, [10-30]%, [10-30]% and [10-30]%, respectively. However, there are a number of strong media owners in Latvia.301 Indeed, the TV advertising includes a number of strong players, including the three private channels (LNT, TV3 and PBK) and the  public  channel  LTV1.  Together  they  held  a  46%  audience  share  in  2011.302 Additionally, in early 2012, Latvia's two largest commercial TV broadcasters merged (MTG and LNT).303

411.   Other types of media: online/mobile and outdoor. Likewise, the online/mobile and outdoor channels are concentrated in Latvia. Google is strengthening its position in Latvia in online advertising,304 particularly in online search advertising, while the outdoor advertising segment is led by JCDecaux, which appears to be the leading outdoor advertising operator in this country,305 followed by Clear Channel, EuroAWK, Pils• tas L•nijas and Tilts Media. Lastly, there are four main players in the cinema channel, namely Forum Cinemas, Multikino, Cinamon and Silver Screen, which together represent a significant proportion of the available ad time in Latvia.306 In addition, cinema ad spend is small and no significant developments are forecasted.307

412.   The merged entity will therefore face strong media owners on the TV, online/mobile, outdoor and cinema channels in Latvia. Even though the Parties will represent a relatively important share of the sales of media owners, the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

413.   Considering the level of the combined market shares, the fact that other competing MBS agencies that can service these segments are present in Latvia and the level of concentration of media owners, it is unlikely that competition concerns will arise as a result of the Transaction.

414.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Latvia.

IV.4.1.4.11.      Lithuania

a.      Sales side

415.   In 2012, the Parties’ combined market share in MBS in Lithuania amounted to [10-20]% including direct sales, and [10-20]-[10-20]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

416.   The merged entity will face competition from three of the four other larger advertising networks308, Dentsu-Aegis, IPG, and Havas, as well as from a number of independent players309. Some of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

417.   Customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of the Parties' internal data for Lithuania indicates that [Summary  of Publicis and Omnicom's win, loss and incumbency  data]. 310

418.   A majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Dentsu-Aegis, while Omnicom's closest competitor is WPP, followed by IPG.311

419.   In view of the above, Omnicom and Publicis are not each other's closest competitors in Lithuania.

420.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Lithuania that contradicts any of the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.312

421.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Lithuania.

b.      Procurement side

422.   Table 16 below presents market shares of the Parties on the procurement side of the MBS market in Lithuania313:

Table 16: Lithuania - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales – Shares of the Parties per cost of media time314)

423.   As can be seen in Table 16, the following media channels are affected by the Transaction in Lithuania: TV, online/mobile, outdoor and cinema.

424.   Outdoor. The outdoor advertising space is led by two main operators, Clear Channel Lithuania and JCDecaux Lietuva UAB. In general, outdoor media owners are constrained by the limited availability of advertising space and by regulatory limitations as regards the choice of space.

425.   The merged entity will therefore face strong media owners on the outdoor channel in Lithuania. Even though the Parties will represent a relatively important share of the sales of these two media owners, the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

426.   Other types of media: TV, online/mobile and cinema. The Parties' combined market shares on the procurement markets on the online/mobile and cinema channels are below [10-30]% while on the TV channel they amount to [10-30]%.

427.   The nationwide segment for TV advertising is led by the private broadcaster group MTG (which owns 32 channels, including TV3), MG Baltic Group and UAB Koncernas Achemos Grupe. Together they accounted for approximately 47% of the audience share in 2011.315 In Lithuania, online advertising is led by Google particularly in relation to online search advertising.316 The cinema advertising segment is led by Forum Cinemas, which owns 6 multiplex cinemas across the country. Other international media chains active in Lithuania are Multikino and Cinamon which owns one multiplex each. Forum Cinemas is a leader in both film distribution and film exhibition. In addition, the advertising space of both Forum Cinemas and Cinamon is sold through the intermediary of the same sales house, Pirmalinja, active in all three Baltic countries.317

428.   As evidenced above, these media channels are concentrated in Lithuania.318 The countervailing power on the media owners' side will be significant enough to offset any increase in negotiation power of the merged entity.

429.   Considering the level of the combined market shares and the fact that other competing MBS agencies that can service these segments are present in Lithuania, it is unlikely that competition concerns will arise as a result of the Transaction.

430.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Lithuania.

IV.4.1.4.12.      The Netherlands

a.      Sales side

431.   In 2012, the Parties’ combined market share in MBS in the Netherlands amounted to [5- 10]% including direct sales, and [10-20]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

432.   The merged entity will continue to face competition from the four other large advertising networks319, WPP, IPG, Dentsu-Aegis and Havas, as well as from a number of independent players320. Several of these independent players have a number of large- scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

433.   The analysis of RECMA Compitches data indicates that the number of participants in the Netherlands was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders321. In all overlap tenders, there were 4 or more participants compared to 27% for the rest of the tenders in the country. There is also at least one other participant in all (100%) overlap tenders.322

434.   Therefore, the merged entity will continue to face competition from a number of competitors in the tenders in which it will take part post-Transaction.

435.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data for the Netherlands shows that in tenders where it was incumbent, Publicis did not win any tenders. No conclusions can be drawn with regard to tenders in which Omnicom was the incumbent as the complete dataset includes only two tenders in which Omnicom was the incumbent.323

436.   According to the analysis of the Parties' participation in 2010-2012 tenders in the Netherlands in the RECMA Compitches data, Publicis and Omnicom overlapped in 3% of the tenders. Out of the tenders in which Omnicom participated, Publicis was also present in 7% of the cases. Out of the tenders in which Publicis participated, Omnicom was also present in 7% of the cases.324 The most prominent competitors in all tenders in which Omnicom participated in the Netherlands was WPP first and Dentsu-Aegis second. Likewise, in all tenders in which Publicis participated in the Netherlands, WPP was first followed by IPG.325 In tenders in which Publicis was the incumbent, it lost to Havas, IPG, and WPP, while Omnicom lost tenders in which it was the incumbent to WPP and Dentsu-Aegis. Publicis won its tenders against WPP, Dentsu-Aegis and IPG. Omnicom won tenders from Dentsu-Aegis, IPG and an independent agency.

437.   Moreover, market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by IPG and Omnicom. Omnicom's closest competitor is also WPP, followed by Dentsu-Aegis.326

438.   The Parties are therefore not each other's closest competitors in the Netherlands.

439.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards the Netherlands that contradicts the conclusions outlined in the general section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.327

440.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in the Netherlands.

b.      Procurement side

441.   Table 17 below presents market shares of the Parties on the procurement side of the MBS market in the Netherlands328:

Table 17 - the Netherlands- MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time329)

442.   As can be seen in Table 17, the print, online / mobile and cinema channels are not affected by the Transaction. They are therefore not further discussed in this section.

443.   The TV, outdoor and radio channels are affected by the Transaction. However, the combined market shares of the Parties ([10-30]% for each affected media, respectively), is rather small, and it is unlikely that competition concerns will arise as a result of the Transaction, given the presence of competing MBS agencies in the country.

444.   Given the low combined market share of the Parties and the concentration of media vendors in the TV, radio and outdoor media respectively, the vendors of advertising space or time in these media in the Netherlands will be in a position to continue to exert a sufficient level of bargaining power versus the merged entity.

445.   Notably, the nation-wide market for TV advertising is led by two private broadcaster groups, RTL Group and Sanoma, and one public broadcaster, NBO, with a combined market share of 73%.330

446.   Radio advertising is also concentrated, with four groups, the public Nederlandse Publieke Omroep (NPO), Talpa Media, Telegraaf Media Group and De Persgroep, with a combined market share of 72% in 2012.331

447.   Finally, the outdoor advertising space is led in the Netherlands by three large international groups: CBS Outdoor, Clear Channel, and JCDecaux. Among them, CBS Outdoor is the market leader and has a market share of over 50% in outdoor advertising.332

448.   Therefore, the merged entity will face strong media owners on the outdoor, TV and radio  media  channels in  the  Netherlands.  The  countervailing power  on  the media owners' side will be significant enough to offset any increase in negotiation power of the merged entity.

449.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in the Netherlands.

IV.4.1.4.13.      Poland

a.      Sales side

450.   In 2012, the Parties’ combined market share in MBS in Poland amounted to [30-40]% including direct sales, and [30-40]-[30-40]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

451.   The merged entity will face competition from the four other large advertising networks333, WPP, IPG, Dentsu-Aegis and Havas, as well as from a number of independent players334. Several of these independent players have a number of large- scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

452.   The analysis of RECMA Compitches data for Poland indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.335 Indeed, in 53% of overlap tenders there were four or more participants compared to 30% for the other tenders in the country.336

453.   WPP participated between 2010 and 2012 in 76% of the tenders in the country. Moreover, the participation for IPG, Dentsu-Aegis and Havas increased significantly between 2010 and 2012.

454.   Therefore, the merged entity will continue to face competition from a number of competitors in the tenders in which it will take part post-Transaction.

455.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data shows that on average, Omnicom won tenders in 27% of the cases where it was the incumbent (through the same agency or another agency within the group) and Publicis in 41% of the cases.337

456.   According to the analysis of the Parties' participation in 2010-2012 tenders in Poland in the RECMA Compitches data, Publicis and Omnicom were both present in 29% of all tenders. Publicis was present in 63% of the tenders in which Omnicom participated. Conversely, Omnicom was present in 44% of the tenders in which Publicis participated338. In addition, the most prominent competitors in the tenders in which Omnicom participated were WPP followed by Publicis. Likewise, in tenders in which Publicis participated, WPP and Omnicom were first and second respectively.339 Omnicom's customers that switched agency, mainly switched to WPP and Publicis. Publicis' customers which switched agency, switched to WPP first, then equally to Omnicom and IPG. Omnicom won tenders from Publicis and WPP, while Publicis won tenders first from Omnicom, and second from IPG and Havas.340

457.   Nevertheless, a majority of market participants that replied to the market investigation considered that WPP is the closest competitor of Publicis, followed by IPG. Omnicom's closest competitor is also WPP, followed by Publicis.341

458.   In view of the above, the Commission considers that WPP, Publicis and Omnicom appear to be equally close competitors on the Polish market.

459.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Poland that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.342

460.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Poland.

b.      Procurement side

461.   Table 18 below presents the market shares of the Parties on the procurement side of the MBS market in Poland by media channel34

Table 18: Poland - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales – Shares of the Parties per cost of media time344)

462.   As can be seen in Table 18, all types of media are affected by the Transaction.

463.   TV. Certain TV media suppliers in Poland raised the concern that the Transaction will strengthen the merged entity's bargaining power and therefore increase the price pressure on media owners345. However, the media suppliers that replied to the market investigation indicated that the Polish TV segment is concentrated.346

464.   Indeed, the TV sector in Poland is concentrated with three players representing around 80% of the prime time viewing. The public service broadcaster group TVP accounts for 36.5% of audience share, the private broadcaster groups TVN and Polsat have respectively 24% and 20% of audience share.347

465.   Even though the Parties will represent an important share of the media owners' sales, in view of the above, the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

466.   Outdoor. There are two main operators in the outdoor advertising sector in Poland: AMS and Stroer Group, with shares of respectively 32% and 29%. Cityboard Media and Clear Channel are two other significant players with respectively 13% and 10% market shares. JCDecaux is also present in Poland, and reinforced its position on the Polish market by acquiring Gigaboard Polska in January 2012.348 The media owners that replied to the market investigation indicated that the outdoor segment is concentrated.349

467.   The merged entity will therefore face strong media owners on the outdoor channel in Poland. Even though the Parties will represent a relatively important share of the sales of media owners, the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

468.   Other types of media: print, radio, online/mobile and cinema. The three main publisher groups (Springer, owner of Fakt, Agora, owner of Gazeta Wyborcza and Metro ZPR, owner of Super Express) represent more than 50% of the daily newspaper ownership.350 Three main publisher groups are present on the magazine sector, namely Springer, Bauer Media and Gruner + Jahr, each of them owning several magazines.351 In the radio sector, three entities (Bauer, which owns Radio RMF and is also active in the magazine sector, Polskie Radio, and Eurozet LTD) have a combined audience market share of 61%.352 Online advertising is led by Google, particularly in relation to online search advertising. Google’s search engine represents 97.5% of user share.353 In the cinema sector, 80% of the cinemas are held by big multiplex cinema chains, among which the leaders are Cinema City Poland, Helios Film Center, and Multikino.354 The merged entity will therefore face strong media owners on the print, radio, online/mobile and cinema channels in Poland.

469.   The countervailing power on the side media owners in the print, radio, online/mobile and cinema channels is significant enough to offset any increase in negotiation power of the merged entity.

470.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Poland.

IV.4.1.4.14.      Portugal

a.      Sales side

471.   In 2012, the Parties’ combined market share in MBS in Portugal amounted to [20-30]% including direct sales, and [20-30]-[20-30]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

472.   The merged entity will face competition from the four other large advertising networks355, WPP, Havas, IPG, Dentsu-Aegis, as well as a number of independent players such as Nova Expressao and Executive Media356. Some of these   independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

473.   The analysis of RECMA Compitches data for Portugal indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.357 In all overlap tenders, there were four or more participants while this ratio was only 37% for the rest of tenders.358

474.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. Indeed, the analysis of RECMA Compitches data for Portugal shows that on average, in tenders where it was the incumbent, Omnicom won tenders in 33% of the cases.359 No conclusions can be drawn with regard to Publics as the complete dataset includes only two tenders in which Publicis was the incumbent.360

475.   According to the analysis of the Parties' participation in 2010-2012 tenders in Portugal in the RECMA Compitches data, Publicis and Omnicom overlapped in tenders to a limited extent (10% of all tenders). Publicis and Omnicom were present in less than half of the tenders in which the other was present (25% and 33% respectively).361 In addition, the most prominent competitors in the tenders in which Omnicom participated in Portugal were WPP and Havas, followed by Dentsu-Aegis. Likewise, in tenders in which Publicis participated, Havas was first, followed by WPP and IPG362. In tenders in which Omnicom was the incumbent, it lost to Dentsu-Aegis, IPG, Havas and independents.363 Publicis won tenders from Dentsu-Aegis and Havas. 364

476.   A majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by IPG. Likewise, Omnicom's closest competitor is also WPP, followed by IPG.365

477.   In view of the above, Publicis and Omnicom are not each other's closest competitors in Portugal.

478.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Portugal that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.366

479.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Portugal.

b.      Procurement side

480.   Table 19 below presents market shares of the Parties on the procurement side of the MBS market in Portugal367

Table 19: Portugal - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales – Shares of the Parties per cost of media time368)

481.   As can be seen in Table 19, the following media channels are affected by the Transaction in Portugal: TV, print, online/mobile and outdoor.

482.   TV. On the one hand, the nationwide market for TV advertising is led by the public broadcaster RTP and the privately-owned groups Grupo Impresa (SIC portfolio) and Media Capital (the owner of Grupo Prisa, which operate the leading commercial television channel TVI, and the radio stations Rádio Comercial, Rádio Clube Português and Cidade FM). Together they accounted for approximately 81% of the audience share in 2011. On the other hand, the end users of the media time are advertisers and no single advertiser accounts for more than 5% of demand. 369

483.   Print. The print segment is likewise concentrated in Portugal. The main publishers are the following: Controlinveste (which owns Jornal de Notícias, Diário de Notícias,   24 Horas, O Jogo, National Geographic and several other specialised TV, a cable television channel), Cofina (which owns the daily newspapers Correio da Manhã, Jornal de Negócios and Record, the free newspapers Destak and Meia-Hora and the magazines Sábado, a news magazine, TV Guia and several other specialised publications), Impresa (the owner of Expresso, Visão, Jornal de Letras, Exame, Telenovelas, Caras and half a dozen specialised magazines), Impala (which owns, inter alia, focus, Maria, Eco, In7Dias, VIP and Gente).370 Together, these main players account for 79% of the Portuguese print segment.371

484.   Outdoor. The outdoor advertising space is led by JCDecaux Portugal, MOP (Multimédia Outdoors Portugal), Cemusa Portugal (Companhia de Mobiliario Urbano e Publicidade), Spetacolor Portugal (Publicidade Informatizada), and PD (Publicidade Dinâmica) (APSmedia). Together JCDecaux, MOP and Cemusa Portugal account for around [90- 100]% of the Portuguese outdoor advertising segment.372 The strength of outdoor advertising space providers also rests on the limited availability of desirable ad space.373

485.   Online. The online advertising segment is led by Google, whose search engine represents approximately 97% of user share.

486.   As evidenced above, these media segments are concentrated in Portugal.374 The countervailing power on the media owners' side will be significant enough to offset any increase in negotiation power of the merged entity.

487.   Considering the level of the combined market shares and the fact that other competing MBS agencies that could service these segments are present in Portugal, it is unlikely that competition concerns will arise as a result of the Transaction.

488.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Portugal.

IV.4.1.4.15.      Romania

a.      Sales side

489.   In 2012, the Parties’ combined market share in MBS in Romania amounted to [30-40]% including direct sales, and [30-40]-[30-40]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1). 

490.   The merged entity will face competition from the four other large advertising networks375, WPP, IPG, Dentsu-Aegis and Havas, as well as from a number of independent players such as Media Investment, United Media, Media-Tique and House of Media.376 Several of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

491.   The analysis of the RECMA Compitches data for Romania indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.377 Indeed, in 83% of overlap tenders, there were four or more participants while this ratio was only 9% for the rest of the tenders.378

492.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data for Romania shows that on average,379 in tenders where they were incumbents, Publicis won in only 14% of the cases while Omnicom won in only 20% of the cases.380

493.   According to the analysis of the Parties' participation in 2010-2012 tenders in Romania in the RECMA Compitches data, Publicis and Omnicom overlapped in tenders to a limited extent (15% of all tenders). Publicis and Omnicom were present in less than half of the tenders in which the other was present (38% and 33% respectively)381. In addition, the most prominent competitors in the tenders in which Omnicom participated in Romania were WPP first and IPG second. Likewise, in tenders in which Publicis participated, WPP was first, followed by Omnicom.382 In tenders in which they were the incumbents, Publicis mainly lost tenders to other international groups such as Dentsu- Aegis and Omnicom while Omnicom mainly lost tenders to independents as well as to WPP and IPG. When Omnicom was not the incumbent, it mainly won from Publicis and independent competitors. On the contrary, when Publicis was not the incumbent, it mainly won tenders from IPG and Dentsu-Aegis.383

494.   A majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP followed by IPG. On the other hand they consider that Omnicom's closest competitor is Publicis, followed by WPP.384

495.   In view of the above, WPP, Publicis, Omnicom, and IPG, appear to be equally close competitors in Romania.

496.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Romania that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.385

497.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Romania.

b.      Procurement side

498.   Table 20 below presents market shares of the Parties on the procurement side of the MBS market in Romania386:

Table 20: Romania - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales – Shares of the Parties per cost of media time387)

499.   As can be seen in Table 20, the following media channels are affected by the Transaction in Romania: TV, print, online/mobile, outdoor and radio

500.   TV and print. First, TV and print owners are concentrated in Romania.The TV advertising segment is led by the CME group (Central European Media Enterprises), the Intact Group and ProSieben which together account for approximately 82% of the total net TV ad spend. Likewise, the print advertising segment is led by six big   publishing groups (Mediafax/Publimedia, Adevarul Holding, Intact Group, Ringier, Sanoma Hearst Romania and Medien Holding) which together have a combined revenue share of 63%.

501.   Second, as of April 2013, Romanian law prohibits any acquisition of advertising space by MBS agencies unless the latter act in the name and on behalf of their advertiser clients. Moreover, advertisers can buy TV advertising only directly from media owners even if they are using the services of an MBS provider388. In particular, the price for the purchase of advertising space on TV must be paid by the advertiser directly to the media owner, based on the invoice to be issued directly to advertisers by the media owner. The invoice must include any rebate or financial benefit granted by the media owner, in order to end the former practice of television broadcasters granting yearly rebate to MBS agencies. In Romania, thus, no rebates and/or other material incentives can be paid by TV owners to MBS providers.

502.   Therefore, even though the Parties will represent a relatively important share of the sales of media owners, the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

503.   Other types of media: radio, outdoor, online/mobile. The Parties' combined market share on the procurement markets on all the other affected channels is equal to or below [10-30]%. These media channels are concentrated in Romania.389 The radio advertisement segment is led by MGSI, Clier Media and Regie Radio Music which together hold more than 90% of the total radio market. There are four main operators in the outdoor segment in Romania: EpaMedia, Affichage Romania, Clear Channel and New Outdoor Romania which together have a share of 67%.390 Google leads the online search engine segment with a user share of 97% in Romania.391 The countervailing power on the media owners' side will be significant enough to offset any increase in negotiation power of the merged entity.

504.   Considering the level of the combined market shares and the fact that other competing MBS agencies that could service these segments are present in Romania, it is unlikely that competition concerns will arise as a result of the Transaction.

505.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Romania.

IV.4.1.4.16.      Slovenia

a.      Sales side

506.   In 2012, the Parties’ combined market share in MBS in Slovenia amounted to [5-10]% including direct sales, and [5-10]% excluding direct sales, according to the Parties' estimates of market shares based on the WARC market size. If the Parties' combined market shares are calculated on the basis of the market size estimated in the Group M report, Slovenia would be an affected market. Their combined market shares would be [10-20]% including direct sales and [20-30]-[20-30]% excluding direct sales (see Section IV.2.2.1).

507.   The merged entity will face competition from three of the four other large advertising networks392, Havas393, Dentsu-Aegis394 and IPG, as well as from other independent players such as Pristop Media, Doticni395. Some of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

508.   Customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of the Parties' internal data for Slovenia shows that [Summary of Publicis and Omnicom's win, loss and incumbency data].396

509.   A majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Omnicom. Omnicom's closest competitor is also WPP, followed by Publicis.397

510.   In view of the above, Omnicom and Publicis are not each other's closest competitors in Slovenia.

511.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Slovenia398 that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction in particular also for large-scale advertisers.399

512.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Slovenia.

b.      Procurement side

513.   As mentioned in Section IV.2.2.2, Table 21 below presents market shares of the Parties on the procurement side of the MBS market in Slovenia400 based on WARC report:

Table 21: Slovenia - MBS Procurement share by channel

Source: Calculated on the basis of the Parties' real media cost and the market size estimated by type of media in the WARC report (Annex 26 to Form CO)401

514.   As can be seen in Table 21, only online/mobile will be affected by the proposed Transaction.

515.   The Parties' combined market shares on the procurement markets on the online/mobile channel is [10-30]%. Considering the small market shares of the Parties in this affected market, the Transaction is unlikely to raise any serious doubts as to its compatibility with the internal market on this segment.

516.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Slovenia.

IV.4.1.4.17.      Spain

a.      Sales side

517.   In 2012, the Parties’ combined market share in MBS in Spain amounted to [20-30]% including direct sales, and [20-30]-[20-30]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

518.   The merged entity will face competition from the four other large advertising networks402, Havas, WPP, Dentsu-Aegis, and IPG403 as well as from a number of independent players404. Several of these independent players have a number of large- scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

519.   The analysis of RECMA Compitches data indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders.405 In 83% of overlap tenders, there were four or more participants compared to 41% for the other tenders in the country.406

520.   Therefore, the merged entity will continue to face competition from a number of competitors in the tenders in which it will take part post-Transaction.

521.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data shows that on average, Omnicom won tenders in 40% of the cases where it was the incumbent (through the same agency or another agency within the group) and Publicis in 30% of the cases.407

522.   According to the analysis of the Parties' participation in 2010-2012 tenders in Spain in the RECMA Compitches data, Publicis and Omnicom were present in 21% of all tenders. Publicis was present in 55% of the tenders in which Omnicom participated. Conversely, Omnicom was present in 38% of the tenders in which Publicis participated.408 In addition, the most prominent competitor in the tenders in which Omnicom participated was WPP followed by Dentsu-Aegis. Likewise, in tenders in which Publicis participated, Havas and Dentsu-Aegis were first and second respectively409. Omnicom's customers that switched agency, mainly switched to the independent player, Ymedia. Publicis' customers that switched agency, switched to Havas first, then to Dentsu-Aegis. The data does not show any customer switching from Omnicom to Publicis and vice-versa. Omnicom won tenders from WPP and the independent player, Remo, while Publicis won tenders first from WPP, and second from Havas.41

523.   Moreover, a majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Havas. Omnicom's closest competitor is Havas, followed by WPP.411

524.   In view of the above, Publicis and Omnicom are not each other's closest competitors in Spain.

525.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Spain that contradicts the conclusions outlined in the general section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.412

526.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Spain.

b.      Procurement side

527.   Table 22 below presents the market shares of the Parties on the procurement side of the MBS market in Spain by media channel: 413

Table 22: Spain - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time414)

528.    As can be seen in Table 22, the print, online / mobile and radio sectors are not affected by the Transaction. They are therefore not further discussed in this section.

529.   Outdoor and TV. The outdoor advertising sector is concentrated with a few strong players, such as Clear Channel and JCDecaux. According to a 2011 Report on the Spanish Outdoor Advertising segment from Outdoormedia (a Spanish specialised outdoor media agency), the following vendors control together over 55% of the available outdoor advertising space in Spain: JCDecaux (24%), followed by    Cabitel, which is part of Telefónica (16%), and Clear Channel (16%).415 The media suppliers that replied to the market investigation indicated that the outdoor segment is concentrated.416

530.    Likewise, the TV sector in Spain is concentrated. Antena 3 and La Sexta, which merged into Media España and Atresmedia in October 2012, represented 86.8% of total TV advertising sales in 2012.417

531.   The merged entity will therefore face strong media owners on the outdoor and TV channels in Spain. Even though the Parties will represent a relatively important share of the sales of media owners, the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

532.   Cinema. Five main cinema groups are present in the cinema channel, namely Cinesa, Yelmo, Ábaco Cinebox, Cines Acec and Kinepolis, which represented in 2011 41% of the total of screens in Spain and absorbed more than 50% of Spanish cinema audience.418

533.   The merged entity will therefore face strong media owners on the cinema channel in Spain. The countervailing power on the media owners' side will be significant enough to offset any increase in negotiation power of the merged entity.

534.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Spain.

IV.4.1.4.18.      Sweden

a.      Sales side

535.   In 2012, the Parties’ combined market share in MBS in Sweden amounted to [10-20]% including direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

536.   The Commission considers that in the particular case of Sweden, the assessment of the competitive position of the merged entity on the MBS market excluding direct sales is not appropriate, considering the high level of these direct sales in this country (except for TV and outdoor – see paragraph 34 above). The Commission has therefore conducted its assessment on a market including direct sales only.419

537.   The Commission further notes that even if direct sales were to be considered as being outside of the relevant market, their high level in Sweden would likely exercise a further significant competitive constraint on the merged entity post-Transaction in this country.

538.   The merged entity will face competition from the  four  other  large  advertising networks420, WPP, Dentsu-Aegis, IPG and Havas 421  , as well as independent players     like Scream and Bizkit partners422. Some of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-   wide or global) reach.

539.   The analysis of the RECMA Compitches data for Sweden indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders423. In 57% of overlap tenders, there were four or more participants while this ratio was only 31% for the rest of the tenders.424

540.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. Indeed, the analysis of RECMA Compitches data for Sweden shows that in tenders where they were incumbents, Publicis never won any contracts while Omnicom won tenders in 25% of the cases.425

541.   According to the analysis of the Parties' participation in 2010-2012 tenders in Sweden in the RECMA Compitches data, Publicis and Omnicom overlapped in tenders to a limited extent (17% of all tenders). Publicis and Omnicom were present in less than half of the tenders in which the other was present (44% and 39% respectively)426. In addition, the most prominent competitors in the tenders in which Omnicom participated in Sweden was WPP first and Publicis second. In tenders in which Publicis participated in Sweden, WPP and IPG were the most prominent competitors427. In tenders in which Omnicom was the incumbent, it mainly lost to WPP, Dentsu-Aegis and independents, while in tenders in which Publicis was the incumbent, it lost to IPG, WPP, Dentsu-Aegis and Omnicom. When Omnicom was not the incumbent, it mainly won from WPP and Dentsu-Aegis. When Publicis was not the incumbent, it mainly won tenders from independents as well as from Dentsu-Aegis and IPG.428

542.   A majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP followed by Dentsu-Aegis. Likewise, Omnicom's closest competitor is WPP, followed by Dentsu-Aegis.429

543.   In view of the above, Publicis and Omnicom are not each other's closest competitors in Sweden.

544.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Sweden that contradicts the conclusions outlined in the Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.430

545.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Sweden.

b.      Procurement side

546.   Table 23 below presents market shares of the Parties on the procurement side of the MBS market in Sweden:

Table 23: Sweden - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time431)

547.   As can be seen in Table 23, the following media channels are affected by the Transaction in Sweden: TV, outdoor432 and cinema.

548.    TV. In Sweden, TV media ownership is led by a few players: two public broadcast channels SVT1/ SVT1 HD and SVT2/ SVT2 HD, and privately-owned television channels, TV3, TV4 and Kanal 5. The two public broadcast channels have no advertising, consequently the shares of advertising revenue are distributed by the following privately-owned television channels: TV4 (29.8%), TV 3 (22.1%), Kanal 5 (20.4%), TV 6 (17.2%), Kanal 9 (5.3%) and TV 4+ (3%). The three main media owners that concentrate the largest part of the privately-owned television channels are Bonnier, MTG and ProSiebenSat.1 Media. Bonnier, which is the largest media owner in Sweden (across all segments), owns TV 4 which had an audience share of 28.9% in 2012. The other two main players, MTG and ProSiebenSat.1 had an audience share of 19.5% and 8.7% respectively. The three main TV media vendors thus have countervailing power in Sweden.433

549.   Cinema. Svensk Filmindustri (SF Bio) controls around 65%-80% of the segment. The other main players are (i) Folkets Hus och Parker, which has more than 200 movie theatres across the country; (ii) Svenska Bio, which has 104 screens across the country, and (iii) Folkets Bio, which has 17 cinemas across the country. Moreover, SF Bio does not only have a high market share (65-80%) but it also owns a 50% stake in one of the other main players, Svenska Bio. An additional factor increasing the leverage of the cinema owners over media buying agencies is the fact that advertising represents only a small portion of their total revenues (their main source being ticket sales, and sales of food and refreshments).434

550.   Outdoor. Regarding the outdoor channel, the Parties estimate that Clear Channel and JCDecaux together hold more than [90-100]% of the market, Clear Channel representing around 60% of these two players’ combined share and JCDecaux around 40%.435

551.   The merged entity will therefore face strong media owners on the TV, cinema and outdoor channels in Sweden. Even though the Parties will represent a relatively important share of the sales of media owners (in some media channels), the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

552.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Sweden.

IV.4.1.4.19.      The United Kingdom

a.      Sales side

553.   In 2012, the Parties’ combined market share in MBS in the United Kingdom amounted to [20-30]% including direct sales, and to [20-30]-[20-30]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1).

554.   The merged entity will face competition from the four other large advertising networks436, WPP, Dentsu-Aegis437, IPG, and Havas, as well as from a number of independent players438. Several of these independent players have a number of large- scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

555.   The analysis of RECMA Compitches data indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders439. Indeed in 56% of overlap tenders, there were four or more participants compared to 26% for the rest of the tenders in the country.

556.   Therefore, the merged entity will continue to face competition from a number of competitors in the tenders in which it will take part post-Transaction.

557.   Moreover, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies. The analysis of RECMA Compitches data shows that on average, Omnicom won tenders in 58% of the cases where it was the incumbent (through the same agency or another agency within the group) and Publicis in 29% of the cases440

558.   According to the analysis of the Parties' participation in 2010-2012 tenders in the United Kingdom in the RECMA Compitches data, Publicis and Omnicom overlapped in tenders to a limited extent: they were present in 11% of all tenders and in around 25-26% of the tenders in which the other one was present.441 In addition, the most prominent competitor in tenders in which Omnicom participated was WPP followed by Publicis. Likewise, in tenders in which Publicis participated, WPP and Dentsu-Aegis were first and  second  respectively.442   Omnicom's  customers  that  switched  agency,    mainly switched to WPP. A majority of Publicis' customers that switched agency, switched to WPP first, to Dentsu-Aegis second. Omnicom won tenders mainly from WPP, and independent players, while Publicis won tenders both from different large groups and independents without clear majority.443

559.   Moreover, a majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Dentsu-Aegis. Likewise, Omnicom's closest competitor is WPP, followed by Dentsu-Aegis.444

560.   Publicis and Omnicom are therefore not each other's closest competitors in the United Kingdom.

561.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards the United Kingdom that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.445

562.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in the United Kingdom.

b.      Procurement side

563.   Table 24 below presents the market shares of the Parties on the procurement side of the MBS market in the United Kingdom by media channel446:

Table 24 – United Kingdom - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time447)

564.   As can be seen in Table 24, the online / mobile and cinema sectors are not affected by the Transaction. They are therefore not further discussed in this section.

565.   TV. The TV sector in the United Kingdom is concentrated with four players representing 85% of the prime time viewing (BBC: 37.2%, ITV: 27.1%, Channel 4: 11.8%, Sky/News Corp: 8.7%, in 2011).448 However, the BBC does not carry advertising on its public service channels, instead financing its broadcasting through a license fee.449 This makes it less dependent on advertising revenue as well as reducing the amount of advertising time available on prime time TV. The three other groups (ITV, Channel 4 and Sky/NewsCorp) account for approximately three quarters of the attainable audience. Sky is subscription-based which means that it is less dependent on advertising revenues. In addition to those offered by the four most popular broadcasting groups, a number of other channels are available through subscription services such as CNN, MTV and Discovery.

566.   The merged entity will therefore face strong media owners on the TV channel in the United Kingdom. Even though the Parties will represent a relatively important share of the sales of media owners the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

567.   Furthermore, OFCOM, the United Kingdom regulator for telecoms, recently analysed in detail the national advertising market and concluded that "Competition issues in the UK: TV advertising trading mechanism Decision" (para 1.27): …We have evidence which suggests that advertisers have access to detailed pricing information and information about the performance of their media buyers and we believe that advertisers are able to (and do) exercise their ability to switch to alternative media buyers or deal direct with broadcasters. We are also satisfied that features of the market do not prevent the trading model from evolving or create barriers to innovation. …" 450

568.   Other types of media: print, outdoor and radio. A large number of newspapers and magazines in the United Kingdom are owned by eight companies.451 In 2005, two of these (News International, and Daily Mail and General Trust) held a 55% share and the top four accounted for 85% of the total national market. A similar pattern can be observed at regional and local level, where the five largest owners account for over 70% of the market. Two of these five (Daily Mail and General Trust, and Trinity Mirror) are also amongst the top national newspaper owner companies.452  This also holds true for magazines: ownership is concentrated in the hands of a small number of players, such as Hachette Filipacchi, Condé Nast and H Bauer.  453

569.   The most popular radio network is the BBC, which offers ten stations in total and accounts for over half of the radio listenership in the United Kingdom.454 BBC Radio 2 is the most successful of the BBC stations – it is listened to by 27.4% of the population.455 As for TV, the BBC does not carry advertising on its public services for radio. This makes it less dependent on advertising and limits the advertising demand on the radio sector. The largest commercial radio group is Global Radio. It owns numerous national and local radio stations. In May 2013, the Competition Commission ordered Global Radio to divest seven stations following its acquisition of the Real and Smooth radio network. The Competition Commission was concerned that the company’s strength could result in higher prices for advertising across the United Kingdom.456 These elements show the strength of the big players in this segment and the concentration of the British radio environment in general.

570.   As regards outdoor advertising, in a 2011 market study into this sector, the OFT found that "three large media owners (CBS Outdoor, Clear Channel, JCDecaux) represent around 80% of the supply of outdoor advertising space" in the United Kingdom.457

571.   The merged entity will therefore face strong media owners on the print, radio and outdoor channels in the United Kingdom. Even though the Parties will represent a relatively important share of the sales of media owners (in some media channels), the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

572.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in the United Kingdom.

IV.4.1.4.20.      Norway

a.      Sales side

573.   In 2012, the Parties’ combined market share in MBS in Norway amounted to [10-20]% including direct sales, and to [10-20]-[10-20]% excluding direct sales, according to the Parties' estimates of market shares based on the Zenith Optimedia market size (see Section IV.2.2.1)

574.   The merged entity will face competition from the four other large advertising networks458, WPP, Dentsu-Aegis, IPG, Havas, as well as from a number of independent players459. Some of these independent players have a number of large-scale advertisers as clients, that is to say clients for MBS requiring large (EEA-wide or global) reach.

575.   The analysis of RECMA Compitches data indicates that the number of participants was higher in tenders where Publicis and Omnicom were both present compared to the rest of the tenders460. Indeed in 67% of overlap tenders, there were four or more participants compared to 16% for the rest of the tenders in the country.

576.   Therefore, the merged entity will continue to face competition from a number of competitors in the tenders in which it will take part post-Transaction.

577.   No conclusion can be reached on the basis of RECMA Compitches as regards incumbent advantage due to the small size of the sample. Nevertheless, in view of the competitive constraint exerted by the number of players on the Norwegian market, customers have, and will still have post-Transaction, the ability to switch to other MBS agencies.

578.   According to the analysis of RECMA Compitches data in Norway, Publicis and Omnicom were both present in 11% of all tenders, and in less than half of the tenders in which the other was present: Publicis was present in 30% of all tenders in which Omnicom participated, and Omnicom in 38% of Publicis tenders461. In addition, the most prominent competitors in tenders in which Omnicom participated was WPP, followed by Dentsu-Aegis and IPG. Likewise, in tenders in which Publicis participated, WPP was first, followed by Dentsu-Aegis and IPG462.

579.   The data does not show any client switching from Omnicom to Publicis, or vice-versa. Nevertheless, in view of the small number of incumbencies of the Parties in Norway, no conclusion can be drawn in this regard. Omnicom won tenders from WPP, Publicis and the independent agencies, Futatsu and Mood, while Publicis won tenders from WPP.

580.   Moreover, a majority of market participants that replied to the market investigation consider that Publicis' closest competitor is WPP, followed by Omnicom. Omnicom's closest competitor is also WPP, followed by Dentsu-Aegis.463

581.   In view of the above, Publicis and Omnicom are not each other's closest competitors in Norway.

582.   Finally, a majority of customers that replied to the market investigation did not raise any country-specific issue as regards Norway that contradicts the conclusions outlined in the general Section IV.4.1.3 above, notably that there will remain a sufficient number of MBS agencies post-Transaction, including for large-scale advertisers.464 A customer indicates in addition that "[i]n Norway, the contract is based on the advertiser's volume not the agency. Small media agencies may well achieve good prices."465

583.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market on the sales side of the market for MBS and any of its possible segments, in Norway.

b.      Procurement side

584.   Table 25 below presents the market shares of the Parties on the procurement side of the MBS market in Norway by media channel466:

Table 25: Norway - MBS Procurement share by channel

Source: Form CO (ZenithOptimedia including direct sales - Shares of the Parties per cost of media time467)

585.     As can be seen in Table 25, the print, online / mobile, radio and cinema sectors are not affected by the Transaction. They are therefore not further discussed in this section.

586.   TV. The TV sector in Norway is concentrated with three players representing more than 75% of TV viewing (NRK, 40%, Egmont/A-Pressen, more than 25%, ProSiebenSat. 1 Medi AG, 12%).468 The merged entity will therefore face strong media owners on the TV channel in Norway. In addition, the degree of media vendor concentration is strengthened by the fact that the State-owned NRK is publicly funded and operates without any advertising.469

587.   Therefore, even though the Parties will represent a relatively important share of the sales of media owners, the countervailing power on the media owners' side will be significant enough to offset the increase in negotiation power of the merged entity.

588.   Outdoor. The outdoor channel is also concentrated, with two leading players: JCDecaux and Clear Channel. JCDecaux offers transport and billboard advertising as well as having the rights for advertising at most airports. According to the Parties, Clear Channel has the exclusive rights for advertising at Oslo Airport (Gardermoen)470. That company also has rights for transit advertising through agreements with Stor Oslo Lokaltrafikk and other bus operators.471

589.  The countervailing power on the media owners' side will be significant enough to offset any increase in negotiation power of the merged entity.

590.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the procurement market for MBS and any of its possible segments, in Norway.

IV.4.1.4.21.      EEA market

591.   In 2012, the Parties' combined market share in the EEA amounted to [10-20]% including direct sales and [10-20]-[10-20]% excluding direct sales, according to Zenith Optimedia's calculations of MBS shares.472

592.   The competitive assessment conducted on the sales side for the market of MBS at the Europe wide level does not materially differ from the one carried out on a country-by- country basis.

593.   Therefore, all the conclusions reached in the general section preceding the country-by- country analysis remain valid in relation to the assessment of the Transaction at EEA level.

594.   In particular, the Commission considers that also at EEA level: (i) the Parties will be constrained post-Transaction by other MBS agencies (including the four large international networks); (ii) the Parties compete infrequently against each other; (iii) if the merged entity raises its costs, customers will have the ability to switch to other MBS agencies, since switching costs are low; (iv) post-Transaction, competition from  other MBS agencies will be sufficient in order to prevent the merged entity from raising its prices also with reference to large-scale customers with global reach.

595.   In light of the above, the Transaction does not raise serious doubts as to its compatibility with the internal market on the market for MBS and any of its possible segments, in the EEA.473.

IV.4.1.5.   CONCLUSION

596.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market as regards the markets for media buying services in all overlap EEA countries, and at EEA level.

IV.4.2      MARKETING AND COMMUNICATION SERVICES

IV.4.2.1.   VIEWS OF THE PARTIES

597.   According to the Parties, the Transaction will not lead to any horizontally affected national markets for marketing and communication services.474

598.   The Parties contend that the MCS market is highly fragmented and creativity-driven. In addition, MCS contracts are awarded mostly by tenders. Barriers to entry are low, as contracts are easy to terminate without incurring in significant costs. Regular performance reviews impair any incumbent advantage. Therefore, the market for the provision of marketing communication services is characterised by a high level of competition. In addition, the Parties submit that competition concerns can be excluded on the basis of the limited combined market share and the limited increment arising from the Transaction.

IV.4.2.2.   COMMISSION'S ASSESSMENT

IV.4.2.2.1.   Introduction

599.   The Parties' activities in the market for supply of marketing communication services overlap in the following countries: Austria, Belgium, Czech Republic, Croatia, Denmark, Finland, France, Germany, Greece, Hungary, Italy, the Netherlands, Norway, Poland, Portugal, Romania, Slovenia, Spain, Sweden and United Kingdom.

600.   Preliminarily, the Commission note that even on the narrowest geographic markets, that is to say national markets, the combined market shares of the Parties for the supply of marketing communication services do not lead to any affected market.

601.   However, in assessing the competitive effects of the Transaction on the relevant segments of the markets for marketing communication services, the Commission has taken into account the following factors, which are analysed in further detail in the sub- sections below: (i) the competitive constraint exerted on the merged entity by other MCS agencies; (ii) whether the Parties are each other closest competitors; (iii) the level of barriers to entry and expansion; (iv) the degree of competition (also intra-group) pre and post-Transaction.

IV.4.2.2.2.   The competitive constraint from other MCS agencies

602.   In addition to the merged entity, a sufficient number of other large advertising networks, including at least three of the four large competing advertising networks, WPP, IPG, Dentsu-Aegis and Havas, are present (directly or through affiliates) in all the countries where the activities of the Parties overlap and will exercise a significant competitive constraint on the merged entity post-Transaction, including for large-scale advertisers, that is to say. for MCS requiring large (EEA-wide or global) reach. Moreover, national MCS markets are fragmented between a number of competitors of various sizes.

603.   In addition, respondents to the market investigation indicated that the market for the supply of marketing communication services is mostly talent-driven. The majority of the respondents identified creativity as key factor on the basis of which MCS agencies are selected by customers475. Other criteria were also identified as being relevant, such as expertise, quality and delivery, price, scale, (absence of) conflict of interest.

604.   Certain respondents pointed out that creativity plays a lesser role for large contracts (in terms of investment and/or international reach), where value and delivery are more crucial. Nonetheless, this does not prevent smaller agencies from obtaining large contracts and competing with larger agencies.

605.   In addition, competition for winning tenders is not necessarily linked to an agency’s size, although for certain type of multi-national contracts, the ability to ensure scale and network appears important.

606.   Even if competitors' replies were mixed with regard the extent to which MCS contracts are concluded by tender or direct negotiation, the majority of customers indicated that common practise is to award contracts through a bidding process476.

607.   Similarly to MBS, the tender procedure encompasses essentially three phases, the RFI, the RFP, and the chemistry stages. 4.9 competing agencies participated in the selection process at the RFI stage and 3.2 at the RFP stage.477

608.   Considering all the above and other available evidence, the Commission considers that there will remain a sufficient competitive constraint on the merged entity post- Transaction, including for large-scale advertisers.

IV.4.2.2.3.   Closeness of competition

609.   Respondents to the market investigation indicated that the Parties are not generally each other's closest competitors on a country-by-country level. In almost all the countries where the activities of the Parties overlap, respondents to the market investigation identified WPP as the closest competitor of both Publicis and Omnicom.478

610.   Even in the small number of countries where the Parties are each other’s closest competitors, this factor alone is not sufficient to amend the Commission's overall conclusion on the absence of serious doubts, given the presence of a sufficient number of competing MCS agencies post-Transaction, and the low level of barriers to entry that characterise the MCS market, as set out in the previous and following sub-sections of the Decision.479

IV.4.2.2.4.   Barriers to entry or expansion

611.   Respondents to the market indicated that barriers to entry and expansion are low in the MCS market.480

612.   Contracts are concluded for a limited duration, ranging from 1 to 3 years. Shorter contracts can be concluded on a project-by-project basis. On average, contracts can be terminated within a short notice period of 3 months.481

613.   Incumbent agencies appear to enjoy advantages, as indicated also by the significant percentage of contracts won or renewed by the incumbent mentioned by the majority of both customers and competitors.482 However, this evidence is counterbalanced by the low barriers to entry that characterise the market

614.   Moreover, should the merged entity raise its prices post-Transaction, customers are likely to switch easily to another MCS agency and without incurring any significant cost (the only costs mentioned by the majority of the respondents are mainly staff – related). Even if the timeframe to switch depends upon the size, and/or the type of contract, respondents to the market investigation indicated some approximate ranges: from 2 to 6 months for competitors and from 6 to 12 months for customers. In general, setting up new agency appears to be easier than in MBS.483

IV.4.2.2.5.   Competition  post-Transaction

615.   Finally, respondents to the market investigation indicated that competition post- Transaction will still be high484. According to a majority of respondents, the existence of many players will prevent the merged entity from raising prices, including for large customers.

IV.4.2.3.   CONCLUSION

616.   In light of the above, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market as regards the market for marketing and communication services, or any of its segments, in all overlap EEA countries, and at EEA level.

IV.4.3      "BIG" DATA ANALYTICS

617.   The Parties submit that one of the rationales of the Transaction is to develop its activity in so-called "big data" analytics. This relates to the process of examining large amounts of data of a variety of types ("big data") to uncover patterns, correlations and other useful information. The primary goal of big data analytics is to help companies make better business decisions by enabling data scientists and other users to analyse large volumes of transaction data as well as other data sources that may not be assessed by conventional business intelligence  programs.

IV.4.3.1.   VIEWS OF THE PARTIES

618.   The Parties submit that in past decisions, the Commission identified the existence of marketing data services, further segmented into: (i) marketing information services, which consist in the provision of data to individual consumers for direct marketing purposes; (ii) market research services, which are aimed at measuring actual purchasing patterns (essentially through the measurement of sales at different retail points – “retail tracking” – or on the basis of data obtained from consumer panels); and (iii) media measurement services, which are aimed at measuring the audience of specific media, such as television and internet.485

619.   Publicis and Omnicom do not provide market research services as stand-alone services, outside the provision of core MCS and MBS. The Parties also do not provide services consisting in measuring the audience of specific media. Beyond audience measurement, the Commission has also included in past decisions the measurement of advertising expenditures within the scope of media measurement services. Omnicom does not provide any such services. Publicis’ agency, ZenithOptimedia, publishes reports estimating and forecasting advertising expenditures. However, while ZenithOptimedia sells these forecasts to the finance and media community, they are primarily a marketing tool. Publicis derives limited revenues from this activity.

620.   The Parties further submit that while they have marketing data analytics capabilities, they are essentially used in-house as a mean of targeting and optimising MCS campaigns or MBS in support of the provision of core MCS or MBS services. Analysing consumer data is an intrinsic part of the services offered in the markets for MCS and MBS.

621.   Publicis has one subsidiary that provides data analytics services to third parties on a stand-alone basis, Ninah, which is active in the United States and in the United Kingdom. Ninah's revenues are […]. As part of its MBS offering, Omnicom provides its clients with data analytics services through Annalect. In 2012, Annalect’s data analytics worldwide revenues […].

622.   According to the Parties, the data analytics space includes at least the following categories of competitors:

-        Measurement and Analytics providers such as Omniture (Adobe), Google Analytics, Coremetrics (IBM), webtrends, comScore, Nielsen, Flurry, Marketshare, Neuralitic, Agent, Localytics, and Tracksimple (Bluekai).

-        Independent DSPs, such as Turn, Invite Media (Google), EfficientFrontier (Adobe), The Trade Desk, MediaMath, DataXU, and Accordant Media.

-        Other ad agencies, such as WPP, IPG, Havas, Dentsu-Aegis, MDC Partners, and iCrossing.

623.   Accordingly, even if Publicis and Omnicom are seen as active in a market for data analytics, their combined share of such a market and any of its possible segmentations is small, and in any case well below the 15% threshold defining an affected market.

624.    The Parties submit that "big data" has not transformed the core business of MBS and MCS providers. Data and data analytics are merely the raw material and a tool used by advertising agencies to deliver their two core services to advertisers: creating the advertisers’ messages on the MCS market and delivering these messages to consumers on the MBS market. The fact that the data used by advertising agencies has become more robust and extensive has no relevance for market definition in relation to the Transaction because the nature of the services delivered remains essentially the same. In addition, neither party controls the primary sources of this data and the combined entity will still need to use the databases of its clients, and purchase such data from various first-party data sources.

IV.4.3.2.   COMMISSION'S ASSESSMENT

625.   The Commission has assessed whether "big data" may become in the near future a key factor in helping advertisers to target better their offers to customers, in particular as regards online advertising. In that context, the Commission has investigated whether the availability of “big data” is or may become a key factor for the merged entity in order to conduct its business and attract new advertisers. Finally, the Commission has sought to identify who are the current main providers of “big data”, and assess whether competing advertising agencies will still have access to “big data” from other providers even if post-merger, the merged entity was to develop its own "big data" analytics platform and not give access to it to its competitors.

626.   Replies from competitors were mixed as regards the extent to which big data is important or not for MBS agencies to conduct their business and attract new customers. A number of competitors explained that in the “offline” world – that is to say in television, print, outdoor, etc. - although it may become increasingly relevant in the future, big data is currently still of little relevance, and generates limited revenues. A number of competitors, however, confirmed that for digital advertising and social media, where data is more easily available and collected, big data is becoming important. The replies given by competitors on the importance of big data in MCS were materially similar to the ones given on MBS. 486 Responses by customers that replied to the market investigation were also mixed as regards the importance of big data.487 Finally, a majority of media owners replied that the concept of big data was still in its infancy stage and that it is not yet or not at all relevant for them. The media owners that replied that big data is important to them are mostly present in online advertising488.

627.   In addition, competitors confirmed that they have access to a large number of third party suppliers that provide big data analytics. Other competitors have also developed their own data analytics tools, with the help of consultancy firms. Finally, some of the larger competitors of the merged entity have also developed their own big data subsidiaries.489

628.   A large majority of competitors also confirmed that they are at least similarly placed to the merged entity to get access to big data analytics. As one competitor, Salesforce.com, Inc., explained, “increasingly data are controlled by third party content providers and aggregate on the web such as Twitter, Facebook, Google and blogs. These sources are open to the public for free or for a fee. We do not anticipate any advantages that the merged entity may have over its competitors in terms of access to such data in result of the merger”.490 A large majority of customers also confirmed that competitors of the merged entity are at least similarly placed to the merged entity to get access to big data analytics.491 Finally, none of the media owners believed that the merged entity is better placed than its competitors post-merger to get access to big data analytics.492

629.   Furthermore, a majority of competitors responded that if the merged entity develops post-Transaction its own big data analytics platform and does not allow access to it, the impact will be limited as they are currently using their own data analytics platform or one from third parties. As one competitor, Dentsu-Aegis, explained, “historically, there has been little success where advertising holding companies have sought to develop such software as technology companies typically have better access to investment, research and development and technical resources. Therefore, (…) it is unlikely that the merged entity could create a platform that could negatively impact” it.493 Replies from customers to the same question were broadly consistent with those of competitors. While a number of customers explained that they are interested in using the database of the merged entity if it were to become the "most comprehensive one available on the market", other customers explained that many other providers are developing big data platforms, or will be able to access similar data and not be disadvantaged. One customer, [Identity of customer], also explained that big data is even less relevant for MCS as creative insight is less reliant on data.494 Finally, even though a small minority of media owners considered that they will be negatively impacted if the merged entity develops its own big data analytics and not share it with its competitors495, a large majority of them explained that they will not be impacted if the merged entity engages in such practice.496

630.   Based on the above, the Commission concludes that, post-Transaction, there will remain a sufficient number of alternative providers of big data analytics to the merged entity and that as a result, no serious doubts are arising from the Transaction in relation to big data.

IV.5.               ANALYSIS OF HORIZONTAL COORDINATED EFFECTS OF THE TRANSACTION

631.   A number of respondents to the market investigation suggested that the Transaction will lead to a duopoly, in particular in certain national MBS markets, between the merged entity and WPP, notably for advertisers with a global reach. This, in turn, will enable both groups together to impose their conditions on media vendors and as a result, attract more advertisers and put smaller MBS competitors at a significant competitive disadvantage in the longer run. Some additional market participants expanded the theory to a set of three or four large groups, including the merged entity, WPP, and other groups, which together will hold a significant share of the MBS markets in some countries.

IV.5.1      VIEW OF THE PARTIE

632.   The Parties submit that the Transaction will not lead to coordinated effects

633.   According to the Parties, the general market conditions in MBS prevent any consensus regarding “terms of coordination” from emerging. That is because of the sheer number and diverse nature of agencies competing for business in any given national market, of the importance of non-price factors such as consumer insights and creative media planning strategies, which are impossible to coordinate, of the disruption resulting from the emergence of advertising opportunities in online and interactive media, and of the concomitant entry of new competitors.

634.   In addition, the Parties submit that monitoring deviations from any hypothetical terms of coordination will be unfeasible given the bilateral and confidential nature of negotiations between agencies and advertisers. Effective retaliation against deviators will be excluded in a segment characterised by large-volume bids, in which the gains of deviating from any hypothetical terms of coordination are large and immediate, whereas the threat of “punishment” is small, uncertain and distant.

635.   Finally, any prospect of successful coordination is further reduced by the large number of “outsiders” to any hypothetical oligopoly, in particular independent agencies and non- agency competitors for new forms of media, as well as the countervailing bargaining power of advertisers. All the above-described features preventing coordinated effects are present in each EEA country.

IV.5.2      COMMISSION'S ASSESSMENT

636.   To assess coordinated effects, settled case law497 and the Commission’s guidelines on the assessment of horizontal mergers498 require evidence that a merger will make coordination more likely, more effective and more sustainable. The analysis needs to focus in particular on: (i) the ability to reach terms of coordination; (ii) the ability to monitor deviations; (iii) the existence of a credible deterrent mechanism if deviation is detected; and (iv) the reaction of outsiders such as potential competitors and customers.

637.   On the procurement side of the MBS market, where MBS agencies meet media vendors, the Commission has investigated whether, and to what extent, the level of discounts that are agreed between an MBS agency and a media vendor are generally known in the market, so that competing MBS agencies are able to detect any deviation from the amount of discounts tacitly agreed between the colluding MBS agencies.

638.   As regards the level of discounts offered to winning MBS agencies, evidence gathered during the market investigation indicated that coordination on the level of discounts is unlikely given that the rates are negotiated individually between the media vendors and the media buying agencies (or sometimes directly with the advertisers) on specific contracts with each customer and are not normally disclosed publicly.499

639.   Furthermore, the Commission notes that there are many types of rebates even if certain common features exist. In general, media vendors offer volume-based rate card discounts, free advertising space/time and other types of specific discounts such as on- top cash discounts or agency related discounts. A number of customers also qualified the system of rebates granted by media vendors to MBS agencies as being "opaque".500

640.   The Commission further notes that the level of price and the ability to achieve discounts, although an important factor for advertisers when selecting an MBS agency, is only one of a range of key parameters on which MBS agencies compete. These other parameters include non-price parameters such as the ability to achieve the right media strategy.

641.     Reaching terms of coordination will therefore be difficult.

642.   The Commission also investigated whether it will be likely that the largest MBS groups, excluding small and medium-sized MBS competitors, and possibly other large competitors – will coordinate their behaviour on the procurement side of the MBS market.

643.   Only a small number of respondents to the market investigation raised the concern that the largest MBS agencies will be able to coordinate their behaviour on the market for the procurement of MBS in any given country. One respondent that substantiated its concern, raised the possible issue that the merged entity and other larger agencies, because of the increased gap between them and smaller MBS agencies, will eventually not pass-on the fruit of their increased power to the advertisers in the form of better prices.

644.   However, evidence collected in the market investigation indicated [Evidence about pass- on]501. In addition, a large majority of customers that responded to the market investigation stated that they are aware of the levels of rebates obtained by their MBS suppliers, and usually contractually require that rebates relating to their activity are passed on to them, and have put in place either internally or externally through media auditors a system of checks to ensure that rebates obtained by MBS agencies are actually passed on to them.502

645.   Therefore, even if the merged entity and the other remaining large competitors could hypothetically collude tacitly and as a result extract better prices and rebates than their smaller competitors from media vendors, they would have to pass these rebates to advertisers. This would affect negatively the incentive to coordinate between the set of large groups with a view to keep the rebates to themselves.

646.   In addition, given the presence of other competing MBS agencies (which will not participate in the tacit coordination), advertisers could switch to these other MBS agencies, to the detriment of the colluding entities. The likely reactions of these outsiders will also make tacit collusion unlikely. Advertisers will further have the necessary countervailing negotiation power to defeat any strategy to increase their prices against them. Media owners are also often concentrated, giving them similar countervailing  negotiation power.

IV.5.3      CONCLUSION

647.   In light of the above, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market as regards the MBS market, arising from a risk of coordinated effects.

IV.6.     ANALYSIS OF VERTICAL EFFECTS OF THE TRANSACTION

648.   The Transaction will create vertical affected markets in relation to: (i) Publicis' activities in the market for the sale of advertising space in cinema in France and in the Netherlands and; (ii) the Parties' activities in the cinema MBS segment.

IV.6.1      VIEW OF THE PARTIES

649.   The Parties submit that the vertical effects brought about by the Transaction between the market for the sale of advertising space in cinema and the procurement side of the MBS market will not have any adverse impact on competition.

650.   Publicis is active in the market for the sale of advertising space in cinemas in France and in the Netherlands through its subsidiary Mediavision NV503. Omnicom is not active in the sale of advertising space (irrespective of the media segmentation), so there is no overlap. The Parties submit that should the cinema channel segment be considered as a market, the Transaction will not raise any vertical foreclosure concerns, as the merged entity will not have the ability or the incentive to foreclosure access to inputs or to downstream markets

651.   Furthermore, the Parties argue that in any event, the Transaction will not modify Publicis' current incentives as: (i) it will not bring about any change on the sale of advertising space in cinema, as Omnicom does not sell advertising space (in any media) (ii) it will not substantially change the market position currently held by Publicis on the media buying side, given Omnicom’s limited market share for both France ([0-10]%) and the Netherlands ([0-10]%).

IV.6.2      COMMISSION'S ASSESSMENT

652.   Vertical mergers pose no threat to effective competition unless the merged entity has a significant degree of market power in at least one market concerned504. Market shares and concentration levels are therefore an important proxy in the Commission's assessment of vertical mergers.

653.   The Commission is unlikely to find concern in non-horizontal mergers where the market share post-Transaction of the merged entity in each of the markets concerned is below 30%505.

654.   In the present case, the Parties' combined market share in the market for the sale of advertising space in France and in the Netherlands and in the vertically related MBS market is low. In particular, in France this share is [0-10]% for the global market of sale of advertising space or time and [20-30]% for MBS, whereas in the Netherlands this share is below [0-10]% for the global market for the sale of advertising space and [0- 10]% for MBS506. In MBS, if the market is segmented by media type, the Parties' market shares are as follows: [10-20]% in France507 and [0-10]% in the Netherlands508.

655.   In the market for the sale of advertising space in cinemas, Publicis’ share of advertising space sold in cinema amounted to [30-40]%509 in France and approximately [70-80]% in the Netherlands.

656.   Therefore, the merged entity will have a market share below the safe harbour set out in the Guidelines in the global market for the sale of advertising space or time and in the MBS market regardless of any possible segmentation my media channel, but above this threshold in the market for the sale of advertising space or time in cinema, in both France and the Netherlands.

657.   In a merger between companies which operate at different levels of the supply chain, anti-competitive effects may arise when the merged entity's behaviour can limit or eliminate rivals' access to supplies (input foreclosure) and/or to a sufficient customer    base (customer foreclosure). 510

-        Input foreclosure

658.   In relation to the Transaction, the Commission has assessed whether, in the downstream markets for the media buying services, the Parties' competitors will be foreclosed, post- Transaction, from having sufficient access to the key inputs, such as the advertising space.

659.   In investigating the likelihood of input foreclosure, the Commission has assessed: (i) whether, post-Transaction, the merged entity will have the ability to substantially foreclose access to input; (ii) the merged entity will have the incentive to do so; and (iii) whether a foreclosure strategy could have a significant detrimental impact on effective competition downstream.

660.   In the case at stake, it seems unlikely that the merged entity will have the ability and the incentive to restrict the access to the advertising space to competitors of the merged entity’s media buying agencies or to otherwise discriminate against these competitors because the merged entity will not be able to negatively affect the overall availability of advertising space in cinema for the competing media buying agencies.

661.   As argued by the Parties, the merged entity will sell advertising space on behalf of cinema owners, which can switch to other suppliers if the merged entity does not act in their interests. Furthermore, there are sufficient alternatives for the supply of advertising space in cinema so that competing media buying agencies can easily switch their demand to alternative suppliers in France and in the Netherlands.

662.   Furthermore, competing media buying agencies are likely to put in place counter- strategies such as reducing their investments in the cinema channel and redeploying them in other media channels, as cinema does not generally represent a key segment for a successful advertising campaign.

663.   No respondents to the market investigation raised any issue about the effects of this vertical relationship.

664.   For all these reasons above, any input foreclosure will also not be profitable.

665.   Finally, and in any case, any vertical foreclose will not raise any serious doubts because the sale of advertising space in cinema is not an important product for the downstream market, as advertising in cinema accounts for a tiny share of total media purchase (0.1% in the Netherlands and 1.1% in France) due to the fact that the audience reached by cinema is limited and occasional.

666.   In light of the above, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market as regards a possible input  foreclosure strategy of the Parties in the market for the sale of advertising space in cinema in the France and in the Netherlands.

-        Customer foreclosure

667.   In relation to the Transaction, the Commission has assessed whether, in the upstream markets for the sale of advertising space in cinema, the Parties' competitors will be foreclosed, post-Transaction, from having sufficient access to downstream customers such as the merged groups' media buying agencies.

668.   In assessing the likelihood of anti-competitive customer foreclosure, the Commission has assessed: (i) whether the merged entity will have the ability to foreclose access to downstream markets by reducing its purchases from its upstream rivals; (ii) whether the merged entity will have the incentive to do so; and (iii) whether such a foreclosure strategy will have a significant detrimental effect on customers in the downstream market.511

669.   In the case at stake, it seems unlikely that the merged entity will have the ability and the incentive to restrict the access to merged group's media buying agencies to competitors on the upstream market for the sale of advertising space in cinema given the large customer base in the downstream market of media buying agencies and the merged entity's limited market power in that market in France and in the Netherlands.512

670.   Similarly to input foreclosure, no respondents to the market investigation raised any issue about the effects of this vertical relationship.

671.   Based on the above, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market as regards a possible customer foreclosure of the Parties in the market for sale of advertising space in cinema in the France and in the Netherlands.

IV.6.3      CONCLUSION

672.   In light of the above, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market as regards the vertical relationship between the Parties on the market for the sale of advertising space in cinema and the MBS markets in France and in the Netherlands.

IV.7.         ANALYSIS OF CONGLOMERATE EFFECTS OF THE TRANSACTION

IV.7.1       VIEW OF THE PARTIES

673.   The Parties submit that the Transaction is unlikely to lead to any conglomerate effects between MCS and MBS because the Parties will not have the ability and incentive   to leverage a position on one market to another by means of tying or bundling or other exclusionary practices (conglomerate  foreclosure).

674.   The Parties argue that there is no vertical or conglomerate relationship between MCS and MBS, as in most of cases they are tendered and contracted separately and are perceived as separate services by clients513.

675.   In any case, the Parties submit that they already offer, as most of their competitors, a wide range of services, and the Transaction does not increase that range or provide any unique advantage that could give rise to conglomerate effects.

IV.7.2       COMMISSION'S ASSESSMENT

676.   Conglomerate effects concern companies that are active in closely related markets. MCS on the one side and MBS on the other side may be considered as neighbouring markets. These services may be provided to the same customers and both belong to the overall sector of AMC services.

677.   In the present case, respondents to the market investigation did not raise any concerns with regard to conglomerate effects.

678.   The majority of the respondents to the market investigation indicated that the Transaction will not have any significant impact on the way in which MBS and MCS are supplied514. The Parties already owns multiple agencies that are capable of offering services in both MBS and MCS and do not need the Transaction to present themselves as a full-service operator. The majority of customers mentioned that they buy MBS and MCS separately from different agencies, which most of the time do not belong to the same group515. Large competitors indicated that even if they are able to offer both services, customers decide to separate them516.

679.   Although one competitor that responded to the market investigation alleged that the merged entity will be able to leverage its market power in price negotiations with media vendors which in turn will lead to increased barriers to entry, the Commission considers that any such effect would be counterbalanced by the broad choice of alternative agencies and by the buyer power of media vendors (see in particular Section IV.4 of this Decision).

IV.7.3       CONCLUSION

680.   In light of the above, the Commission considers that the Transaction does not raise any doubts as to its compatibility with the Internal Market with regard to conglomerate effects.

V.         CONCLUSION

681.   For the above reasons, the Commission considers that the proposed concentration does not raise serious doubts as to its compatibility with the internal market.

682.   It has therefore decided not to oppose the Transaction and to declare it compatible with the internal market and with the EEA Agreement. This decision is adopted in application of Article 6(1)(b) of the Merger Regulation.

1         OJ L 24, 29.1.2004, p. 1 ("the Merger Regulation"). With effect from 1 December 2009, the Treaty on the Functioning of the European Union ("TFEU") has introduced certain changes, such as the replacement of "Community" by "Union" and "common market" by "internal market". The terminology of the TFEU will be used throughout this decision.

2         Publicis offers advertising services mainly through five global networks: Publicis Worldwide, Leo Burnett, Saatchi & Saatchi, Fallon and Bartle Bogle Hegarty ("BBH"). Publicis’ media strategy, planning and buying services are structured around two independent entities: ZenithOptimedia and Starcom MediaVest Group. In certain EEA countries, another Publicis entity called Vivaki serves as an umbrella organisation bringing together ZenithOptimedia and Starcom MediaVest Group in their purchasing negotiations with media vendors.

3         Omnicom offers advertising services mainly through three global networks: BBDO, DDB and TBWA. It offers media planning and purchasing services under the Omnicom Media Group ("OMG"),  which consists of two full service media networks: Optimum Media Direction ("OMD") and Pattison Horswell Durden ("PHD").

4         Turnover calculated in accordance with Article 5(1) of the Merger Regulation and the Commission Consolidated Jurisdictional Notice (OJ C 95, 16.04.2008, p. 1).

5         However, Publicis does not own or control the cinema ad space. Rather, it acts as a media selling agent for cinema owners in these countries.

6         Form CO, paragraph 204.

7         Form CO, paragraph 205.

8         Form CO, paragraph 206.

9         Form CO, paragraph 136.

10      See Commission Decision of 10 June 1999 in case IV/M.1529 – Havas Advertising/Media Planning, recitals 11 and 13; Commission Decision of 18 June 2002 in case COMP/M.2785 – Publicis/BCOM3, recital 9; and Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recital 21.

11      See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 4; replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 3; replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 3.

12      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 4.

13      See Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recital 19.

14      See Commission Decision of 18 June 2002 in case COMP/M.2785 – Publicis/BCOM3; and Commission Decision of 24 January 2005 in case COMP/M.3579 –WPP/Grey.

15      See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, questions 8.2 and 8.3.

16      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 7.2 and 7.3.

17      See Commission Decision of 10 June 1999 in case IV/M.1529 – Havas Advertising/Media Planning, recital 12; and Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recital 33.

18      See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 9.

19      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 8 and 9; replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 10.

20      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 11.

21      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 14; replies   to Commission questionnaire Q1 to Competitors of 25 November 2013, question 15.

22      See Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recital 23.

23      See Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recitals 25-32.

24      See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 16.

25      See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 17.

26      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 15.

27      Such segmentation was considered in the Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recitals 62 et seq.

28      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 35.5.

29      See replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 20.1.

30      See Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recital 40.

31      See Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recital 41.

32      See replies to Commission questionnaire Q3 to Media owners of 25 November 2013, questions 7 and 8.

33      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 13.

34      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 12; replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 13.

35      See Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recitals 42-48.

36      See footnote 27.

37      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 35.5.

38      See replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 20.1.

39      Submission of the Parties of 17 December 2013. Responses to the market investigation did not suggest the existence of any significant variations among media channels in other EEA countries where the Parties' activities overlap.

40      See Commission Decision of 10 June 2005 in case IV/M.1529 – Havas Advertising/Media Planning, recitals 14-17; Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recitals 34- 35.

41      See Commission Decision of 10 June 1999 in case IV/M.1529 – Havas Advertising/Media Planning, recital 14; Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recital 35.

42      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 17 and 19.

43      See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 22.

44      See Commission Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recital 49.

45      See Commission Decision of 11 March 2008 in case COMP/M.4731 – Google/DoubleClick, recitals 83- 84.

46      See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 23.1; replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 20.1; replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 11.1.

47      See Commission Decision of 24 January 2005 in case COMP/M. 3579 – WPP/Grey, recital 13.

48      See Commission Decision of 24 January 2005 in case COMP/M.3579 WPP/Grey, recital 13 and Commission Decision of 24 August 2008 in case COMP/M.2000 WPP Group/ Young & Rubicam, recitals 10 et seq.

49      Form CO, paragraph 108 et seq.

50      Form CO, paragraph 114.

51      See replies to Commission questionnaires Q1 to Competitors of 25 November 2013, question 52 and Q2 to Customers of 25 November 2013, question 41

52      The Parties' 2012 combined market share for the supply of public relations services was below [5-10]% in each EEA country. As a result, even if such a separate segment were considered, the Transaction would not give raise to any serious doubts. See Form CO, paragraph 115, footnote 31 and paragraph 783, submission of the Parties "PR analysis" of 16 December 2013.

53      See Commission Decision of 24 January 2005 in case COMP/M. 3579 WPP/Grey, recital 15.

54      See replies to Commission questionnaires Q1 to Competitors of 25 November 2013, question 55 and Q2 to Customers of 25 November 2013, question 44.

55      See Commission Decision of 24 January 2005 in case COMP/M. 3579 WPP/Grey, recital 11.

56      The Parties have not been able to provide an estimation of their combined share of supply of healthcare MCS as no existing source provides the total value of such a potential market. Due, however, to the large number of players providing healthcare MCS, the Parties believe that their share in such discipline would be within the same order of magnitude as in the overall MCS market or in the PR segment.

57    See Commission Decision of 25 July 2003 in case COMP/M.3209 WPP/Cordiant, recital 22 and Commission Decision of 24 August 2000 in case COMP/M.2000 WPP Group/Young & Rubicam, recitals 15, 16 and 17.

58      See replies to Commission questionnaires Q1 to Competitors of 25 November 2013, questions 56 and 57 and Q2 to Customers of 25 November 2013, question 46.

59      See Commission Decision of 24 January 2005 in case COMP/M. 3579 WPP/Grey, recitals 16-18

60      See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, questions 58, 59 and 60 and Q2 to Customers of 25 November 2013, questions 48 and 49.

61      See Commission Decision of 14 June 2013 in case COMP/M.6866 – Time Warner/CME, recitals 58-68, Commission Decision of 8 September 2009 in case COMP/M.5533 – Bertelsmann/KKR/JV, recitals 46- 47; Commission Decision of 14 September 2001 in case COMP/M. 2529 – JCD/RCS/Publitransport/IPG, recitals 7-9.

62  See Commission Decision of 24 January 2005 in case COMP/M.3579 WPP/Grey, recital 49.

63      Annex 22 to the Form CO.

64      Annex 26 to the Form CO.

65      Annex R24 to Parties' response to Commission's RFI of 26 November 2013.

66      With respect to six EEA countries (Belgium, France, Greece, Lithuania, the Netherlands and Portugal), the report indicates that expenditures estimates include agency commissions and fees. In addition, for Slovenia, the data includes production costs.

67      The limited variations mentioned in footnote 66 above do not materially alter the share estimates.

68      See Commission Decision of 24 January 2005 in case COMP/M.3579 WPP/Grey, recital 59.

69      See Commission Decision of 24 January 2005 in case COMP/M.3579 WPP/Grey, recital 59.

70      These countries are Belgium, Croatia, Germany, Greece, Ireland, Italy, the Netherlands, Slovenia, the United Kingdom and Norway. These estimations are available only in a limited number of EEA countries and are based on surveys from only a sample of media owners which may affect their accuracy.

71      Annex 23 to the Form CO.

72      Annex R24 to Parties' response to Commission's RFI of 26 November 2013.

73      The data collected by RECMA from advertisers and media agencies estimates ‘activity’ which is intended to represent gross billings.

74      For example, RECMA estimates Omnicom’s ‘activity’ in Europe to be approximately EUR 13 billion in 2012 whereas Omnicom’s actual gross billings were […].

75      Annex 36 to the Form CO.

76      These countries are Austria, Belgium, Greece, Hungary, Latvia, Spain, Sweden, Norway and the United Kingdom. In the following EEA countries there is either no difference at all or a difference of less than 5 percentage points: Austria, Belgium, Greece, Hungary, Latvia, Spain and Sweden. For Norway the difference is less than 10 percentage points; however even if the difference is higher compared to the other countries, the combined market shares of the Parties' would increase from [10-20]% to [20-30]% only (or from [10-20]% to [20-30]% excluding direct sales) and the Commission's conclusions would not change. See also the country-specific sections below. The United Kingdom is the only EEA country for which other sources than the three reports estimate a considerably smaller market size. These other sources are the Kingston Smith report (http://www.kingstonsmithw1.co.uk/annual-survey-2012/), the IBIS (Advertising Agencies in the UK – Market Research Report, May 2013), Campaign Top 50 Media Agencies (http://www.campaignlive.co.uk/wide/1175163/Top-50-media-agencies-2013/) and Nielsen (Nielsen Online report). However, these local sources either have only a limited coverage (Kingston Smith report and Campaign Top 50 Media Agencies both of which estimate the total cost of media purchases of the top 50 media buyers), or they underestimate a segment (for example Nielsen underestimates the online segment). In addition, the IBIS does not aim at estimating advertising expenditures but at providing a general overview of the advertising agencies industry, with the result that it under-estimates total cost of media.

77      In addition to these EEA countries, Publicis is active in Bulgaria and Omnicom in Estonia, Finland and Slovakia. However the Parties' activities do not overlap in these countries. None of the Parties is active in Cyprus, Malta, Iceland, Lichtenstein and Luxembourg.

78      The WARC report does not cover the Czech Republic.

79      Publicis has acquired a 75% stake in Walker Media, a media buying company in the United Kingdom (and Cyprus, where Omnicom is not active). Walker Media's revenues have been included to the Parties' combined revenues in the United Kingdom.

80      The level of direct buying in each EEA country is based on the Parties' estimates.

81      As regards market shares of competitors if direct sales are excluded, the effect of excluding direct sales from the total market size will in principle be proportional for the Parties and their competitors because the market size will decrease and therefore the market shares will increase for the Parties and their competitors to a proportional extent so that there will be no significant difference for the purpose of the competitive assessment.

82      The Commission notes that the Office of Fair Trading of the United Kingdom also relied on the ZenithOptimedia reports in its recent assessment of the acquisition of Aegis by Dentsu in 2012. See OFT's Decision in Case ME/5636/12 of 3 October 2012, recital 39. In addition, the Commission used the WARC reports in its Decision of 24 January 2005 in case COMP/M.3579 – WPP/Grey, recital 59.

83      Taking into consideration that Group M report does not estimate the markets for radio and cinema in Slovenia for 2012, the Commission will rely on the alternative source WARC to estimate the market shares of the Parties in TV, print, radio, outdoor, online/mobile and cinema on the procurement side of the MBS market in Slovenia.

84      Form CO, paragraph 347.

85      Submission of the Parties of 17 December 2013.

86      The local sources include AGCOM figures for Italy and Monitor Polsky B for Poland. See Form CO, page 229.

87      These EEA countries are: Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, the Netherlands, Norway, Poland, Portugal, Spain, Sweden, and the United Kingdom.

88      See Commission Decision of 17 September 2012, OJ L 251, 18.09.2012, p. 49 and Article 2(1) and 12 (b) of Regulation (EC) No 223/2009, OJ L 87, 31.3.2009,p. 169

89      As a further example to show that the adjustment mechanism will not lead to any affected market, the Parties mentioned that MCS demand estimate needs to be reduced by more than 90% for the Parties’ combined market share to reach 15% in some EEA countries, which is not a realistic hypothesis.

90      The EEA totals are based on parties' actual revenues and Eurostat market size adjusted for MBS revenues, as it was not possible to estimate EEA market shares on the basis of the Barnes report, as it includes only 18 EEA countries.

91      GRPs are a measure of the volume of the audience watching the client’s commercial advertisement. One GRP represents 1% of the available audience. Specifically, GRPs quantify impressions as a percentage of the population reached rather than in absolute numbers reached. GRP values are commonly used by media buyers to compare the advertising strength of various media vehicles.

92      The Cost Per Rating ("CPR") metric represents the cost of buying one rating point, or 1% of the targeted population (which may depend on sex, age group of socio-professional category). It is calculated by dividing the cost of the campaign by the GRPs and enables media buyers to compare the costs of different outlets.

93      Cost per mille, often abbreviated to CPM, means that advertisers pay for every thousand displays of their message to potential customers (mille is the Latin word for thousand). In the online context, ad displays are usually called "impressions."

94      A rate card is a document containing prices and descriptions for the various ad placement options available from a media vendor.

95      See Form CO, paragraph 258

96      See Annex 32 of Form CO, Compass Lexecon report on RECMA Compitches Win/Loss Analysis.

97      See Commission Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, OJ C 31, 05.02.2004, page 7, point 17.

98      See paragraph 195 above.

99      See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 15.

100    See Commission Decision of 24 January 2005 in case COMP/M.3579 WPP/Grey, recital 71.

101    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 40.1 and 65.

102    See replies to Commission questionnaire Q3 to Media vendors of 25 November 2013, question 25.

103    Compitches 2012, page 8.

104    Market shares at national level are presented in Table 1 and Table 2b of the Decision.

105    See CompassLexecon analysis of RECMA Compitches data for 2010, 2011, and 2012. The data analysed included Austria, Bulgaria, Belgium, Croatia, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Italy, Ireland, the Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Spain, Sweden, and the United Kingdom. It did not include Latvia, Lithuania, and Slovenia, as no relevant data is available for these countries. Furthermore, Bulgaria, Finland and Slovakia are not assessed in the present Decision as the Parties activities do not overlap in these three EEA countries. The CompassLexecon report contains information about 3,029 tenders. Out of these tenders, the contenders are identified in 972 tenders, and the contenders, winner and incumbent MBS agency, are identified in 856 tenders; the rest of tenders have incomplete information. The Commission performed its competitive assessment using the set of 856 tenders with complete information, in order to portray a picture that is as close as possible to reality. However, it appears that the results identified using either the set of tenders with complete information, or the entire set of tenders, are not different to an extent that it would alter the conclusions of the Commission's assessment. This assessment is reported in more detail in the general and in the country- specific competitive assessment section of the present Decision.

106    See reply of CompassLexecon, on behalf of the Parties, to the questions of the Commission of 26 November 2013, question 15.

107    In addition, RECMA data includes a value threshold for each country (for instance EUR 2 million in the United Kingdom and Germany, EUR 1 million in France, etc. The lowest thresholds apply to Greece (EUR 0.2 million), and Croatia and Romania (EUR 0.1 million). The Parties and their competitors participate in tenders which fall below these thresholds. However, the Commission does not expect the competitive situation to be significantly different for lower value tenders; if different, it would rather be expected to have more competition from independent, smaller agencies.

108    Market shares by channel at national level are presented in Table 4 of the Decision.

109    Compitches 2012, page 30. The statement refers to the "top 212" countries, including countries both in the EEA (the United Kingdom, Germany, France, Italy, Spain) and outside (USA, Canada, Mexico, Russia, China, India, Australia).

110    See replies to Commission questionnaires Q1 to Competitors, of 25 November 2013, question 63 and Q2 to Customers of 25 November 2013, question 24.

111    Compitches 2012, page 4. In the advertising industry, a pitch is when an advertising agency is trying to win the business of an advertiser, either through direct negotiations or through a competitive tender process involving other advertising agencies.

112    See replies to Commission questionnaires Q1 to competitors of 25 November 2013, question 65 and Q.2 to customers of 25 November 2013, questions 25.1-25.2.

113    The Parties confirmed that RECMA only includes contenders that were invited to pitch whereas more agencies may be initially approached when advertisers put out a RFI. The use of RECMA data would therefore be cautious in this context.

114    See also Section IV.4.1.4.

115    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 40.

116    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 32.

117    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 46.

118    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 39.

119    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 39.1.

120    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 41.

121    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 33.

122    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 23.

123    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 42.

124    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 42.1.

125    That is to say the same agency was the incumbent and winner.

126    See submission of the Parties of 18 December 2013, tab "EU Wide_Comp", table 15.

127    See submission of the Parties of 18 December 2013, tab "EU Wide_Comp", table 15.

128    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 45.1.

129    See replies to Commission questionnaire Q3 to Media Owners of 25 November 2013, questions 20.1 and 20.2.

130    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 35.1.

131    See submission of the Parties of 18 December 2013, tab "Countries_Comp", table 26.

132    See submission of the Parties of 18 December 2013, tab "Countries_Comp", table 31.

133    See replies to Questionnaire Q1 to Competitors of 25 November 2013, question 39, Questionnaire Q2 to Customers of 25 November 2013, questions 31.1 and 31.2, and Questionnaire Q3 to Media owners of 25 November 2013, question 19.

134    See Commission Decision of 24 January 2005 in case COMP/M. 3579 WPP/Grey, recital 47.

135    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, questions 41.

136    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, questions 47 and 50.

137    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 47.

138    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 30.

139    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 50.

140    See submission of 5 December 2013, and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 31.

141    Annexes 46 and 47 to the Form CO.

142    See replies to Commission Questionnaire Q2 to Customers of 25 November 2013, question 29.

143    Which are present directly or through affiliates.

144    Such as Mundomedia, 6+1, Action Marketing, Ad-sys, Fantastic, Impact Diffusion, Impuls, Mediawize, PTOC, Robert & Marien, Wondergarden and Zigt.

145    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

146    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

147    The ratio takes into account wins through the same incumbent agency or another agency within the group. Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

148    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

149    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30.

150    No conclusions can be drawn from the competitors against which Omnicom and Publicis won tenders in Belgium as the complete data set was too limited. Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

151    See replies to Commission questionnaire Q1 to competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to media owners of 25 November 2013, question 19.

152    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

153    Even considering the estimated market shares based on the Group M report, the Commission's conclusions do not change.

154    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

155    See Section IV.2 on methodologies for estimating market shares for MBS. Rounding effects may intervene in all tables.

156    Which are present directly or through affiliates.

157    Such as Czech Promotion and VCCP.

158    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

159    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

160    The ratio takes into account wins through the same incumbent agency or another agency within the group. Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

161    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

162    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30.

163    No conclusions can be drawn with regard to Publicis' losses as the complete dataset only includes one tender in which Publicis was the incumbent.

164    Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

165    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

166    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

167    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

168    See Section IV.2. on methodologies for estimating market shares for MBS.

169    See replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 23.

170    The Parties explained that it also owns Outdoor Akzent as of recently.

171    As the level of direct sales is very low (1%) for TV, including or excluding direct sales from the MBS sales and procurement markets has no material effect on the Commission's assessment.

172    Which are present directly or through affiliates.

173    Such as Syntese, OMI, Mindmill, Atcore, and Brandspot.

174    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

175    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

176    The ratio takes into account wins through the same incumbent agency or another agency within the group. Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

177    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

178    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30.

179    Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

180    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

181    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

182    See Section IV.2 on methodologies for estimating market shares for MBS.

183    Submission of the Parties of 17 December 2013.

184    As the level of direct sales is very low (1%) for TV, including or excluding direct sales does not materially affect the market share of the merged entity on the procurement side of the MBS market for TV.

185    http://www.b.dk/kultur/skaerpet-kamp-paa-biografmarkedet

186    Which are present directly or through affiliates.

187    Such as MyMedia, Agence Business, Australie, Climat Media, CoSpirit Media Track, Fred & Farid Group, FullSIX, Oconnection.

188    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

189    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

190    The ratio takes into account wins through the same incumbent agency or another agency within the group. See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

191    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

192    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30.

193    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

194    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

195    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

196    The exclusion of direct sales would not materially affect the conclusion of the competition assessment on these channels.

197    See Section IV.2 on methodologies for estimating market shares for MBS.

198    Conseil Supérieur de l’Audiovisuel, Les chiffres clés de l’audiovisuel français, 1er semestre 2013, page 9, available at: http://www.csa.fr/content/download/33430/432161/file/Chiffres_cles_de_laudiovisuel-2013- S1.pdf

199    CNC, Ressources, Statistiques par secteur, Financement de la télévision, data available at http://www.cnc.fr/c/document_library/get_file?uuid=c5d48b33-3388-430e-933d- bd306a6406c8&groupId=18

200    NRJ Group (24%), RTL Group (23%), Lagardère (18%) and Next Radio TV (15%) – Source: Conseil Supérieur de l’Audiovisuel, Les chiffres clés de l’audiovisuel français, 1er semestre 2013, page 31, available at: http://www.csa.fr/content/download/33430/432161/file/Chiffres_cles_de_laudiovisuel-2013- S1.pdf

201    Form CO, paragraph 396. Moreover, CBS Outdoor, previously Giraudy Viacom Outdoor, is also a significant actor in outdoor advertising in France. More information about CBS Outdoor available at http://www.cbsoutdoor.fr/

202    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 37; replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 17; and replies to Commission questionnaire Q2 to Customers of 25 November 2013.

203    Which are present directly or through affiliates.

204    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

205    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

206    The ratio takes into account wins through the same incumbent agency or another agency within the group. See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

207    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

208    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30.

209    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

210    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

211    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

212    See Section IV.2 on methodologies for estimating market shares for MBS.

213    The public broadcasters only play a minor role in the sale of airtime.

214    See           RTL         Group     Corporate                Presentation,           March     2013,       p.52.        Available at http://www.rtlgroup.com/www/assets/file_asset/RTL_Group_Corporate_Presentation_March_2013.pdf. and various decisions of the German Competition Authority (Bundeskartellamt) involving these two TV broadcasters: Decision of 17 March 2011, B6-94/10 ProSiebenSat.1 Media AG/RTL interactive GmbH, paras. 43, 66, 101; Decision of 11 April 2006, B 6-142/05 RTL Television/n-tv, pages 12, 17 ff.; Decision of 19 January 2006, B6-92292-Fa-103/05 Axel Springer/ProSiebenSat.1 Media AG, pages 25, 30 ff. The trade association "Zentralverband der deutschen Werbewirtschaft ZAW"; in its yearbook "Werbung in Deutschland 2013" estimates the combined share of private TV broadcasters based on net TV advertising revenues is at approximately 93%.

215    Annex 41 to the Form CO, para 135.

216    Bundeskartellamt decision of 15 August 2001, B6-92202-TX-127/99 RMS/Radio Aachen, page 17 ff. For 2012, this share is confirmed by trade association „Zentralverband der deutschen Werbewirtschaft ZAW“; in its yearbook „Werbung in Deutschland 2013“ the combined share of RMS and AS&S based on net revenues for radio advertising is estimated at 89.5 %

217    Annex 41 to the Form CO, para. 135.

218    Annex 41 to the Form CO, para 135.

219    See           the           Bundeskartellamt's  press       release     available  at: http://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2007/30_11_2007_share- deals.html

220  See Parties' response to Commission's RFI of 3 January 2014.

221    Bundeskartellamt decision of 17 March 2011, B6-94/10 ProSiebenSat.1 Media AG/RTL interactive GmbH, paras. 99 and 125.

222    One specificity to the negotiation process in Germany is that some negotiations are done directly by the advertiser with the owner, although the media buying agency is in charge of putting together the media mix and strategy. The consequence of this process is that the volume directly purchased by these clients has no impact on the agency’s AVB.

223    Annex 46 and 47 to the Form CO.

224    Annex 41 to the Form CO, para. 137.

225    Form CO, footnote 46.

226    Which are present directly or through affiliates.

227    WPP offers MBS in Greece through its subsidiary MEC; Mediacom, another subsidiary of WPP is also present in Greece. It is however not clear if, or to what extent it offers MBS in Greece.

228    Such as Fortune, Havas, Leoussis group, and Oxygen.

229    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

230    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

231    The ratio takes into account wins through the same incumbent agency or another agency within the group. Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

232    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

233    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30.

234    Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp". No conclusion can be drawn as regards Publicis' wins since the complete dataset only includes one win of Publicis.

235    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

236    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

237    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

238    See Section IV.2 on methodologies for estimating market shares for MBS.

239    European Audiovisual Observatory 2012 Yearbook (Vol. 1), page 147.

240    Which are present directly or through affiliates.

241    Such as Berg Media, Allison Advertising and Café Communications.

242    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

243    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

244    The ratio takes into account wins through the same incumbent agency or another agency within the group. See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

245    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

246    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30.

247    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

248    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

249    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

250    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

251    See Section IV.2 on methodologies for estimating market shares for MBS.

252    Form CO, paragraph 569.

253   See Council of Europe, “TV and on-demand audio visual services in Hungary”, report available at http://mavise.obs.coe.int/country?id=16.

254    See http://www.jcdecaux.com/en/Newsroom/Press-Releases2/2012/JCDecaux-reenters-Hungary, consulted on 19 December 2013.

255    Which are present directly or through affiliates.

256    Although Havas is not mentioned by the Parties in the form CO, Havas has indicated to the Commission that it is present in MBS in Ireland. See reply of Havas to Commission questionnaire Q1 to Competitors of 25 November 2013, question 5.

257    Such as Gaffney McHugh Advertising, GT Media, Pierce Media & Advertising, Shanley Media Solutions, Southern Advertising Agency.

258    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

259    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

260    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

261    See submission of the Parties of 18 December 2013, tab " Country_Comp", table 26.

262    See submission of the Parties of 18 December 2013, tab " Country_Comp", table 30.

263    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

264    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

265    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

266    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

267    See Section IV.2 on methodologies for estimating market shares for MBS.

268    See Outdoor Media Association: http://www.oma.ie/oma/www/index.asp?magpage=2. According to the Parties, the concentration in the Irish outdoor advertising segment increased further in April 2013 when JCDecaux, the leading provider, acquired Bravo Outdoor Advertising.

269    Form CO, Annex 41 Competitive Analysis of Ten Additional EU Member States, paragraph 190.

270    According to a report from "Autorità per le garanzie nelle comunicazioni" ("the AGCOM report") in Italy, the MBS market had the following main competitors in 2010: WPP (market share: 40-60%), Aegis (0- 20%), Publicis (0-20%), Omnicom (0-20%) and other international and independent operators with lower market shares. The position of different players did not change if MBS by media type were considered. Report of the “Autorità per le Garanzie nelle Comunicazioni – Indagine Conoscitiva sul Settore della Raccolta Pubblicitaria”, p.60, annexed to Delibera n. 551/12/CONS (Chiusura dell'indagine conoscitiva sul settore della raccolta pubblicitaria, avviata con Delibera n. 402/10/CONS) of 28 November 2012

271    Which are present directly or through affiliates.

272    Such as Ammiro Partners, Cayenne, FullSIX.

273    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

274    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

275    The ratio takes into account wins through the same incumbent agency or another agency within the group.

276    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

277  See submission of the Parties of 18 December 2013, tab " Country_Comp", table 26.

278  See submission of the Parties of 18 December 2013, tab " Country_Comp", table 30.

279    No conclusions can be drawn with regard to Omnicom's losses as the complete dataset only includes one tender in which Omnicom was the incumbent and lost the tender.

280    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

281    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

282    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

283    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

284    The Parties' combined market shares on outdoor was calculated by the Parties on the basis of a different market size than the one estimated in the ZenithOptimedia report (EUR 390 million estimated on the basis of the Italian Association Assocom). The Parties explained that, as the ZenithOptimedia report for outdoor was based on AC Nielsen data, it underestimated total outdoor spending by around 50%. This under- estimation was due to both the scattered nature of the monitored formats and the absence of monitoring of some major formats, such as "maxi or big size" posters, and of special supports such as street furniture and long-term sites.

285    See Section IV.2 on methodologies for estimating market shares for MBS.

286    The Parties explain that these two cinema groups offer a total capacity of approximately 180.000 seats across Italy, which, taking into account the multiple screenings of films, represents a very significant portion of the available ad time.

287    Response to Commission questionnaire Q3 to Customers of 25 November 2013 question 23. This is are confirmed by the AGCOM report, which states that this market is competitive and barriers to entry were low, The AGCOM report stated in that respect that "Proprio l'affermazione di Moviemedia, che in pochi anni è riuscita a guadagnare una quota del 26% in volume e dell'11% in valore, è di per sé evidenza di una certa contendibilità del mercato. Esiste, infine, una non marginale porzione di esercenti, allo stato attuale, non aderente a nessun network (67% in termini di cinema e 38% in termini di schermi), che lascia spazio all'ingresso e/o alla crescita di nuovi soggetti" ("The very fact that Movimedia in a few years managed to gain a share of 26% in volume and 11% in value demonstrates the existence of a contestable market. Finally, the not insignificant presence of independent operators not belonging to any network (67% in terms of film and 38% in terms of screens) leaves room for the entry and / or the expansion of new players"). See AGCOM report, page 178.

288    Response to Commission questionnaire Q3 to Media Owners of 25 November 2013, question 24.

289    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 38.

290    For all these media segments see AGCOM report, pages 149, 151-161, 162-174 and 212-213.

291    Form CO, paragraph 441.

292    See AGCOM report, Table 4.9 page 135.

293    See AGCOM report, page 185.

294    Which are present directly or through affiliates.

295    Such as MediaFlux and Creative Media Baltic.

296    Submission of the Parties, 4 December 2013 (Question 16 and Annex 2A).

297    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

298    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

299    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

300    See Section IV.2 on methodologies for estimating market shares for MBS.

301    Form CO, paragraph 596.

302    See also Comparison of share of national TV channels in Latvia for May – June 2013: http://www.tns.lv/?lang=en&fullarticle=true&category=showuid&id=4271 (consulted on 19 December 2013).

303    See Aegis Global Advertising Expenditure Report, May 2013, page 69.

304    See Aegis Global Advertising Expenditure Report, May 2013, page 69. Google was also identified as the most popular online/mobile resources for Latvian Internet users http://ejc.net/media_landscapes/latvia (consulted on 19 December 2013).

305    See   http://www.jcdecaux.lv/en/jcdecaux-latvia

306    The advertising space of all these four operators is sold through the intermediary of the same sales house, Pirmalinja.

307    See Aegis Global Advertising Expenditure Report, May 2013, page 69.

308    Which are present directly or through affiliates.

309    Such as Creative Media Services and Mediapool.

310    Submission of the Parties, 4 December 2013 (Question 16 and Annex 2A).

311    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

312    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

313    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

314    See Section IV.2 on methodologies for estimating market shares for MBS.

315    The third biggest broadcaster in terms of audience share is LNRT, with audience share of 11.3%. However, LNRT could not be considered as a big player on the Lithuanian TV advertising segment, given the low operating revenues realised in 2011. See Form CO, paragraph 627.

316    See Form CO, paragraph 627.

317    See: http://www.pirmalinija.lv/en/.

318    Form CO, paragraph 627.

319    Which are present directly or through affiliates.

320    Such as Brandwebbing, De Media Balie, MediaXplain, SVBmedia and Services, Media Balance, Zigt, M2 Media, De Media Maatschap, Stroom, and Zuiver Media.

321    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

322    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

323    The ratio takes into account wins through the same incumbent agency or another agency within the group. Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

324    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

325    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30.

326    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

327    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

328    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

329    See Section IV.2 on methodologies for estimating market shares for MBS.

330    European Audiovisual Observatory 2012 Yearbook (Vol. 1), page 241.

331    See     http://www.mediamonitor.nl/mediamarkten/radio/radio-in-2012/.

332    See www.g2mi.com/country_sector_info.php?sectorName=Outdoor%20advertising&countryName=Netherlan ds&id=80.

333    Which are present directly or through affiliates.

334    Such as Ecorys, Infinity Media, Media Concept, Pro Media House, Star Media.

335    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

336    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

337    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

338    See submission of the Parties of 18 December 2013, tab " Country_Comp", table 26.

339    See submission of the Parties of 18 December 2013, tab " Country_Comp", table 30.

340    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

341    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

342    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

343    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

344    See Section IV.2 on methodologies for estimating market shares for MBS.

345    See replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 27.

346    See replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 23.

347    European Audiovisual Observatory 2012 Yearbook (Vol. 1), page 262.

348    Parties' submission of 17 December 2013. See also Global Media Marketing Intelligence, available at: http://www.g2mi.com/country_sector_info.php?sectorName=Outdoor%20advertising&countryName=Pol and&id=174.

349    See replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 23.

350    Official promotional Website of the Republic of Poland, Press and publishers, available at http://en.poland.gov.pl/Press,and,publishers,7226.html. European Journalism Centre, available at http://ejc.net/media_landscapes/poland indicates that each of Fakt and Gazeta Wyborcza represent around 15% of the market, which would in any event not change the assessment on the press channel.

351    Form CO, paragraph 691.

352    European Journalism Centre, available at http://ejc.net/media_landscapes/poland.

353    Joinville, available at http://joinville.se/guide-to-the-online-advertising-market-in-poland/

354    More information may be found at www.filmneweurope.com/country-profiles/45-poland/menu-id-235.

355    Which are present directly or through affiliates.

356    Other players on this market include Milenar and Mediagate.

357    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

358    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

359    The ratio takes into account wins through the same incumbent agency or another agency within the group.

360    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

361    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

362    See submission of the Parties of 18 December 2013, tab " Country_Comp", table 30.

363    No conclusions can be drawn with regard to Publicis' losses as the complete dataset only includes one tender in which Publicis was the incumbent and lost. See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

364    No conclusions can be drawn with regard to Omnicom's wins as the complete dataset only includes one tender in which Omnicom won. See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

365    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

366    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

367    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

368    See Section IV.2.on methodologies for estimating market shares for MBS.

369  See Form CO, paragraph 722.

370    See Form CO, paragraph 722.

371    Submission of the Parties of 7 January 2014. See also http://www.alexandrepais.pt/cofina-aumenta-de-32- para-34-a-sua-quota-de-mercado/ (the figures in the table correspond to the period January to August 2012).

372    Submission of the Parties of 8 January 2014.

373    See Form CO, paragraph 722.

374    See Form CO, paragraph 722.

375    Which are present directly or through affiliates.

376    Other players on this market include Media Today, Next Advertising and Springer&Jacoby.

377    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

378    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

379    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

380    The ratios for both Publicis and Omnicom take into account wins through the same incumbent agency or another agency within the group.

381    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

382    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30.

383    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

384    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

385    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

386    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

387    See Section IV.2 on methodologies for estimating market shares for MBS.

388    The Romanian Audiovisual Law no 504/2002 as amended by the Government Emergency Ordinance 25/12.04.2013, published in the Official Gazette N° 208/12.04.2013.

389    Form CO, paragraph 751 and Aegis Global Advertising Expenditure Report, May 2013, page 78.

390    http://mediafactbook.ro/ebook/wp-content/uploads/2012/06/MediaFactBook2012_Initiative.pdf

391  http://mediafactbook.ro/ebook/wp-content/uploads/2012/06/MediaFactBook2012_Initiative.pdf

392    Which are present directly or through affiliates.

393    The Commission's market investigation indicated that Havas is present in Slovenia.

394    Dentsu-Aegis is present in Slovenia through Carat Slovenia (see http://www.dentsuaegisnetwork.com/en/network/#/emea) and, according to the Parties, it is also affiliated with an independent Slovenian agency Media Publikum (see the Parties' response to Commission's RFI of 13 December 2013).

395    Other competitors include Httpool, iProm and Mediamix.

396    Submission of the Parties, 4 December 2013 (Question 16 and Annex 2A).

397    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

398    One customer expressed concerns as regards the transaction on the MBS market in Slovenia, but these concerns were not substantiated and not in line with the market share data of the Parties or the other repliers to the market investigation.

399    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

400    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

401    See Section IV.2 on methodologies for estimating market shares for MBS.

402    Which are present directly or through affiliates.

403  See Form CO, paragraph 61, and https://www.interpublic.com/our-agencies/find-an-agency

404    Such as Veritas Media, Media by Design, Alma Media, Equmedia, Moon Media, Ymedia, Zertem Communication Group.

405    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

406    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

407    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

408    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

409    See submission of the Parties of 18 December 2013, tab " Country_Comp", table 30.

410    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

411    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

412    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

413    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

414    See Section IV.2 on methodologies for estimating market shares for MBS.

415    Reply of the Parties to a Commission RFI of 27 November 2013 (Question 4), paragraph 73.

416    See replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 23. Apart from JCDecaux and Clear Channel, CBS Outdoor and Espacio exterior have been indicated by media owners as being present in the Spanish outdoor sector.

417    Estudio IndoAdex de la Inversión Publicitaria en España 2013.

418    See CNC’s (Comision Nacional de Competencia) decision of 6 April 2011 in case C-0339/11, CINESA/ACTIVOS UGC.

419    As the level of direct sales is very low for TV (6%) and outdoor (16%), including or excluding direct sales from the MBS sales and procurement markets has no material effect on the Commission's assessment.

420    Which are present directly or through affiliates.

421    The Parties have not provided market shares for Havas; however the Commission's market investigation indicated that Havas is present in Sweden.

422    Other players on this market include Tre Kronor, HowCom, Klirr Stockholm.

423    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

424    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 29.

425    The ratio takes into account wins through the same incumbent agency or another agency within the group. Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

426    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 26.

427    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30.

428    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

429    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

430    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

431    See Section IV.2 on methodologies for estimating market shares for MBS.

432    As the level of direct sales is very low for TV (6%) and outdoor (16%), including or excluding direct sales does not materially affect the market share of the merged entity on the procurement side of the MBS market for TV and outdoor.

433    Form CO, Annex 41, paragraph 278, and Parties' submission of 16 December 2013. See also Parties' submission of 8 January 2014.

434    Parties' submission of 16 December 2013.

435    Parties' submission of 17 December 2013.

436    Which are present directly or through affiliates.

437    In its decision in case ME/5636/12 Dentsu/Aegis, the Office of Fair Trading stated that, based on a combination of internal data (for their shares) and a Zenith Optimedia report (for market size), Dentsu- Aegis would have a market share of [5-15]% of the MBS market in the United Kingdom.

438    Such as CMS Music Media, M.i. Media, NetBooster UK, the7stars, Total Media.

439    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

440    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

441    See submission of the Parties of 18 December 2013, tab " Country_Comp", table 26.

442    See submission of the Parties of 18 December 2013, tab " Country_Comp", table 30.

443    See Commission calculation based on submission of the Parties of 18 December 2013, tab "Addit AnnexC_V_Comp".

444    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

445    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

446    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

447    See Section IV.2 on methodologies for estimating market shares for MBS.

448    European Audiovisual Observatory 2012 Yearbook (Vol. 1).

449    BBC website, How we work: Policies and Guidelines: Advertising, available at http://www.bbc.co.uk/aboutthebbc/insidethebbc/howwework/policiesandguidelines/advertising.html.

450    Extract             from            Statement             published             by            OFCOM            on            15/12/2011 (http://stakeholders.ofcom.org.uk/consultations/tv-advertising-investigation/statement)

451    Newspaper and magazine distribution in the United Kingdom – Introductory overview paper on the newspaper and magazine supply chains, Office of Fair Trading Report, October 2008, available at http://www.oft.gov.uk/shared_oft/reports/comp_policy/oft1028.pdf.

452    European Journalism Centre, Media Landscapes, UK, available at http://ejc.net/media_landscapes/united- kingdom.

453    Ibid.

454    Ibid.

455    BBC Annual Report 2011/12 (part 2), available at http://downloads.bbc.co.uk/annualreport/pdf/bbc_executive_2011_12.pdf.

456    Press Release announcing Global/GMG final report, 21/05/2013, available at http://www.competition- commission.org.uk/media-centre/latest-news/2013/May/global-gmg-final-report.

457    OFT Market Study, Outdoor Advertising, February 2011, OFT1304, para. 1.5, available at http://www.oft.gov.uk/shared_oft/market-studies/oft1304.pdf.

458    Which are present directly or through affiliates.

459    Such as Futatsu Industries, InSight, Mood Communications, Re:Media.

460    The comparison is between overlap tenders and non-overlap tenders, that is to say either tenders in which only one of the Parties is present or tenders in which none of them is present.

461    See submission of the Parties of 18 December 2013, tab " Country_Comp", table 26.

462    See submission of the Parties of 18 December 2013, tab "Country_Comp", table 30

463    See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 39, replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 31 and replies to Commission questionnaire Q3 to Media owners of 25 November 2013, question 19.

464    See replies to Commission questionnaire Q2 to Customers of 25 November 2013, questions 39 and 39.1.

465    See reply to Commission questionnaire Q2 to Customers of 25 November 2013, question 36.1.

466    The exclusion of direct sales does not materially affect the conclusion of the competition assessment on these channels.

467    See Section IV.2 on methodologies for estimating market shares for MBS.

468    Market shares by TV audience. Source: European Audiovisual Observatory 2012 Yearbook, page 251.

469    http://www.nrk.no/informasjon/about_the_nrk/1.3698330.

470    According to the Parties, the Oslo Airport is the largest airport in the Oslo area, and the busiest airport in Scandinavia. Moss Lufthavn Rygge (Rygge Airport) mostly serves low cost airlines and serves Oslo. To the Parties’ knowledge, Sageng Media is responsible for all advertising spaces in the airport. Torp Sandefjord Lufthavn (Torp Airport) also serves Oslo. It manages and sells its own advertising spaces.

471    Form CO, paragraph 657.

472    No analysis is conducted on the procurement side because the procurement markets are national.

473    The Transaction does not raise serious doubts at worldwide level, where the Parties will face competition from the other four large international networks. Respondents to the market investigation refer to the Parties' combined market shares worldwide at [30-40]% (however using RECMA as a basis), See also Annex 23 to the Form CO. Furthermore, the Transaction was cleared by the United States Federal Trade Commission and the Canadian Competition Bureau.

474    No affected market exists at EEA level.

475    See replies to Commission questionnaires Q1 to Competitors of 25 November 2013, question 61 and Q2 to Customers of 25 November 2013, question 50.

476    See replies to Commission questionnaires Q1 to Competitors, of 25 November 2013, question 63 and Q2 to customers of 25 November 2013, question 51.

477    See replies to Commission questionnaires Q1 to Competitors of 25 November 2013, question 65 and Q2 to customers of 25 November 2013, question 52.

478    See replies to Commission questionnaires Q1 to Competitors of 25 November 2013, question 66 and Q2 to customers of 25 November 2013, question 54.

479   The EEA countries where one of the parties is the other's closest competitor are: (i) Belgium and the Netherlands, where Publicis is Omnicom's closest competitor together with WPP; and (ii) Germany, where Omnicom is Publicis' closest competitor, followed by WPP. See replies to Commission questionnaire Q1 to Competitors of 25 November 2013 question 66 and Q3 to Customers of 25 November 2013, question 54.

480  See replies to the Commission questionnaires Q1 to Competitors of 25 November 2013, questions 67, 68, 69, 70, and Q2 to Customers of 25 November 2013 question 57.

481    See replies to Commission questionnaires Q1 to Competitors of 25 November 2013, question 64 and Q.2 to customers of 25 November 2013, question 53.

482    See replies to Commission questionnaires Q1 to Competitors of 25 November 2013, question 62 and Q.2 to customers of 25 November 2013, question 55.

483    See replies to Commission questionnaires Q1 to Competitors of 25 November 2013, questions 67-68 and Q2 to Customers of 25 November 2013, question 57.

484    See replies to Commission questionnaires Q2 to Competitors of 25 November 2013, questions 71-74 and Q3 to Customers of 25 November 2013, questions 58-61.

485    Commission Decision of 12 February 2001 in case M.2291 VNU/AC Nielsen, recitals 10 et seq; Commission Decision of 23 September 2008 in case M.5232 WPP/TNS; Commission decision of 4 September 2012 in Case M.6314 Telefonica UK/Vodafone UK/Everything Everywhere/JV, recitals 197 et seq.

486    See replies to Commission Questionnaire Q1 to Competitors of 25 November 2013, question 76.1.

487    See replies to Commission Questionnaire Q2 to Customers of 25 November 2013, question 63.1.

488    See replies to Commission Questionnaire Q3 to Media owners of 25 November 2013, question 28.1.

489    See replies to Commission Questionnaire Q1 to Competitors of 25 November 2013, question 76.2.

490    See replies to Commission Questionnaire Q1 to Competitors of 25 November 2013, question 76.3.

491    See replies to Commission Questionnaire Q2 to Customers of 25 November 2013, question 63.2.

492    See replies to Commission Questionnaire Q3 to Media owners of 25 November 2013, question 28.2.

493    See replies to Commission Questionnaire Q1 to Competitors of 25 November 2013, question 76.4.

494    See replies to Commission Questionnaire Q2 to Customers of 25 November 2013, question 63.3.

495    In particular, one media owner raised the concern that the merged entity could, post-merger, force them to sell their data only to the merged entity and not to their competitors, and that the merged entity could force them to use the merged entity's data analytics service and not the one of their competitors. The Commission considers that such concern is unfounded, as the merged entity will not have the necessary market power to engage in such foreclosure strategies against media vendors, as set out in section IV.4.

496    See replies to Commission Questionnaire Q3 to Media owners of 25 November 2013, question 28.3.

497    Case T-342/99 Airtours plc v Commission, ECR [2002] II-2585, paragraphs 62 Case C-413/06 P Bertelsmann AG and Sony Corporation of America v Independent Music Publishers and Labels Association (Impala), ECR [2008] I-4951, paragraphs 123 and 124.

498    Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, OJ C 31, 5.2.2004, p.5, points 39-57.

499    See replies to Commission Questionnaire Q2 to Customers of 25 November 2013, question 29.

500    See replies to Commission Questionnaire Q2 to Customers of 25 November 2013, question 29.

501    Annexes 46 and 47 to the Form CO.

502    See replies to Commission Questionnaire Q2 to Customers of 25 November 2013, question 29.

503    Mediavision NV is in turn controlled by Publicis' subsidiary Médias & Régies Europe. Mediavision NV is a "régie" which sells advertising space on behalf of cinema owners. Form CO, paras 404-405 and Annex 41, paragraph 223.

504    Commission Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentration between undertakings OJ C 265, 18 October 2008, page 6, Article 23.

505    Commission Guidelines on the assessment of non-horizontal mergers, Article 25.

506  Form CO, paragraphs 194 and 196.

507  The Parties' shares are as follows: [0-10]% for Publicis and [0-10]% for Omnicom.

508  The Parties' shares are as follows: [0-10]% for Publicis and [0-10]% for Omnicom.

509    The Parties contend that this share, based on ZenithOptimedia Advertising Expenditure Forecast, would be overestimated. The Parties state that it has been calculated on the basis of underestimated data relating to the total demand in the cinema channel. According to another source (France Pub), Publicis' share for the selling of advertising space in cinema would amount to [20-30]%. Form CO, paragraph 196, footnote 60.

510    Commission Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings, OJ C 265, 18.10.2008, p.7, point 30.

511    Commission Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings, OJ C 265, 18.10.2008, p.7, point 59.

512    For an indication of the Parties' market shares on the MBS market and on its segmentation related to cinema, see Section IV.4.1.4.

513    Form CO, paragraphs 758 – 760.

514  See replies to Commission questionnaires of 25 November 2013, Q.1 to Competitors, questions 4 and  77, Q.2 to Customers, questions 3 and 64 and Q.3 to Media Owners, question 29.

515  See replies to Commission questionnaire Q2 to Customers of 25 November 2013, question 4.

516  See replies to Commission questionnaire Q1 to Competitors of 25 November 2013, question 6