Booking Holdings Inc. v European Commission (T-1139/23)

Contents

Judgment of the General Court (Tenth Chamber, sitting with five Judges)9 September 2026^note(Competition — Concentrations — Market for online travel agencies in the hotel sector — Decision declaring the concentration incompatible with the internal market and the EEA Agreement — Guidelines on the assessment of non-horizontal mergers — Counterfactual scenario — Significant impediment to effective competition — Efficiencies)In Case T‑1139/23,Booking Holdings Inc., established in Norwalk, Connecticut (United States), represented by F. González Díaz, R. Snelders, lawyers, and D. Beard, Barrister-at-Law,supported byEtraveli Group AB, established in Uppsala (Sweden), represented by C. Esteva Mosso, L. Crocco, S. Völcker, C. Malamataris and L. Bitsakou, lawyers,vEuropean Commission, represented by F. Castillo de la Torre, P. Caro de Sousa, M. Domecq and T. Franchoo, acting as Agents,THE GENERAL COURT (Tenth Chamber, sitting with five Judges),Composed, at the time of the deliberations, of M. van der Woude, President, M. Jaeger, L. Madise, P. Nihoul and S. Verschuur (Rapporteur), Judges,Registrar: I. Kurme, Administrator,having regard to the written part of the procedure,further to the hearing on 8 and 9 July 2025,gives the following

Judgment

1By its action under Article 263 TFEU, the applicant, Booking Holdings Inc., seeks the annulment of European Commission Decision C(2023) 6376 final of 25 September 2023 declaring a concentration to be incompatible with the internal market and the functioning of the EEA Agreement (Case M.10615 – Booking Holdings/Etraveli Group), as found in Annex A1 to the application (‘the contested decision’).

I. The background to the dispute

A. The entities in question

2The applicant and Flugo Group Holdings AB, which operates under the trade name Etraveli Group (‘the intervener’), are online travel agencies (‘OTAs’) and are mainly active in different service markets in the European Economic Area (EEA).

3Online travel agencies are online retailers which resell to consumers one or more types of travel service, such as flights, accommodation and car rental, which are supplied by travel service providers (‘TSPs’), such as airlines, accommodation or car rental companies (recital 51 of the contested decision).

4OTAs provide, more specifically, search, comparison and booking services to consumers (‘business-to-consumer’ or ‘B2C’ segment of the market), as well as marketing services and booking functionalities for TSPs (‘business-to-business’ or ‘B2B’ segment of the market). Accordingly, OTAs act as intermediaries between consumers and TSPs, but derive most of their revenues from the commissions they charge to TSPs (recitals 51 and 52 of the contested decision).

5The applicant is a company established and listed on the stock exchange in the United States, which, while it operates OTAs for a number of travel services under various brands, such as Rentalcars, Priceline and Agoda, is mainly active as an accommodation OTA, under the Booking.com brand. It also provides access to its accommodation functionalities via commercial affiliated agreements which it concludes with rival hotel OTAs which do not have such capability themselves (recital 2 of the contested decision).

6The intervener is a company established in Uppsala (Sweden) which operates as a flight OTA via various brands, such as Gotogate, Mytrip, Flightnetwork, travelstart, Seat24 and SuperSaver (recital 3 of the contested decision). It has also concluded a number of commercial affiliated agreements with other OTAs, namely the applicant, [confidential]^note (recitals 665 and 966 of the contested decision).

7In 2019, the applicant launched its flight offering by concluding a first commercial affiliated agreement with the intervener. In that regard, it is apparent from the documents before the Court that the cooperation between the parties began on 10 January 2019, under a simplified contract, which was replaced by the so-called ‘Phase 1’ agreement on 27 August 2019. After a series of amendments and following negotiations which lasted from January to June 2021, the parties decided to deepen their cooperation by signing the so-called ‘Phase 2 Agreement’ on 9 June 2021 (‘the Phase 2 Agreement’). Lastly, on 10 August 2023, the applicant and the intervener agreed an amendment to the Phase 2 Agreement.

8By means of the concentration at issue, the applicant would acquire sole control, as referred to in Article 3(1)(b) of Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations between undertakings (the EC Merger Regulation) (OJ 2004 L 24, p. 1), of the intervener’s business, excluding the subsidiary Flightmate AB, which operates a metasearch services (‘MSS’) business under the Flygresor brand and is mainly active in Sweden (‘the transaction’) (recitals 3 and 4 of the contested decision).

B. The administrative procedure

9Given that the transaction did not have a Community dimension within the meaning of Article 1(2) and (3) of the EC Merger Regulation and could have been reviewed under the national merger control laws of Germany, Cyprus and Austria, the applicant and the intervener submitted, in accordance with Article 4(5) of that regulation, a reasoned submission to the Commission on 14 February 2022 requesting that the transaction be referred to it. Since no Member State objected to that request within the time limit laid down, it was accepted on 9 March 2022 (recitals 6 and 7 of the contested decision).

10The administrative procedure commenced on 10 October 2022, as the transaction was notified to the Commission on that date (recital 9 of the contested decision).

11On 16 November 2022, the Commission adopted a decision to initiate proceedings pursuant to Article 6(1)(c) of the EC Merger Regulation.

12On 9 June 2023, the Commission sent a statement of objections to the applicant (recital 22 of the contested decision), in which it preliminarily concluded that the transaction would significantly impede effective competition in a substantial part of the internal market within the meaning of Article 2(3) of the EC Merger Regulation, by strengthening the applicant’s dominant position on the hotel OTA market in the EEA. On 24 June 2023, the applicant responded to that statement of objections, disputing those preliminary conclusions (recital 24 of the contested decision).

13On 31 July 2023, the applicant offered commitments pursuant to Article 8(2) of the EC Merger Regulation, in order to address the concerns identified by the Commission (recital 36 of the contested decision).

14On 15 August 2023, the applicant informed the Commission that it and the intervener had agreed to modify the Phase 2 agreement (recital 614 of the contested decision).

15On 25 August 2023, the applicant submitted a revised set of commitments (recital 45 of the contested decision).

16Following the meeting of the Advisory Committee of 13 September 2023 and the favourable opinion of the Hearing Officer on the conduct of the procedure set out in her report of 15 September 2023, the Commission adopted the contested decision (recitals 48 and 49 of the contested decision).

II. The contested decision

17By the contested decision, the Commission declared the transaction incompatible with the internal market and the EEA Agreement, finding, inter alia, as set out below.

A. The relevant markets and the position of the players present on those markets

18In the contested decision, the Commission identified, as relevant markets, the hotel OTA market, on the one hand, and the flight OTA market, on the other, each covering the entire EEA (Section 5 of the contested decision).

19As regards the hotel OTA market, the Commission found that, in 2022, the applicant was the largest player with market shares of approximately [60-70]% for B2B and [60-70]% for B2C, followed by Expedia, which had a market share of [10-20]% for B2B and [5-10]% for B2C. The rest of the market consisted of competitors, such as HRS, Airbnb, OYO, Travelminit, Weekendesk, Trip.com, Lastminute, TUI, eDreams Odigeo, Kiwi and Tix, which had modest market shares (recitals 215 and 216 of the contested decision).

20As regards the flight OTA market, the Commission estimated that, in 2022, the intervener ([10-20]% market share) and the applicant ([5-10]% market share), together occupied the position of the second largest player, after eDreams Odigeo, which held a [20-30]% market share, with the market shares of other flight OTA being as follows: Trip.com [10-20]%; Lastminute [5-10]%; Kiwi [5-10]%; Travelgenio [5-10]%; Expedia [0-5]%; and others [20-30]% [recital 824 (Table 13) of the contested decision].

B. The applicant’s dominant position on the hotel OTA market

21In the contested decision, the Commission found that the applicant held a dominant position on the hotel OTA market, not only as a result of its market share (see paragraph 19 above and recitals 340 to 362 of the contested decision), but also because it received an average effective commission which was [confidential]% higher than that received by rival OTAs (recitals 373 to 375 of the contested decision), because it was an unavoidable trading partner for hotels and was able to impose certain general conditions on them (recitals 410 to 426 of the contested decision), because it was not subject to sufficient competitive pressure from rival OTAs or sufficient out-of-market constraints (recitals 435 and 436 of the contested decision) and because there was no evidence of entry or expansion likely to act as an effective constraint on it (recitals 472 to 488 of the contested decision).

22In that context, the Commission noted that, for each parameter determining the competitiveness of a hotel OTA, the applicant outperformed its competitors, whether it be the breadth of its hotel portfolio, its customer base, the level of the actual fees which it charged hotels, its advertising strength, its ability to attract online traffic, brand recognition or access to customer data (recitals 491 to 557 of the contested decision).

23The Commission also found that the applicant’s dominant position on the hotel OTA market was strengthened by the fact that there were high barriers to entry and expansion in that market, on account of the significant network effects on that market and the high degree of customer inertia, in so far as a significant proportion of customers would go directly to the website or the app which they would generally use to make bookings, without comparing prices on other sites (recitals 558 to 565 of the contested decision, referring to recitals 217 to 251 of that decision).

24In that context, the contested decision explains that the attractiveness of an OTA for a hotel depends on the number of end customers that the OTA is able to attract. The greater that number, the more attractive it is for hotels to be listed on the platform of the OTA at issue and to offer better content and rates there. Similarly, the greater the number of hotels on an OTA, the more attractive the OTA is for end customers. The Commission added that the interdependence between the number of hotels on the hotel OTA platform and the number of end customers using that platform resulted in a self-reinforcing dynamic which is difficult to break for smaller OTAs and for new entrants (recitals 217 to 225 of the contested decision).

25Lastly, it is argued that network effects were aggravated by additional barriers which smaller hotel OTAs and new entrants would face, in particular the difficulty in entering into and managing bilateral contracts with a large number of hotels and an inability to generate online traffic to their platform by investing in online advertising and in search engine optimisation (recitals 226 to 232 of the contested decision).

C. The incentive for a hotel OTA to include a flight offering on its platform

26In the contested decision, the Commission highlighted the advantages for a hotel OTA in including a flight offering in its platform, namely the ability to generate additional customer traffic and better access to customer data making it possible to target potential hotel customers with specific offers.

27Accordingly, the Commission explains that flights are the ‘entry point’ of the trip, since they often constitute the beginning of the booking journey and the service with the highest opportunity to be booked on the same website as accommodations (recitals 587, 753, 756 and 928 to 930 of the contested decision). In addition, it is stated that flights create a significant opportunity for cross-selling other travel services in general and accommodation in particular (recitals 590 to 592 of the contested decision) and that, in that context, they represent the most effective ‘vertical’ for attracting new customers (recital 605 of the contested decision). Lastly, the Commission observes that the connected trip strategy, that is to say, the sale of a range of travel services, of which flights are an essential part, is likely (i) to increase the loyalty of existing customers and to decrease cancellations (recitals 594 and 596 of the contested decision) and (ii) to attract and retain high-value customers who [confidential] (recitals 594, 596, 598 and 933 to 938 of the contested decision).

28In that regard, it is stated that, in order to enable further growth on the hotel OTA market, any additional customer traffic, regardless of whether flights are sold, presented an opportunity for the applicant to attract new hotel customers, to create a new customer relationship and to cross-sell (recitals 769, 770 and 977 to 981 of the contested decision).

29More specifically, when searching for a flight on an OTA platform, a customer necessarily provides information regarding their profile and planned journey, which could be collected by the OTA and used to target that customer with hotel offers tailored to their planned journey and preferences (recital 771 of the contested decision).

30Indeed, if the customer carried out the search using a personal account, the OTA would have access not only to their email address, which would enable it to send targeted advertisements by email, but also information on the customer’s booking history, which would further enable it to tailor its offers to the customer’s profile; such offers could also be displayed either on the OTA’s website when the customer was visiting it, or elsewhere whilst the customer was surfing the internet (recital 771 of the contested decision).

31However, if the customer carried out the search using an app on a mobile device, the OTA could send the customer notifications to that device or show other types of prompts on the app used (paragraph 771 of the contested decision).

32As such, the possibility for cross-sales is not limited to the actual moment of a sale of the flight, but could also occur prior to the sale or after it, because of the data in the possession of the OTA (recital 772 of the contested decision).

33Lastly, the Commission observed that when a customer actually books flights on an OTA platform, that platform has more accurate information, allowing it to target the customer with advertisements in a similar way to that outlined in paragraphs 29 to 32 above. In that context, it is explained that (i) where a customer books a flight, the customer’s travel plans are confirmed, making it more likely that there is demand for a hotel booking, (ii) when booking flights, the customer typically has to provide more personal details, (iii) a customer who booked a flight on a specific platform is more inclined to also book a hotel with that platform and (iv) after booking the flight, the customer has the possibility to consult the booking on the platform, which offers OTA additional cross-selling opportunities by way of targeted offers (recitals 775 and 776 of the contested decision).

D. How the applicant currently operates its flight offering integrated into its platform

34In the contested decision, the Commission explains that the transaction forms part of the applicant’s connected trip strategy, by which it seeks to create a comprehensive travel ecosystem to enable customers to book multiple travel services on the same platform, such as different types of accommodation, flights, taxis, car rental and the purchase of tickets to attractions (footnote 229 to the contested decision and recitals 572, 577 and 585 of that decision). Indeed, according to the Commission, developing a flight offering is the cornerstone of that strategy and an important tool for stimulating growth in the applicant’s accommodation business (recital 747 of the contested decision).

35In that context, it is stated that, under the simplified contract referred to in paragraph 7 above, which was in force between 10 January and 27 August 2019, customers searching for a flight on the applicant’s platform were redirected to the intervener’s platform (recital 686 of the contested decision), whereas the so-called ‘Phase 1’ agreement, in force between 27 August 2019 and 9 June 2021, allowed the applicant to integrate the intervener’s flight offerings onto its platform, both on its website and its mobile app, since it received the intervener’s flight content via an application programming interface. The Commission observed that, since that time, the applicant has been offering the intervener’s content under its own brand. According to the Commission, [confidential] (recital 687 of the contested decision).

36Under the Phase 2 agreement, in force between 9 June 2021 and 15 August 2023, the applicant was authorised, for the first time, to sell the intervener’s content via MSSs, such as Google Flights or Skyscanner, but, given the importance of that channel for flight OTAs in general and for the intervener in particular, that agreement provided for a cap on the number of flights which the applicant could sell via MSSs (‘the meta cap’) (recital 612 of the contested decision). The Commission stated, in that regard, that that clause was intended to protect the intervener from the ‘cannibalisation’ of its flight OTA business in the B2C segment of the market, which refers, in that context, to competition from the applicant selling the intervener’s flight content under its own brand via the MSS channel (recital 708 of and footnote 949 to the contested decision).

37Lastly, the Commission observed that, pursuant to the amendment to the Phase 2 agreement dated 15 August 2023, inter alia, [confidential] (recital 614 of the contested decision).

38In that regard, in recitals 778 to 792 and 958 to 962 of the contested decision, the Commission described how the applicant already uses the intervener’s flight offering on its own platform to strengthen its hotel OTA business, thereby enabling it to create additional customer traffic and giving it access to customer data which it uses to target potential hotel customers with specific offers.

E. The Commission’s theory of harm

39In Section 6.7.2 of the contested decision, the Commission found that the transaction would make the applicant’s dominant position less contestable by increasing barriers to entry and expansion for rival hotel OTAs and would likely lead to harm for hotels and consumers.

40In the first place, the Commission recalled, as stated in Section 6.4 of the contested decision, that, in its view, the applicant already held a dominant position on the hotel OTA market, on which other competitors were weak and unable to constrain it.

41In the second place, the Commission set out, in Section 6.7.2.2 of the contested decision, how the transaction would enable the applicant to acquire its own flight solution in order to use its flight offering as leverage to acquire customers for its hotel booking business in a context where it already dominated the other customer acquisition channels (recital 748 of the contested decision).

42Indeed, according to the Commission, [confidential] (recital 591 of the contested decision) and have [confidential] (recital 592 of the contested decision) as well as the highest potential for cross-selling to hotel OTA services (recitals 750 to 759 of the contested decision).

43The Commission noted that, according to the applicant’s own estimates, between [confidential]% and [confidential]% of total demand for hotel OTA services also required flight OTA services (recital 748 of the contested decision). According to the Commission, that evidence demonstrates that, by acquiring a flight OTA, the applicant would obtain an additional means of capturing a significant part of the demand for hotel OTA services.

44Furthermore, the Commission stated that, by acquiring the intervener, the applicant would obtain a significant volume of additional customer traffic, which, in turn, would offer it additional cross-selling opportunities for its hotel OTA services (recitals 767 to 800 of the contested decision). The Commission found that the intervener was already a leading flight OTA, with a consistent growth path (recitals 826 to 839 of the contested decision), and that the transaction would enable it to accelerate the growth of its business to become the leading provider in that area by 2026 (recitals 841 to 868 of the contested decision) and to develop the opportunities for cross-selling, from which the applicant would benefit in the short and medium term.

45In the third place, in Section 6.7.2.3 of the contested decision, the Commission explained how the acquisition of the intervener would increase barriers to entry and expansion for other existing hotel OTAs and potential new competitors.

46First, the Commission found that the significant network effects which already characterised the hotel OTA market would increase, since the applicant would be able to strengthen its position in the B2C segment of the market, which, in turn, would make its platform even more attractive for hotels (recitals 926 to 932 and 955 of the contested decision).

47The Commission stated, in particular, that, as a result of the acquisition of the intervener, the applicant would have access to additional customer data, which would enable it to re-target consumers more efficiently and to attract more traffic (recitals 510, 511, 588, 589, 769 to 772 and 946 of the contested decision).

48Furthermore, according to the Commission, the transaction would strengthen customer loyalty and customer inertia. It is apparent from the applicant’s internal documents that having an ecosystem of travel services makes its platform, in its own words, ‘stickier’ for new and existing customers (recitals 933 to 938 of the contested decision).

49Second, the Commission noted that, because the applicant would become the leading flight OTA, rival hotel OTAs and potential new competitors would find it increasingly difficult to build a customer base that would make them attractive to hotels (recital 993 of the contested decision).

50According to the Commission, the applicant already dominates the main hotel customer acquisition channels, whereas other travel services or ‘verticals’, such as those relating to the rental of private accommodation or cars, do not give rise to as much traffic as flights (recitals 972, 973, 976, 986 to 992 and 996 to 1000 of the contested decision). It adds that it is unlikely that the applicant would maintain or conclude cooperation agreements with rival OTAs which would allow them to build their own customer base for flights (recitals 974 and 993 of the contested decision).

51In the fourth place, according to the Commission, that increase in barriers to entry and expansion constitutes a structural change in the hotel OTA market affecting the ability of existing and potential competitors to compete with the applicant.

52In that context, the Commission presents various calculations which, in its view, reflect the applicant’s expectations of the increment in its hotel OTA market share, which serve to support its position on that market being uncontestable. More specifically, in Section 6.7.2.4 of the contested decision, the Commission calculated the increase in the applicant’s market share on the basis of two methods derived from the transaction valuation model (‘TVM’) developed by the applicant to assess the value of the intervener’s business and the synergies resulting from acquiring it, including cross-sales of hotel rooms to flight customers. In the context of those methods, which the applicant designates as the ‘modified TVM methodology’ and ‘Figure 123 methodology’, referring to Figure 123 of the contested decision, the Commission claims to have followed, with certain adjustments, the TVM established by the applicant. Accordingly, the increment in the applicant’s hotel OTA market share resulting solely from the gains from combined sales of flights and hotels is [0-5]% under the modified TVM methodology and [0-5]% under the Figure 123 methodology, whereas the increment resulting from all the gains from the connected trip is [0-5]% under the modified TVM methodology and [0-5]% under the Figure 123 methodology. According to the Commission, it follows that the applicant does not dispute that its share of the hotel OTA market post-transaction would increase, but only the extent of that increase.

53Lastly, the Commission stated, in Sections 6.7.2.5 and 6.7.2.6 of the contested decision, that, due to the difficulty in contesting the applicant’s dominant position on the hotel OTA market, hotels and consumers would likely be harmed as a result of the transaction.

54First, according to the Commission, the applicant’s incentives to lower commissions and provide more beneficial terms and conditions would decrease. Similarly, given that the transaction would result in an increase in the number of consumers reserving through the applicant, the costs for hotels would probably be increased, since the applicant is already one of the most expensive sales channels available to hotels (recitals 1079 and 1080 and 1084 to 1111 of the contested decision).

55Second, the Commission observed that (i) hotels could pass on those costs to customers, in particular because of their inertia and their loyalty and (ii) a greater number of customers would book their accommodation via the applicant, even though its prices were, in the majority of cases, higher than in the cheapest alternative (recitals 1099, 1124, 1125 and 1140 of the contested decision). On that basis, the Commission concluded that, post-transaction, end customers could pay higher prices than if the transaction did not proceed (recitals 1134 to 1139 of the contested decision).

III. Forms of order sought

56The applicant and the intervener claim that the Court should:– annul the contested decision;– order the Commission to pay the costs.

57The Commission claims that the Court should:– dismiss the action;– order the applicant to bear its own costs and to pay those incurred by the Commission.– order the intervener to bear its own costs.

IV. Law

58In support of its action, the applicant raises three pleas in law.

59In the first plea, the applicant submits that, in the contested decision, the Commission erred in departing from the Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings (OJ 2008 C 265, p. 6; ‘the Non-Horizontal Merger Guidelines’) and in failing to justify such departure from them in favour of a markedly lower standard.

60In the second plea, the applicant maintains that the Commission made a number of errors when identifying the counterfactual scenario.

61In the third plea, the applicant submits that the contested decision does not meet the standard of proof necessary to demonstrate a significant impediment to effective competition, that it does not prove any of the steps on which the theory of harm is based and that it incorrectly dismisses the efficiencies which the transaction would bring about.

62It is appropriate to examine, first of all, the first plea relating to the application of the Non-Horizontal Merger Guidelines, next, the second plea relating to the appropriate counterfactual scenario, then the first and second parts of the third plea, concerning the Commission’s theory of harm and the standard of proof required in that context, and, lastly, the third part of the third plea, relating to efficiencies.

A. The first plea: an alleged departure when applying the Non-Horizontal Merger Guidelines and the alleged error in classifying pro-competitive advantages

63In the first part of the first plea, the applicant raises three complaints, relating to (i) the allegation that, in the contested decision, the Commission departed from the Non-Horizontal Merger Guidelines by failing, even though the transaction involves a conglomerate merger, to demonstrate the likelihood of anticompetitive foreclosure of the market, as referred to in those guidelines, (ii) the allegation that the Commission did not establish that it was justified in departing from those guidelines in favour of a markedly lower intervention standard and (iii) the allegation that, by departing from those guidelines, the Commission failed to have due regard for the legitimate expectations of the entities concerned.

64In the second part of the first plea, the applicant submits that the Commission was, in any event, required to account for the two principles which underpin the analysis of non-horizontal mergers, namely first, the fact that conglomerate mergers, such as the transaction, are likely to create pro-competitive benefits and, second, the fact that any impact on rivals flowing from such benefits is the result of competition on the merits and not of an impediment to competition.

1. The Non-Horizontal Merger Guidelines and the legal standard for applying Article 2(2) and (3) of the EC Merger Regulation were applied incorrectly

65In the first complaint in the first part of the present plea, the applicant and the intervener submit, in essence, that, since they are active on different markets, the transaction involves a non-horizontal merger, more specifically a conglomerate merger. Accordingly, the Commission ought, in accordance with the Non-Horizontal Merger Guidelines, to have assessed possible anticompetitive foreclosure and, in particular, the ability and incentive of the merging undertakings to foreclose their rivals and any resulting anticompetitive effects. The contested decision incorrectly concludes that the transaction would lead to a significant impediment to effective competition on the basis of a novel and unsubstantiated theory that the transaction would result in a reverse leveraging effect in which the merged entity would use its position on a market in which it does not have market power (the flight OTA market) to strengthen its allegedly dominant position on another market (the hotel OTA market).

66In that context, in response to a question put by the Court in a document dated 10 June 2025, the applicant indicated, at the hearing, the provisions of the Non-Horizontal Merger Guidelines with which, in its view, the Commission’s approach in the contested decision conflicted.

67Accordingly, the applicant maintains, first, that, by relying on a reverse leveraging effect, the Commission failed to observe paragraphs 93 to 104 of the Non-Horizontal Merger Guidelines and, second, that, by assuming that any increase in its position on the hotel OTA market, however minimal, would be sufficient to establish a significant impediment to effective competition, the Commission departed from paragraphs 10, 112 and 113 of those guidelines.

68In addition, the applicant, supported by the intervener, adds that the transaction cannot have horizontal effects, since it combines complementary – not competing – offerings, and that, by alleging in recitals 194 and 195 of the contested decision that the operator ‘active in a different market’ could have ‘horizontal effects’, that decision is at odds with how horizontal and non-horizontal mergers and their respective effects have been distinguished.

69The Commission disputes those arguments.

70As a preliminary point, it should be borne in mind that the Court of Justice has already held, first, that while internal measures adopted by the administration may not be regarded as rules of law which the administration is always bound to observe, they nevertheless form rules of conduct, from which the administration may not depart, in an individual case without giving reasons which are compatible with the principle of equal treatment and, second, that that case-law applies a fortiori to rules of conduct intended to produce external effects, as is the case with guidelines which are aimed at traders (see, to that effect, judgment of 28 June 2005, Dansk Rørindustri and Others v Commission, C‑189/02 P, C‑202/02 P, C‑205/02 P to C‑208/02 P and C‑213/02 P, EU:C:2005:408, paragraphs 209 and 210 and the case-law cited).

71In adopting such rules of conduct and announcing by publishing them that they will henceforth apply to the cases to which they relate, the Commission imposes a limit on the exercise of its discretion and cannot depart from those rules under pain of being found, where appropriate, to be in breach of the general principles of law, such as equal treatment or the protection of legitimate expectations (see judgment of 28 June 2005, Dansk Rørindustri and Others v Commission, C‑189/02 P, C‑202/02 P, C‑205/02 P to C‑208/02 P and C‑213/02 P, EU:C:2005:408, paragraph 211 and the case-law cited).

72First of all, as regards the question whether the transaction would have horizontal or conglomerate effects, it should be recalled that a horizontal merger is defined as a concentration between undertakings which are actual or potential competitors on the same market (paragraph 5 of the Guidelines on the assessment of horizontal mergers under the [EC Merger] Regulation (OJ 2004 C 31, p. 5) ‘the Horizontal Merger Guidelines’), whereas a conglomerate merger is a merger of undertakings which, essentially, do not have a pre-existing competitive relationship (see, to that effect, judgment of 25 October 2002, Tetra Laval v Commission, T‑5/02, EU:T:2002:264, paragraph 142, and paragraph 91 of the Non-Horizontal Merger Guidelines). As stated in paragraph 12 of the Non-Horizontal Merger Guidelines, unlike horizontal mergers, non-horizontal mergers, such as conglomerate mergers, do not entail the loss of direct competition between the merging firms in the same market.

73Accordingly, it should be noted that the activities of the applicant and the intervener are largely complementary, in that the applicant is mainly active as a hotel OTA and the intervener mainly as a flight OTA. It is true that their activities overlap on the flight OTA market, but the applicant does not yet have its own flight solution to supply the flight offering on its platform and it currently purchases flight content from the intervener on the basis of a commercial affiliated agreement. It follows that the relationship between the parties is primarily non-horizontal in nature, despite that overlap.

74Similarly, it must be stated that the theory of harm adopted by the Commission is based largely on leveraging, which is typical of conglomerate mergers, brought about by the transaction. Such leveraging results from the fact that the merged entity would use its position on a certain market as leverage to strengthen its position on another market. According to the Commission, the applicant would, post-transaction, be able to strengthen its position on the flight OTA market, which would enable it to increase the number of flight customers to whom it could sell hotel rooms (see paragraphs 26 to 33 above) and, therefore, to strengthen its position on the hotel OTA market. Accordingly, the transaction would be likely to increase barriers to entry and expansion on that market, thus limiting potential and actual rival hotel OTAs’ access to hotel customers and therefore limiting their ability to access that market and expand in it and to compete in that market (recitals 206, 734, 917, 1079 and 1090 of the contested decision).

75Therefore, the ‘horizontal effects’ identified by the Commission in the contested decision, even if proved, do not follow from the loss of a direct competitor of the applicant on the hotel OTA market, but from a reduction in actual or potential rival hotel OTAs’ ability to compete (paragraph 93 of the Non-Horizontal Merger Guidelines).

76Next, as regards the applicant’s and the intervener’s allegation that the Commission failed, when analysing anticompetitive foreclosure, to assess the ability and incentive of the merging undertakings to foreclose their rivals and any resulting anticompetitive effects, in the first place, it should be noted that in the Non-Horizontal Merger Guidelines the Commission reserved the right to rely on theories of harm other than those expressly set out in those guidelines. As the Commission has correctly pointed out, paragraph 8 of those guidelines states that ‘the principles contained [in this notice] will be applied and further developed and refined by the Commission in individual cases’.

77Furthermore, it should be noted, as the Commission did, that paragraph 93 of the Non-Horizontal Merger Guidelines states that the main competition concern in the context of conglomerate mergers is foreclosure, but does not rule out the possibility that other concerns may arise. In that context, the Commission correctly observes that digital markets have certain specific characteristics and may therefore give rise to competition concerns which were not sufficiently taken into account when those guidelines were adopted (recital 202 of the contested decision).

78Although the Commission is bound by the notices which it issues in the area of merger control, it is not required to undertake an examination in every case of all the factors which are referred to therein. By contrast, the competitive analysis to be carried out by the Commission must not be based on a mechanical application of those factors, but on an overall assessment of the foreseeable impact of the merger, in the context of which not all the elements which that notice identifies are relevant in each and every case (see, to that effect, judgment of 26 June 2025, EVH and Others v Commission, C‑464/23 P, C‑465/23 P, C‑467/23 P, C‑468/23 P and C‑470/23 P, EU:C:2025:478, paragraph 176).

79It follows that the Non-Horizontal Merger Guidelines lay down a general framework for the assessment of such mergers, without thereby limiting the Commission in the exercise of its discretion by all of the elements referred to therein.

80In the second place, it should be noted that the concept of ‘foreclosure’ is defined in paragraph 18 of the Non-Horizontal Merger Guidelines as ‘any instance where actual or potential rivals’ access to supplies or markets is hampered or eliminated as a result of the merger, thereby reducing these companies’ ability and/or incentive to compete’.

81It must be observed that the Commission’s approach in the contested decision aligns with the definition of ‘foreclosure’ set out in paragraph 80 above.

82Indeed, in recital 205 of the contested decision, the Commission states that, in order to demonstrate foreclosure, it is sufficient that rivals’ ability to access the market, or expand in it, is hampered, while stating in other parts of that decision that the transaction would be likely to increase barriers to entry and expansion on the hotel OTA market, thereby limiting potential and actual rival hotel OTAs’ access to hotel customers and therefore their ability to enter that market or expand and compete.

83It follows that the test used in the contested decision is comparable to that used in the Non-Horizontal Merger Guidelines and is consistent with, inter alia, the definition of ‘foreclosure’ laid down therein. However, rather than disputing the fact that the transaction results in a certain foreclosure effect on the market, the applicant calls into question the fact that that effect is sufficiently significant to be regarded as ‘anticompetitive foreclosure’, within the meaning of paragraph 18 of those guidelines, enabling the merged entity profitably to increase its prices, which, according to the applicant and the intervener, is necessary in order to justify prohibiting the transaction under Article 2(3) of the EC Merger Regulation.

84In the third place, it should be recalled that paragraph 94 of the Non-Horizontal Merger Guidelines states that the Commission examines, when assessing the likelihood of foreclosure, first, whether the merged entity would have the ability to foreclose its rivals, second, whether it would have an economic incentive to do so and, third, whether a foreclosure strategy would have a significant detrimental effect on competition, thus causing harm to consumers.

85In that regard, in relation to, first, the applicant’s allegation that the Commission failed to comply with paragraphs 93 to 104 of the Non-Horizontal Merger Guidelines by relying on a reverse leveraging effect (see paragraph 67 above), it should be recalled that it is common ground that one of the objectives pursued by the applicant via the transaction was to increase cross-sales of hotel rooms to flight customers, with the result that it would strengthen its position on the hotel OTA market.

86Consequently, it is also common ground that the applicant would have the ability and incentive to cross-sell hotel rooms to flight customers post-transaction, which corresponds, in broad terms, to the first and second steps set out in paragraph 94 of the Non-Horizontal Merger Guidelines (see paragraph 84 above).

87In addition, as indicated in recital 202 of the contested decision, the Non-Horizontal Merger Guidelines do not preclude a theory of harm from being based on a reverse leveraging effect, given that the Commission did not already have experience of potential concerns regarding such an effect when those guidelines were adopted in 2008.

88In that regard, it should be noted, first, that leveraging is a generic term in relation to the impact which a practice identified on one market may have on another market (judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T‑612/17, EU:T:2021:763, paragraph 163) and, second, that the potential impact on competition, that is to say, the possible strengthening of the merged entity’s position to the detriment of its competitors, is essentially the same, irrespective of the direction of the leveraging effect.

89As such, it would be contradictory to allow the Commission to rely on leveraging only where that leveraging would strengthen a non-dominant position, and not where it would strengthen an already dominant position. Such an approach would be liable to undermine the effectiveness of merger control.

90It follows that, in view of the evolving nature of the Commission’s decision-making practice, it was open to it to base its theory of harm on reverse leveraging, with the result that it is necessary to reject the applicant’s allegation that the Commission did not observe paragraphs 93 to 104 of the Non-Horizontal Merger Guidelines.

91As regards, second, the applicant’s allegation that the Commission departed from paragraphs 10, 112 and 113 of the Non-Horizontal Merger Guidelines when it found that any increase in its position on the hotel OTA market, however small, would be sufficient to prove a significant impediment to effective competition (see paragraph 67 above), it should be noted that those paragraphs refer, in essence, to the need to demonstrate a significant increase in market power post-transaction (paragraph 10 of the Non-Horizontal Merger Guidelines) and to the principle that it is only where a sufficient fraction of market output is affected by foreclosure that the concentration may significantly impede effective competition (paragraphs 112 and 113 of the Non-Horizontal Merger Guidelines).

92As the Commission correctly points out, the departures relied on by the applicant concern alleged divergences from isolated and specific points of the Non-Horizontal Merger Guidelines, which cannot, in general terms, call into question the compatibility of the Commission’s approach in the contested decision with those guidelines and which will be assessed in the context of the first and second parts of the third plea, by which the applicant disputes more specifically a number of aspects of the Commission’s theory of harm.

93Therefore, it is necessary to reject, in the examination of the first plea, the applicant’s allegation that the Commission failed to have due regard for paragraphs 10, 112 and 113 of the Non-Horizontal Merger Guidelines.

94In the light of the foregoing, it must be concluded that the applicant has failed to prove that the Commission’s approach in the contested decision was incompatible with the Non-Horizontal Merger Guidelines.

95Therefore, it is necessary to reject the first complaint of the first part of the first plea and, consequently, the second and third complaints of the first part of the first plea. Indeed, in the context of those last two complaints, the applicant submits that the Commission failed to establish that it was justified in departing from the Non-Horizontal Merger Guidelines and that by departing from those guidelines, the Commission acted in breach of the legitimate expectations of the parties concerned (see paragraph 64 above), which is, however, based on the incorrect premiss that the contested decision departs from the Non-Horizontal Merger Guidelines.

2. The alleged pro-competitive advantages were classified incorrectly

96In the second part of the first plea, the applicant submits, in the alternative, that the Commission was, in any event, required to take into account the two principles underpinning the analysis of non-horizontal mergers, that is to say, in the first place, the principle that conglomerate mergers, such as the transaction, are likely to create pro-competitive benefits (see paragraph 11 of the Non-Horizontal Merger Guidelines) and, in the second place, that any impact on rivals flowing from those benefits results from competition on the merits, and not an impediment to competition.

97The applicant maintains, in particular, that the Commission conducted no analysis to differentiate between the pro-competitive effects and the anticompetitive effects and that it classified pro-competitive advantages and competition on the merits, namely an increase in its ability to cross-sell hotel rooms to flight customers by offering them a one-stop shop facility, as being anticompetitive.

98The Commission disputes those arguments.

99In that regard, in the first place, it should be recalled that recital 29 of the EC Merger Regulation states that in order to determine the effect of a concentration on the structure of competition, the Commission is required to take into account the likely efficiencies put forward by the undertakings concerned.

100In addition, it is apparent from paragraphs 76 and 77 of the Horizontal Merger Guidelines, which are applicable to non-horizontal mergers pursuant to paragraph 6 of the Non-Horizontal Merger Guidelines, that it is possible that efficiencies counteract the adverse effects of a merger on competition. The Commission may therefore decide that there are no grounds for declaring a concentration incompatible with the internal market where that institution is in a position to conclude on the basis of sufficient evidence that the efficiencies generated by the merger are likely to enhance the ability and incentive of the merged entity to act pro-competitively for the benefit of consumers.

101However, it is for the party which has notified a concentration to demonstrate that the efficiencies are likely to counteract any adverse effects on competition that might otherwise result from the merger, and therefore benefit consumers. The argument regarding efficiencies seeks to counteract the Commission’s conclusions that the proposed merger would probably significantly impede effective competition by creating a dominant position (judgments of 6 July 2010, Ryanair v Commission, T‑342/07, EU:T:2010:280, paragraph 427, and of 9 March 2015, Deutsche Börse v Commission, T‑175/12, not published, EU:T:2015:148, paragraphs 262, 361 and 362). It is therefore for the Commission to demonstrate, as a first step, that the transaction would significantly impede effective competition, whereas it is for the parties to the transaction to demonstrate, as a second step, that the efficiencies are likely to counteract the adverse effects which might result from the transaction.

102Accordingly, in so far as the applicant asserts that the Commission ought, at the stage of finding an impediment to competition, to have taken account of the alleged efficiencies, that assertion reflects a misunderstanding of the EC Merger Regulation.

103In so far as the applicant wished to claim that the Commission did not examine sufficiently the efficiencies brought about by the transaction, it should be noted, as the Commission did, that (i) it did in fact conduct such an examination, as is apparent from Section 7 of the contested decision, and (ii) the applicant did not dispute that examination in the context of the second part of the first plea.

104In the second place, as regards the applicant’s allegation that the Commission found competition on the merits which brought pro-competitive advantages to be anticompetitive (see paragraph 97 above), it should be recalled that the concept of ‘competition on the merits’ has been developed in the case-law relating to Articles 101 and 102 TFEU, in order to distinguish between, on the one hand, conduct which is part of normal competition based on the performance of economic operators and, on the other hand, anticompetitive or abusive conduct capable of infringing those articles (see, to that effect, judgments of 12 May 2022, Servizio Elettrico Nazionale and Others, C‑377/20, EU:C:2022:379, paragraphs 73 to 79, and of 21 December 2023, International Skating Union v Commission, C‑124/21 P, EU:C:2023:1012, paragraph 125).

105It is true that the cross-selling envisaged by the applicant does not involve any commercial pressure on customers, but rather leaves them, following the purchase of a flight, the choice whether also to book a hotel room via the applicant. In that regard, such cross-selling differs from tying or bundling, which are referred to in the Non-Horizontal Merger Guidelines as classic situations which may give rise to competition concerns in the context of a conglomerate merger.

106However, as the Commission correctly pointed out at the hearing, the concept of ‘competition on the merits’ does not play a similar role in the context of merger control.

107First, the implementation of a concentration cannot be regarded as a form of competition on the merits and, second, the EC Merger Regulation seeks to establish effective control of all concentrations in terms of their effects on the structure of competition in the European Union (judgment of 13 July 2023, Commission v CK Telecoms UK Investments, C‑376/20 P, EU:C:2023:561, paragraph 106), without any assumption being made that the entity resulting from the concentration will engage in abusive behaviour (see, to that effect, judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T‑612/17, EU:T:2021:763, paragraph 507).

108Therefore, it is necessary to reject the applicant’s argument that the Commission incorrectly found conduct to be anticompetitive which the Non-Horizontal Merger Guidelines identify as competition on the merits providing pro-competitive advantages.

109Consequently, it is necessary to reject the second part of the first plea and, therefore, the first plea in its entirety must be rejected.

B. The second plea: the counterfactual scenario

110In the second plea, the applicant and the intervener submit that, in Section 6.6 of the contested decision, the Commission erred in using a ‘zero-flights’ counterfactual scenario and made a number of other errors when it identified the appropriate counterfactual scenario.

111As a preliminary point, it should be borne in mind that the Commission may declare a concentration incompatible with the internal market only if the significant impediment to competition is the direct and immediate effect of the concentration (see judgment of 11 December 2013, Cisco Systems and Messagenet v Commission, T‑79/12, EU:T:2013:635, paragraph 118 and the case-law cited).

112Accordingly, the Commission compares the conditions of competition that would result from the notified merger with the conditions that would have prevailed without the merger, the latter conditions constituting the ‘counterfactual’ scenario, that is to say, the most likely scenario absent the merger (see, to that effect, judgment of 13 July 2023, Commission v CK Telecoms UK Investments, C‑376/20 P, EU:C:2023:561, paragraphs 84 and 88 and the case-law cited).

113However, even if the competitive conditions existing at the time when the Commission must decide to authorise or prohibit the transaction often constitute the relevant comparison for evaluating the effects of a merger, in some circumstances, the Commission may find it necessary to take into account future changes to the market that can reasonably be predicted and, therefore, form part of the scenario which is most likely absent the merger (paragraph 9 of the Horizontal Merger Guidelines and paragraph 20 of the Non-Horizontal Merger Guidelines), which, although hypothetical, must be realistic and credible (see, to that effect, judgment of 11 September 2014, MasterCard and Others v Commission, C‑382/12 P, EU:C:2014:2201, paragraphs 166 to 169 and 173; see also, by analogy, judgment of 27 June 2024, Commission v Servier and Others, C‑176/19 P, EU:C:2024:549, paragraph 353).

114Accordingly, it is necessary to examine the various arguments by which the applicant and the intervener dispute the Commission’s counterfactual scenario, which are to be grouped together into six complaints, the first three of which relate to the alleged ‘zero-flights’ counterfactual scenario and the last three of which relate to the errors allegedly made by the Commission when identifying that scenario.

1. The alleged zero-flights counterfactual scenario

115In the first place, the applicant and the intervener observe that, while the contested decision itself repeatedly emphasises the importance for the applicant of having a flight offering as the cornerstone of its connected trip strategy, the Commission indicated that the most likely scenario absent the merger would be the applicant withdrawing completely from the flight OTA market.

116In the second place, the applicant disputes the allegation that it would end its cooperation with the intervener absent the merger, in particular if the intervener were sold to one of the applicant’s competitors, arguing that it is in its interest to maintain such cooperation. It would, more specifically, find it advantageous to execute the Phase 2 Agreement and, on the expiry of that agreement, either to extend that cooperation, in which case it would continue to have access to the intervener’s flight solution, or to terminate it, having found a better flight solution elsewhere.

117In the third place, the applicant and the intervener submit that the fact that the merger would transform a contractual relationship into a structural link, as indicated in paragraphs 622, 628 and 673 to 703 of the contested decision, does not mean that, absent the merger, the Phase 2 Agreement would not have remained in force. That fact is therefore irrelevant to identifying the counterfactual scenario.

118The Commission disputes those arguments.

119In that regard, in the first place, it should be observed that, contrary to the applicant’s assertions, the contested decision does not indicate that, absent the transaction, the applicant would withdraw from the flight OTA market.

120Rather, it is apparent from the contested decision that the counterfactual scenario relied on by the Commission takes account of some degree of cooperation between the applicant and the intervener, since it is stated in that decision that the Phase 2 Agreement would not remain in force in the same form, since, according to the Commission, the intervener would, in the short term, apply pressure to renegotiate that agreement, in order to recover, at least in part, the losses resulting from the sales cannibalised by the applicant (recitals 704 to 720 of the contested decision), and that, in the mid- to long-term, the applicant would prefer to have its own flight solution (recitals 677 to 692 of the contested decision).

121In addition, both in its written pleadings and at the hearing, the Commission, in essence, confirmed that it was likely that the applicant would maintain a flight offering on its platform in all scenarios. It follows that the Commission did not rule out the possibility that the cooperation between the applicant and the intervener would be taken into account in the counterfactual scenario on which it relied.

122Therefore, in so far as the applicant’s and the intervener’s line of argument must be interpreted as seeking to refute a ‘zero-flights’ counterfactual scenario, it must be rejected.

123That conclusion cannot be called into question by the fact that the contested decision refers, on a number of occasions, to the concept of ‘zero-flights’.

124As is apparent from recitals 727, 728, 1008 and 1009 of the contested decision, the Commission referred solely to the concept of ‘zero-flights’ when calculating the increase in the applicant’s hotel OTA post-merger market share, since it would be impossible for the Commission, over the short term, to assess a counterfactual scenario on the basis of more limited cooperation between the applicant and the intervener in the absence of reliable information enabling such an assessment to be made and, in the mid- or long-term, to determine whether and by when the applicant would be able to develop its own flight solution and what its capabilities would be.

125In the second place, it must be stated that it is true that, in recitals 693 to 703 of the contested decision, the Commission maintains, first, that, absent the transaction, the intervener would either be sold to a strategic buyer or an IPO would take place and, second, that it is likely that the applicant would end its cooperation with the intervener, since [confidential].

126In that regard, the Commission stated in its defence, first, that it had not concluded that, in the counterfactual scenario, the applicant would immediately end the cooperation, but rather that it would only be likely to do so if the intervener were sold to one of its competitors, and, second, that it had not concluded that such a sale was the most likely scenario, which it also confirmed at the hearing.

127It must therefore be concluded that the Commission did not find that the cooperation between the applicant and the intervener coming to an end, in particular following a sale of the intervener to a competitor of the applicant, was the most likely scenario (see paragraph 112 above).

128In the third place, it is necessary to reject the argument of the applicant and the intervener that the fact that the transaction would transform their cooperation into a permanent structural link is irrelevant for the purpose of determining the counterfactual scenario. Indeed, in the relevant passages of the contested decision, in particular recitals 673 to 703 thereof, the Commission refers to the inherently temporary nature of a contractual collaboration in support of its position that the applicant would prefer, in the mid- or long-term, to have its own flight solution, so as not to depend on third parties for that strategically important activity. Following the logic of the counterfactual scenario as defined by the Commission, that consideration is not irrelevant.

129Therefore, it is necessary to reject the third argument of the second complaint of the first part, the third complaint of the first part and the second part of the second plea.

2. The errors allegedly made by the Commission in identifying the appropriate counterfactual scenario

130In the first place, the applicant and the intervener submit that the Phase 2 Agreement represents the competitive situation existing at the date of the transaction and, therefore, represents the appropriate counterfactual scenario.

131In the second place, the applicant and the intervener dispute the Commission’s allegation that they would not have accepted the terms of the Phase 2 Agreement if they had not envisaged the transaction.

132In that context, the applicant and the intervener maintain that nothing in the history of the negotiation of the Phase 2 Agreement makes that agreement conditional on the transaction being completed and that there is no evidence capable of demonstrating that the conditions of their cooperation were influenced by the prospect of an acquisition. In addition, and above all, the only relevant question in that regard is whether the terms negotiated and signed by them would remain in force absent the transaction, which is the case, with the result that they constitute the counterfactual scenario.

133The applicant and the intervener dispute the allegation that, absent the transaction, the intervener would apply pressure to renegotiate the Phase 2 Agreement.

134In that regard, the applicant and the intervener submit, first, that the intervener cannot unilaterally terminate or renegotiate the Phase 2 Agreement until 2028 and, second, that their cooperation proved to be, for the intervener, a lucrative and important channel for flight sales, with the result that the intervener has no incentive to attempt to renegotiate the terms of that agreement.

135Lastly, the applicant states that the Commission’s theory of harm is necessarily vitiated by an error in the identification of the counterfactual scenario, given the importance of the quantitative assessment of the likely extent of the anticompetitive effects caused by the transaction.

136The Commission disputes those arguments.

137In that regard, it should be recalled, first, that the Commission’s theory of harm must be based on the competitive effects which may be attributed to the transaction, found via a comparison between the merger scenario and the counterfactual scenario at issue (see judgment of 11 December 2013, Cisco Systems and Messagenet v Commission, T‑79/12, EU:T:2013:635, paragraph 118 and the case-law cited; see, to that effect, judgment of 13 July 2023, Commission v CK Telecoms UK Investments, C‑376/20 P, EU:C:2023:561, paragraphs 84 and 88 and the case-law cited) and, second, that the EU Courts may reject as ineffective a plea or complaint where they find that that plea or complaint is not, if it is well founded, such as to give rise to the annulment sought (see judgment of 30 March 2022, Latam Airlines Group and Lan Cargo v Commission, T‑344/17, EU:T:2022:185, paragraph 135, and the case-law cited).

138Accordingly, as the Commission observes in paragraph 84 of the defence, it is necessary to examine whether the errors it is alleged to have made in identifying the counterfactual scenario, even if established, are capable of affecting the examination of the substance of the theory of harm adopted in the contested decision. In particular, it is necessary to examine whether the difference between the counterfactual scenarios put forward by the parties is sufficiently significant to be capable of affecting the outcome of the present action.

139In that regard, it should be noted that that difference is found primarily in the extent of the growth in the applicant’s flight OTA business and the subsequent increase in its share of the hotel OTA market.

140Indeed, although it has been established that, irrespective of the counterfactual scenario adopted, the applicant would have a flight offering (see paragraphs 120 to 122 above), it must be recalled that, in the scenario put forward by the Commission, in particular the scenario of lesser cooperation between the intervener and the applicant after a renegotiation in the short term, the number of flights sold by the applicant would be lower than in the scenario put forward by the applicant, that is to say, cooperation under the Phase 2 Agreement, on account of the more favourable contractual terms to which the applicant would be subject in the latter scenario.

141According to the Commission, absent the transaction, cannibalisation, that is to say, competition from the applicant selling the intervener’s flight content under its own brand via the MSS channel (see paragraph 36 above), would, in particular, be at the heart of the renegotiation of the cooperation, which would be initiated by the intervener in the counterfactual scenario used by the Commission (recitals 704 to 720 of the contested decision).

142In that regard, the Commission explains that, absent the transaction, the intervener would have an incentive to renegotiate in order to be compensated for the cannibalisation and that such compensation, in particular in the form of a lower meta cap or a higher booking fee, would result in the applicant selling fewer flights than would be the case in the counterfactual scenario put forward by the applicant (recital 719 of the contested decision).

143It follows that, if a quantitative assessment of the increment in the applicant’s post-merger share of the hotel OTA market were found to be necessary in order to examine the validity of the Commission’s theory of harm, the question arises as to the extent to which the counterfactual scenario put forward by the Commission and that put forward by the applicant lead to the increment in the applicant’s share of the hotel OTA market differing from the increment under the merger scenario.

144However, irrespective of the counterfactual scenario, the increment in the applicant’s share of the hotel OTA market in 2025, compared with the merger scenario, is in any event marginal (see paragraph 375 below). Therefore, it must be held that the difference, in quantitative terms, between the counterfactual scenarios put forward by the parties is negligible, in that each of those scenarios leads to a percentage of the same order of magnitude, with the result that the choice between one of the two scenarios has no bearing on the outcome of the present action.

145Accordingly, it is necessary to reject the first complaint of the first part and the first and second arguments of the second complaint of the first part of the second plea as being ineffective.

C. The first and second parts of the third plea: the legal standard and the theory of harm

146In Section 6.7.2.1 of the contested decision, the Commission relied on a theory of harm made up of five steps, which it summarised in paragraphs 18 to 26 of the defence as set out below.

147In the first place, the applicant was already dominant on the hotel OTA market.

148In the second place, the transaction would enable the applicant to acquire its own flight solution, in order to use its flight offering as leverage to acquire customers for its hotel OTA business.

149In the third place, the acquisition of the intervener would lead to increased network effects in favour of the applicant and higher barriers to entry and expansion for other hotel OTAs.

150In the fourth place, the increment in the share of the hotel OTA market which the applicant could achieve post-transaction would be a maximum of [confidential]%.

151Finally, in the fifth place, as a result of the transaction, hotels and consumers were likely to be harmed.

152More specifically, the Commission submits in Section 6.4 of the contested decision, entitled ‘Booking has a dominant position in the hotel OTA market in the EEA’, that it adduced evidence of the applicant’s dominant position and, in Section 6.7 of that decision entitled ‘Strengthening of Booking’s dominant position on the hotel OTA market in the EEA’, that it demonstrated that the transaction would lead to a strengthening of that position, making it more difficult to contest.

153In that context, it should be stated that, in terms of its form, the contested decision does not contain a separate analysis of a significant impediment to effective competition, even though certain recitals of that decision (recitals 734, 743 and 746) contain references to that concept.

154It is settled case-law that where it is clear from the grounds of a decision finding that a concentration is incompatible with the internal market, including those formally devoted to an analysis of the strengthening of a dominant position, that that transaction would give rise to a significant impediment to effective competition, that decision cannot be held to be vitiated by illegality solely because the Commission did not expressly analyse those two criteria (see, to that effect, judgments of 21 September 2005, EDP v Commission, T‑87/05, EU:T:2005:333, paragraphs 49 and 50, and of 14 December 2005, General Electric v Commission, T‑210/01, EU:T:2005:456, paragraphs 87 and 89).

155Therefore, in order for the Commission to be able to declare the transaction incompatible with the internal market, any strengthening of the applicant’s alleged dominant position on the hotel OTA market, even if it were established, must be capable of being regarded, on reading the contested decision, as also resulting in a significant impediment to effective competition on that market (see, to that effect, judgment of 22 October 2002, Schneider Electric v Commission, T‑310/01, EU:T:2002:254, paragraphs 321, 349, 380 and 402).

156In that context, it is for the EU Courts, after examining the substance of the complaints made against the Commission’s assessment of the relevant factors and in the light of the outcome of those complaints, to conduct an overall assessment of whether all the relevant factors and evidence on which the Commission has relied and which may be regarded as established, including those which have not been disputed, are sufficient to demonstrate the existence of a significant impediment to effective competition (see, to that effect, judgment of 13 July 2023, Commission v CK Telecoms UK Investments, C‑376/20 P, EU:C:2023:561, paragraph 262).

157Accordingly, it is necessary to examine the various arguments by which the applicant disputes the validity of the theory of harm adopted by the Commission, which will be grouped together into five complaints, concerning (i) the applicant’s allegedly dominant position on the hotel OTA market, (ii) the importance of a flight offering in general and the intervener’s flight solution in particular, (iii) the increase in the applicant’s share of the hotel OTA market, (iv) the alleged increase in barriers to entry and expansion in that market and (v) the harm which hotels and consumers would suffer post-transaction. Next, the Court will conduct an overall assessment in order to determine whether the Commission erred in finding that there was a significant impediment to effective competition post-transaction, while also examining the first part of the third plea by which the applicant challenges the legal standard applied by the Commission in that context.

158In that regard, it must be recalled that, although the prospective analysis specific to the preventive system of review of concentrations falls within the margin of discretion enjoyed by the Commission with regard to economic matters, the EU Courts must not refrain from reviewing the Commission’s interpretation of information of an economic nature. Not only must the EU Courts establish, in particular, whether the evidence relied on is factually accurate, reliable and consistent but also whether that evidence contains all the information which must be taken into account in order to assess a complex situation and whether it is capable of substantiating the conclusions drawn from it (see, to that effect, judgment of 13 July 2023, Commission v CK Telecoms UK Investments, C‑376/20 P, EU:C:2023:561, paragraphs 82, 84 and 125 and the case-law cited).

1. The first complaint: the applicant disputes that it has a dominant position on the hotel OTA market

159In recitals 363 and 364 of the contested decision, the Commission observed that the applicant held a dominant position on the hotel OTA market, since it was able to act independently of its competitors and customers.

160According to the Commission, the observation referred to in paragraph 159 above follows from (i) the applicant’s share of the hotel OTA market (see paragraph 19 above and recitals 340 to 362 of the contested decision), (ii) the fact that it receives an average effective commission [confidential]% higher than that received by rival OTAs (recitals 373 to 404 of the contested decision), (iii) the fact that it is an unavoidable trading partner for hotels and that it is in a position to impose certain general terms and conditions on them (recitals 405 to 426 of the contested decision), (iv) the fact that it is not subject to sufficient pressure from rival OTAs or sufficient out-of-market constraints (recitals 427 and 471 of the contested decision), (v) the fact that on each parameter determining the competitiveness of a hotel OTA, the applicant outperforms its competitors (recitals 491 to 557 of the contested decision), (vi) the fact that there are high barriers to entry and expansion to that market on account of significant network effects on that market as well as customer loyalty and inertia (recitals 558 to 565 of the contested decision) and (vii) the lack of likely entry or expansion which would effectively constrain the applicant (recitals 472 to 488 of the contested decision).

161It must be stated that, in the context of the second part of the third plea, which concerns the Commission’s theory of harm, the applicant does not call into question either the fact that it is an unavoidable trading partner for hotels, allowing it to impose certain general terms and conditions on them, or the fact that it outperforms its competitors on each parameter determining the competitiveness of a hotel OTA, or even the lack of likely entry or expansion which would effectively constrain it. By contrast, the applicant challenges four of the elements referred to in paragraph 160 above, namely its share of the hotel OTA market (paragraphs 157 to 164 of the application), the average effective commission it would receive (paragraphs 216 to 219 of the application), the fact that it would not be subject to sufficient out-of-market constraints (paragraphs 205 to 214 of the application), and its customers’ loyalty and inertia (paragraphs 230 to 233 of the application).

162Those factors will be examined in the order indicated.

(a)The applicant’s share of the hotel OTA market

163The applicant submits that the Commission overestimates its share of the hotel OTA market in 2022, by underestimating the size of that market in its market reconstruction exercise, as set out in recitals 295 to 325 of the contested decision.

164The Commission took as a basis, in particular, the applicant’s calculations based on the size of the hotel OTA market for 2022 as initially estimated by Euromonitor in August 2022 (‘Euromonitor’s initial market size’), whereas the applicant stated on a number of occasions that those calculations were unreliable. In the applicant’s view, that was confirmed by the fact that, in July 2023, Euromonitor had provided the Commission with a revised estimate for 2022, in order to take account of a methodological flaw and the rapid rebound in the travel market after the COVID-19 pandemic, which Euromonitor anticipated would occur only in 2023 (‘Euromonitor’s revised market size’).

165While the parties repeatedly raised that mistake during the administrative procedure, the Commission refused to correct it, with the result that the size of the hotel OTA market reconstructed by the Commission was [confidential] than Euromonitor’s revised market size.

166According to the applicant, that refusal leads to a significant overestimation of its market share, which increased from approximately [confidential]%, based on Euromonitor’s revised market size, to [60-70]%, as the Commission maintains in recital 216 of the contested decision.

167The Commission disputes the applicant’s arguments.

168In that regard, it must be stated that it is apparent from the file that on 11 March 2023, in response to the Commission’s request for information No 17, the applicant submitted to the Commission estimates of its share of the hotel OTA market and those of other hotel OTAs (‘the applicant’s March 2023 estimates’).

169Accordingly, as the applicant indicated, without being challenged by the Commission, its March 2023 estimates were prepared on the basis of three elements, all expressed in euro, more specifically, the total transaction value (‘TTV’), that is to say, (i) Euromonitor’s initial market size, (ii) the sales by the applicant and by the intervener as hotel OTAs and (iii) the estimates of the market shares of certain other hotel OTAs.

170In that context, the applicant first calculated the market share of the merged entity by dividing its sales and those of the intervener in 2022 by Euromonitor’s initial market size, giving a percentage of [confidential]%.

171Next, the applicant estimated the (percentage) market shares of certain competitors, [confidential].

172[confidential].

173When providing its March 2023 estimates, the applicant had expressed reservations as to the reliability of those percentages, stating that ‘Euromonitor has confirmed that its current 2022 forecast is conservative and that it expects to revise 2022 accommodation OTA market size estimates upwards (as it has recently done for 2011 to 2021) in the course of its normal review process in or around August 2023’.

174It is common ground that, in July 2023, Euromonitor provided the Commission with the revised market size, which was approximately one and a half times greater than the initial market size previously estimated. Therefore, based on Euromonitor’s revised market size, the applicant’s market share was only approximately [confidential]%.

175However, according to the Commission, the Euromonitor market sizes, whether initial or revised, are unreliable for a number of reasons, which the applicant did not dispute in its written pleadings.

176First, Euromonitor does not know the turnover achieved by the various market participants and its estimates are largely based on public sources. Second, Euromonitor’s data do not contain a specific classification for hotel OTAs, but rather refers to a broader ‘lodging’ category. Third, under Euromonitor’s market sizes, the applicant’s market share would exceed 100% in a number of Member States. Fourth, the origin of Euromonitor’s revised market size is uncertain and the Commission has reason to believe that that revised size was overestimated.

177Therefore, the Commission carried out a market reconstruction exercise (recitals 295 to 305 of the contested decision).

178More specifically, the Commission requested the figures relating to hotel OTA sales, expressed in TTV, not only from the applicant and the intervener, but also from other hotel OTAs, seven of which had submitted a response to the Commission, namely [confidential] (recitals 288, 289 and 296 to 302 of the contested decision).

179On the basis of those figures, the Commission reconstructed the size of the hotel OTA market, assuming that the nine entities which provided figures together covered, based on the applicant’s March 2023 estimates, [confidential]% of the market, that is to say, [confidential], while the ‘other’ hotel OTAs, including [confidential], which had not provided the data requested by the Commission, were deemed together cover the remaining [confidential]% (see paragraphs 171 and 172 above and recitals 290 and 303 of the contested decision).

180Lastly, the Commission used the data provided by the applicant and its competitors to calculate their shares of the relevant market based on the reconstructed market size as set out in paragraph 179 above (recital 305 of the contested decision).

181As such, the Commission claims to have followed its long-standing practice by relying on the market share estimates provided by the notifying party, in the present case the applicant, as the most reliable indicator of hotel OTA sales for which it did not have actual figures. In that regard, the Commission maintains, in essence, that the applicant’s March 2023 estimates were more reliable than Euromonitor’s initial or revised market sizes.

182However, as the applicant submits, it must be stated that the estimates referred to above are derived, to a large extent, from Euromonitor’s initial market size.

183Indeed, the vast majority of the [confidential]% (see paragraph 179 above), on the basis of which the Commission estimated the market size, is made up of the market share of the merged entity ([confidential]% (see paragraph 179 above)), which the applicant calculated based on Euromonitor’s initial market size.

184Yet, if the applicant had relied on Euromonitor’s revised market size, the merged entity’s market share would have been only approximately [confidential]%, rather than [confidential]%, whereas the nine entities which provided figures to the Commission would together have covered only [confidential]% of the market rather than [confidential]% (see paragraphs 171 and 179 above).

185It must therefore be observed, as the applicant did, that it is inconsistent for the Commission to reject Euromonitor’s market sizes as unreliable, while relying, de facto, on Euromonitor’s initial market size.

186That being said, it must also be observed, as the Commission did, that, in so far as the Commission indirectly relied on the data published by Euromonitor, the size of the relevant market was rather an overestimation, given that the Commission thereby equated the size of the broader ‘lodging’ market, as estimated by Euromonitor, with the hotel OTA market, which artificially inflated the size of the hotel OTA market (recital 323 of the contested decision) and, therefore, reduced the applicant’s share of that market.

187In addition, it is not apparent from the file that the Commission had more reliable data than the data submitted by the applicant, such that the Commission did not err in relying on that data in order to calculate the applicant’s share of the hotel OTA market.

188It should, in particular, be noted that Euromonitor’s revised market size implies that the collective market share of unidentified hotel OTAs (the ‘Others’ category), expressed in euro (TTV), would be around [confidential]%, which appears to be abnormally high (recital 292 of the contested decision). While it is not exceptional that market share estimates lead to a failure to identify a certain part of the market because of a lack of data concerning all market players, a percentage of that size raises doubts as to the reliability of the estimate of the relevant market size, as the Commission correctly observed in recital 292 of the contested decision.

189When questioned on that subject at the hearing, the applicant was unable to clarify which OTAs fall within the ‘Others’ category, although, as the main provider of hotel content to other OTAs, it was best placed to provide such data. Indeed, those OTAs, if they exist, would probably be as small as, or even smaller than, the smallest OTAs which provided figures to the Commission (see paragraph 179 above), with the result that they are unlikely to have their own hotel content (recitals 254 and 264 to 268 of the contested decision). It is therefore likely that a considerable proportion of them obtain such content from the applicant, as it is the main provider of hotel content to other OTAs (recital 258 of the contested decision), which ought to have enabled the applicant to identify and specify the sales volumes of a significant proportion of the OTAs falling within the ‘Others’ category. Nevertheless, it must be observed that the applicant failed to do so, either during the administrative procedure (paragraph 292 of the contested decision) or during the proceedings before the General Court. Therefore, the Commission did not err in finding that it was more realistic to assume that the ‘Others’ category covered approximately [confidential]% of the market, as the Commission concluded, rather than [confidential]%, as claimed by the applicant.

190Lastly, it should be recalled that, in the context of the market investigation, a number of industry experts and other OTAs stated that they also were of the view that the applicant’s share of the hotel OTA market was at least [60-70]%, which tends to support the Commission’s estimates (recitals 353 to 360 of the contested decision).

191It follows that the applicant has not succeeded in calling into question the Commission’s estimates of its share of the hotel OTA market.

(b)The average effective commission received by the applicant

192In recitals 372 to 404 of the contested decision, the Commission observed that the average effective commission charged by the applicant to hotels was [confidential]% and was therefore [confidential]% higher than that received by rival OTAs, that is to say [confidential]%, which the applicant disputes, raising, in essence, three arguments.

193In the first place, the average rate of [confidential]% is based on a gross calculation consisting of dividing a revenue figure by a TTV figure provided by five hotel OTAs. It is not apparent from the contested decision that the Commission made the appropriate methodological adjustments to account for various differences between the undertakings’ revenues. In that regard, the applicant observes that [confidential] maintained that it was difficult to compare the commissions paid to it and its competitors, that [confidential] indicated that the exact commission could change significantly depending on the country or the hotel brand and that [confidential] explained that there was no one-size-fits-all commission rate.

194In the second place, the applicant refers to a statement made by the HOTREC hotels association, according to which the commission rates charged by hotel OTAs to its members were within a range of [confidential]%; the applicant and [confidential] were both within that range. That statement is confirmed by explanations from a number of hotels, which indicated that most hotel OTAs charged [confidential]% and upwards.

195In the third place, contrary to the Commission’s assertions, an [confidential] survey of accommodation partners did not support the claim that Expedia charges lower commission rates than the applicant.

196The Commission disputes that line of argument.

197In the first place, the Commission stated that it had relied on comparable actual TTV and revenue data provided by the applicant and its main competitors, which together account for approximately 70% to 80% of the hotel OTA market.

198In particular, the Commission asked the applicant’s main hotel OTA competitors which had provided data to confirm that they had followed the same methodology as the applicant. In that context, four rival OTAs provided data on a comparable basis, whereas one rival OTA was not able to do so, with the result that the latter’s data were excluded from the Commission’s calculation (recitals 374 to 380 of the contested decision).

199It should be observed that the applicant has provided no evidence from which it could be concluded that the Commission was required to make other adjustments to its calculations or to verify its results by reference to other sources.

200In the second place, the Commission submits, first, that there is a difference between standard commissions, which are published by OTAs and which are broadly similar on the market, and the commissions actually charged, which can vary on a case-by-case basis depending on the negotiating position of the hotel (and the OTA) and the hotel’s willingness to pay a premium in order to be ranked higher on the website of the OTA concerned. However, in the statement referred to by the applicant, HOTREC did not specify the source of its data or whether it was referring to standard commissions or commissions actually charged. In fact, it simply provided a range for the commissions applied by accommodation OTAs and indicated that the commissions of the main OTAs fell within that range.

201Second, the Commission indicates that the hotel statements quoted by the applicant are general in nature and need to be read in their proper context, since they were made in response to a question about market definition and, more specifically about the interchangeability between the services provided by accommodation OTAs and those provided by brick-and-mortar travel agencies. In that context, the hotels were merely concerned with indicating generally the level of the commission charged by OTAs to enable the Commission to assess whether accommodation OTAs and brick-and-mortar agencies were in the same product market.

202It must be noted that the applicant has not disputed those findings.

203In the third place, the Commission maintains that the [confidential] survey does not establish a comparison between commission levels charged by the applicant and by Expedia respectively. In addition, according to the Commission, although the proportion of respondents to that survey who agreed with the statement that ‘this platform is expensive for us to use’ is slightly higher in terms of the percentage of respondents for Expedia than for the applicant, the actual numbers tell a different story, in so far as the number of those persons who regarded the applicant as an expensive platform was more than 10 times higher than the number of those who regarded Expedia as expensive.

204It must be observed that the applicant has not disputed those findings.

205Furthermore, it should be noted, as the Commission correctly submits, that the statements of the various hotels and HOTREC and the [confidential] survey referred to above have low evidentiary value in that they contain only general statements which do not make it possible to determine precisely the rate of the applicant’s commissions and that of its main competitors. That finding is not called into question by the applicant’s statement that the actual data were collected by those sources from the undertakings paying those commissions, which, moreover, the applicant has not substantiated.

206It must therefore be held that the applicant has not succeeded in calling into question the Commission’s finding that the applicant charges significantly higher commissions than other rival hotel OTAs.

(c)The lack of sufficient pressure from rival OTAs, on the one hand, and of sufficient out-of-market constraints, on the other

207The applicant disputes the Commission’s definition of the product market in recitals 88 to 170 of the contested decision, submitting that (i) that definition ought to have included accommodation OTAs and (ii) the Commission disregarded competition from the hotels’ direct channel. According to the applicant, those errors invalidate, from the outset, the finding that it holds a dominant position on the unduly narrow market for hotel OTAs.

208In the first place, according to the applicant, the correct product market definition ought to have included accommodation OTAs, which are direct competitors of hotel OTAs, since the accommodation they provide is a substitute for that offered by hotel OTAs and their websites also include hotels, which demonstrates supply-side substitutability.

209In the second place, the applicant is of the view that the Commission disregarded the competition that the hotels’ direct channel and hotel OTAs engage in for the sale of hotel rooms, which is fierce and increasing because of the growing importance of MSSs such as Google Hotels, which enable consumers to compare offers from a number of different hotels before booking directly with one of them. Accordingly, direct sales by hotels are increasing, as stated in recital 432 of the contested decision.

210In that regard, the applicant submits that the Commission cannot maintain that the transaction would be likely to move sales from the hotels’ direct channel to it, while also stating that hotel OTAs and direct channels do not compete. Therefore, according to the applicant, in order to argue that its prices are higher than those of the ‘cheapest alternative’, the Commission refers to evidence which shows that [confidential]. Since the hotels’ direct channel does not fall within the relevant market as defined in the contested decision, the Commission’s approach is contradictory.

211Lastly, according to the applicant, the contested decision provides no assessment of how hotels might respond to the applicant potentially gaining a small volume of additional hotel OTA customers. Therefore, hotels could redirect volume to their websites if an increased price through one of their OTA channels made that more profitable or attractive. That is confirmed by the fact that hotels already provide lower prices through the direct channel than through OTAs.

212The Commission disputes that line of argument.

213In that regard, in the first place, it must be stated that the applicant’s arguments are not capable of calling into question the detailed explanation in recitals 98 to 116 of the contested decision of the reasons why the Commission takes the view that hotel OTAs and accommodation OTAs belong to different markets.

214First, in the contested decision, the Commission found that some OTAs which offered hotel rooms also offered private accommodation and vice versa, and those which did so often offered different services for each of those markets with differing degrees of success in each of them (recitals 99 to 105 of the contested decision).

215Second, the Commission stated that there were significant barriers to entry to the hotel OTA market as well as differences between hotels and other types of accommodation, as regards the applicable regulatory requirements, price levels and customer profiles (recitals 106 to 108 of the contested decision).

216Third, the Commission found a difference in price, in that hotels were generally more expensive than private accommodation, and a difference in customer profile, in that users of private accommodation tended to be younger than those of hotel accommodation (recitals 109 to 114 of the contested decision).

217In so far as the applicant has failed to put forward in its written pleadings any facts or circumstances capable of calling into question the factors mentioned above, its argument that accommodation OTAs ought to have been included in the relevant market as defined in the contested decision must be rejected.

218In the second place, as regards the alleged competition from the hotels’ direct channel, in particular due to the growing importance of MSSs, it should be noted that the Commission explains that it does not deny that hotels exert competitive pressure on hotel OTAs, but that it found that that pressure was not sufficient to constrain the applicant’s conduct (recital 429 of the contested decision).

219In that regard, the Commission recalls three of those findings in its defence.

220First, in recitals 1086 and 1087 of the contested decision, the Commission refers to an average [confidential]% hotel OTA commission rate and observes that the applicant’s rate, that is to say, [confidential]% is [confidential]% higher.

221As stated in paragraphs 192 to 206 above, the applicant has not succeeded in calling into question the Commission’s finding that it charged significantly higher commissions than other rival hotel OTAs.

222Second, in recitals 436 to 451 of the contested decision, the Commission states that the applicant has, over time, become relatively more expensive compared with the cheapest alternative offered to end customers.

223In that regard, the applicant correctly observes that that cheapest alternative is, according to the Commission, the hotels’ direct channel. However, contrary to the applicant’s assertions, the finding that it has become more expensive over time compared with that channel does not prevent that channel from being excluded from the relevant product market, in so far as that finding serves to support the fact that the applicant did not feel constrained by the hotels’ direct channel, despite that channel being consistently cheaper.

224Indeed, in recitals 436 and 437 of the contested decision, the Commission states that, if the applicant had felt sufficient pressure from the hotels’ direct channel due to the fact that hotels were actually able to redirect volume to their websites, it would have been compelled to display lower prices in order to capture more end customers and increase the attractiveness of its platform vis-à-vis hotels’ websites. That statement has not been disputed by the applicant in its written pleadings.

225Third, in recitals 432 and 467 of the contested decision, the Commission states that, despite the increase in direct sales by hotels and the increased importance of Google Hotels as an MSS, the applicant increased its share of bookings and, consequently, its share of the relevant market, which the applicant did not dispute in its written pleadings.

226In that regard, it should be recalled that, in recital 432 of the contested decision, the Commission found that it was true that direct sales by hotels had increased, but the size of the hotel OTA market had also increased, as had the applicant’s own sales. That finding is not disputed by the applicant in its written pleadings.

227In the light of the foregoing, it is necessary to reject the applicant’s argument based on competition from the hotels’ direct channel and the growing importance of MSSs.

(d)Loyalty and inertia on the part of the applicant’s customers

228The applicant submits that the Commission erred in stating on numerous occasions that the high levels of customer loyalty were due to customer inertia (see recital 1138 of the contested decision). According to the applicant, its customers are not inert; rather, they are intensely price sensitive and aware of a range of commercial conditions such as cancellation policies, the ease of use of the platform and the discounts offered, which explains why they multi-home. Accordingly, the fact that users return to the applicant’s platform after an initial purchase is indicative not of inertia but of satisfaction with the service.

229In that regard, the applicant disputes the Commission’s allegation that about half of its customers are inert, on the ground that, according to an internal survey carried out by the applicant, when searching for accommodation, they used a single source. According to the applicant, multi-homing cannot be measured meaningfully in that way, whereas the same internal survey shows that [confidential]% of its customers had also booked on other websites over the previous 12 months.

230The Commission disputes that line of argument.

231As a preliminary point, it should be recalled, first, that, in the field of digital services in general and online platforms in particular, the concept of ‘multi-homing’ describes the situation in which users use several competing digital services in parallel (judgment of 17 July 2024, Bytedance v Commission, T‑1077/23, under appeal, EU:T:2024:478, paragraph 171) and, second, that the concept of customer inertia refers to the fact that certain customers remain with a certain service provider despite more competitive offers being available (see, to that effect, judgment of 20 December 2023, Naturstrom v Commission, T‑60/21, not published, EU:T:2023:839, paragraph 241).

232In that context, the applicant disputes, first, that approximately half of its customers do not multi-home and, second, that that occurs due to inertia.

233As regards, first, the proportion of customers who use only the applicant’s platform, it must be observed that the Commission relied on a series of customer surveys, [confidential] conducted by the applicant in 2020 and 2021, each of which involved 500 respondents in Germany, Spain, France, Italy, the United Kingdom and the United States, in which approximately [confidential]% of respondents indicated that they went directly to the platform via which they usually booked, without comparing prices on other websites (recitals 234, 235 and 912 of the contested decision).

234In that regard, the applicant submits, without being challenged by the Commission, that that series of surveys shows that [confidential]% of its customers had also booked travel services on other platforms during the previous 12 months. However, that fact does not make it possible to determine the extent to which they make reservations on a particular platform without first consulting other platforms and does not therefore prove the degree of multi-homing among its customers.

235According to the Commission, it is also apparent from that series of surveys, and from a customer survey by [confidential], that approximately [confidential] of bookings with hotel OTAs in general and with the applicant in particular are made without the customer having consulted a source other than the platform on which the customer made the booking (recitals 237 and 912 of the contested decision).

236Lastly, it should be noted that Figure 5 of the contested decision indicates [confidential] (recital 238 of the contested decision).

237It follows that the applicant has not succeeded in calling into question the plausibility of the Commission’s conclusion that a significant proportion of hotel OTA customers, which could reach [confidential]%, tends not to compare the prices displayed on the various platforms and therefore tends to reserve directly with the platform of its choice.

238As regards, second, the alleged inertia of some of the applicant’s customers, the Commission submits that it is apparent from recitals 1127 to 1133 of the contested decision that the loyalty of the applicant’s customers is not simply due to the quality of its platform.

239In that regard, it must be stated, first, that the customer experience data set out in Figure 128 (recital 1126 of the contested decision) suggest that [confidential], second, that Figure 129 (recital 1130 of the contested decision) indicates that [confidential] and, third, that the same conclusions emerge from the applicant’s internal documents cited in recitals 1127 to 1129 of the contested decision.

240Nevertheless, it should be noted that the arguments raised by the applicant in that context are limited to general assertions that its customers are price-sensitive and aware of commercial conditions, which, however, do not demonstrate that those customers’ loyalty is due to the price and quality of its services. It must, therefore, be concluded that the applicant has not succeeded in calling into question the plausibility of the Commission’s conclusion that customer loyalty was mainly due to those customers’ inertia.

(e)Conclusion

241In the light of all of the factors set out above concerning the applicant’s position on the hotel OTA market, it must be concluded that the applicant has not succeeded in calling into question the Commission’s conclusion that it had a dominant position on that market.

242It is therefore appropriate to continue the analysis on the basis of the findings that (i) the applicant holds a significant market share reflecting its dominant market position on the hotel OTA market, (ii) it charges higher commissions than other hotel OTAs, (iii) it is an unavoidable trading partner for hotels, (iv) it outperforms its competitors on all the parameters determining the competitiveness of a hotel OTA and it is not subject to sufficient pressure from rival OTAs or sufficient out-of-market constraints, (v) on the hotel OTA market there are strong barriers to entry and expansion resulting from significant network effects, (vi) there is customer loyalty and inertia and, finally, (vii) there is a lack of likely entry or expansion which would constrain the applicant effectively.

2. The second complaint: the applicant disputes the importance of a flight offering in general and of the intervener’s flight solution in particular for acquiring hotel customers

243In the second part of the third plea (paragraphs 138 to 146 and 199 to 204 of the application), the applicant contests the Commission’s statement in particular in Section 6.7.2.2 of the contested decision, that, via the transaction, it would purchase an important hotel customer acquisition channel, which would enable it to create a travel ecosystem leveraging its brand strength and customer inertia, which would be difficult for its competitors to replicate.

244The arguments which the applicant puts forward, in that regard, may be divided into two groups, namely (i) those relating to the general importance of a flight offering in order to acquire hotel customers and (ii) those concerning the more particular importance of the intervener’s flight solution for acquiring those customers.

(a)The general importance of a flight offering for hotel customer acquisition

245The applicant submits that it is not necessary to have a flight OTA offering in order to compete effectively on the hotel OTA market.

246According to the applicant, supported by the intervener, the impact of a flight offering on sales of hotel rooms is negligible.

247First, most flight customers who also wish to book accommodation do not do so at the same time, but book the accommodation separately, either in advance of the flight or, more typically, much later in time.

248Second, at least [confidential]% of flights are booked directly with airlines and only [confidential]% of traffic on the 10 main hotel OTA platforms, including the applicant’s platform, concerns visitors who have been rerouted from flight OTA platforms or airlines.

249Third, rival hotel OTAs which also offer flights confirmed that it was difficult to cross-sell accommodation from flights, which is reflected in the fact that the attach rates reported by the major OTAs in the market investigation were all below 15% and their shares of the hotel OTA market were minimal.

250The Commission disputes the applicant’s arguments.

251As a preliminary point, it should be recalled that the transaction forms part of the applicant’s connected trip strategy, that is to say, the sale of a range of travel services in parallel.

252In that context, the Commission explained, without being challenged by the applicant, that the applicant experienced a slowdown in the growth of its revenue from its traditional hotel business and that it therefore attempted to find new avenues for growth and that [confidential] (recitals 571 to 576 and 584 of the contested decision).

253More specifically, the Commission indicated, without being challenged by the applicant, that the connected trip strategy would enable the applicant to attract new customers and increase the loyalty of existing customers (recitals 577 to 605 and 933 to 938 of the contested decision), including high-value customers who, according to the contested decision, [confidential] (recitals 520 to 524 of the contested decision).

254In addition, the Commission pointed out, without being challenged by the applicant, that a flight offering was of critical importance for the applicant’s connected trip strategy.

255In that context, the applicant’s argument that it is not necessary to have a flight offering to be able to compete effectively on the hotel OTA market must be rejected (see paragraph 245 above).

256First, it must be stated that, contrary to the applicant’s assertion, the Commission does not regard the mere ability to offer various travel services or to develop a travel ecosystem as, in itself, a potential source of competitive harm, or that a hotel OTA must have a flight offering in order to be competitive in the hotel sector. On the contrary, the Commission merely found that it was one of the few customer acquisition channels still available for rival OTAs (recitals 988 to 992 of the contested decision).

257Second, even if a complementary business, in the present case a flight OTA business, is not indispensable to efficiently conduct a main business, in the present case a hotel OTA business, the fact remains that the strengthening of the complementary business may, in certain circumstances, adversely affect competition on the market for the main business, in particular where the undertaking concerned has a dominant position on that market.

258In that context, the Commission set out in recitals 747 to 793 of the contested decision the importance of a flight offering as an acquisition channel for hotel customers, both in qualitative and quantitative terms.

259In qualitative terms, the Commission explained that flights were the ‘entry point’ for the trip, since they were generally booked before the hotel and could be used by a hotel OTA in general and by the applicant in particular, in order to generate additional customer traffic and obtain access to customer data. An OTA could, in particular, use those data to target potential hotel customers with specific offerings, before, at the same time or after the sale of the flight, by means of advertisements by email or on the OTA website when the customer navigated to the OTA website, or notifications on that customer’s smartphone (recitals 753, 767 to 793 of the contested decision and paragraphs 28 to 33, 38 and 44 above).

260In that context, it is necessary to reject the applicant’s argument that many flight customers who also wish to book a hotel do not do so at the same time, but do so separately, either in advance of the flight or, more typically, much later in time (see paragraph 247 above). That does not alter the fact that the OTA concerned remains in possession of the customer’s personal data which are directly related to his or her planned trip, in particular the dates and place where that customer will stay, which allows the OTA to target that customer with specific offers for a subsequent hotel booking (recitals 790 to 792 of the contested decision).

261In quantitative terms, the Commission states that flights often constitute (in [confidential]% of cases) the beginning of the booking journey and the service with the highest opportunity to be booked on the same website as accommodation (recitals 754 to 758 of the contested decision). Accordingly, compared with other services forming part of the connected trip strategy, such as car rental, taxi transport or attractions, flights are said to constitute the most important ‘vertical’ for attracting new customers (recital 752 of the contested decision and paragraphs 27 and 42 above).

262In that regard, the applicant maintains that the cross-selling potential of a flight OTA platform to a hotel OTA platform is nevertheless limited.

263First, it should be observed that, whilst it is true that flights often (in [confidential]% of cases) mark the beginning of the booking journey, the applicant has stated, without being contradicted by the Commission, that approximately [confidential]% of flights are purchased directly from airlines, such that only [confidential]% to [confidential]% of hotel OTA customers also require the services of a flight OTA (recital 748 of the contested decision). It follows that the percentage ([confidential]%) of cases in which the flight constitutes the beginning of the booking journey overestimates the overlap between flight OTA and hotel OTA customer groups and, therefore, the cross-selling potential of the former to the latter.

264Second, the applicant maintains, without being challenged by the Commission, that other OTAs stated in the context of the market investigation that their ability to convert that potential into hotel bookings was limited, as demonstrated by attach rates below 15%, which reflect the share of customers who, after purchasing a particular travel service (for example, a flight ticket), purchase (or ‘attach’) an additional service (for example, a hotel room), as stated in footnote 794 to the contested decision (see paragraph 249 above).

265As regards the applicant itself, it should be observed that it is apparent from the contested decision that, while it is true that its attach rate for new customers is barely higher than the percentage referred to in paragraph 264 above, that rate is much higher for existing customers, approximately [20-30]% of whom are regarded by the applicant as ‘high value’ (see paragraph 27 above), reflecting loyalty or even inertia on the part of a considerable proportion of its customers (recitals 561, 764, 1046, 1047 and 1125 of the contested decision). Accordingly, the attach rate applied by the applicant in the TVM, which the Commission reproduced in its calculations in Section 6.7.2.4 of the contested decision, is [confidential]%.

266Third, it is necessary to reject the applicant’s argument that the low percentage of traffic of the 10 main hotel OTAs, including the applicant, from flights confirms the limited importance of a flight offering for a hotel OTA. As the Commission correctly stated, traffic data relating to users’ navigation routes between different platforms do not demonstrate the potential cross-selling of hotel rooms from flights. In particular, such traffic is not representative of the TTV (recital 281 of the contested decision), that is to say, the amount actually spent by the applicant’s customers on overnight hotel stays, and, therefore, is also not representative of the commissions actually received by the applicant.

267In any event, it must be observed that the figures submitted by the applicant in order to demonstrate the negligible impact of a flight offering on the sale of hotel rooms are not sufficient to call into question the importance, for the applicant, of being able to offer flights on its platform.

268Indeed, it is apparent from the file, in particular from the applicant’s internal documents, that, from a strategic point of view, it is essential for the applicant to be able to offer flights on its platform (recitals 577 to 584 of the contested decision) as an instrument for acquiring or keeping hotel customers in its travel ecosystem.

269It is therefore necessary to reject the arguments of the applicant and the intervener by which they dispute the importance, in abstract terms, of a flight offering as an important channel for acquiring hotel customers.

(b)The unique attributes of the intervener and the travel ecosystem which the applicant could create by acquiring it

270In the first place, the applicant and the intervener dispute that the intervener is a ‘rare asset’, that is to say, a flight OTA with unique attributes compared with its competitors, since (i) on the flight OTA market, which the contested decision itself describes as highly competitive, the intervener has only a limited presence with a market share of [10-20]% via its own flight platforms and [5-10]% via its cooperation with the applicant and (ii) the market investigation confirmed that the intervener is only one of many OTAs with similar capabilities.

271The applicant adds that, if the intervener really were a ‘rare asset’, that would factor directly into an increment in its hotel OTA market share, which the applicant could achieve post-transaction.

272Similarly, the Commission’s market investigation confirmed that, on average, the intervener ranked only fourth out of the seven flight OTAs highlighted by the Commission. On price, the intervener ranked second, after Expedia, followed closely by Kiwi, eDreams and Lastminute, while, in terms of brand recognition and reputation, the intervener ranked fourth, after Expedia, Lastminute and eDreams.

273In the second place, the applicant disputes that the transaction would enable it to create a travel ecosystem which would be difficult for other OTAs to replicate. The Commission has failed to demonstrate that rival hotel OTAs are unable to develop their own ‘ecosystems’, in particular by accessing the numerous alternative sources of flights. Indeed, all of the applicant’s main competitors already offer travel ecosystems including hotels, flights and cars, and those which do not currently do so can easily develop that by partnering with other flight OTAs.

274In the third place, the applicant is of the view that the Commission’s attempt to differentiate between its offers and those of rival hotel OTAs by referring to its brand strength is unfounded, since Expedia, Lastminute and eDreams are all well-recognised brands and Expedia and Airbnb have higher brand awareness than it.

275In the fourth place, the applicant submits that the ‘qualitative’ factors put forward by the Commission in support of its theory of harm, such as additional customer traffic, additional access to customer data and increased customer loyalty, give rise to double-counting and confuse cause and effect. Indeed, those factors are nothing more than the cause of the growth in cross-sales of hotel rooms from flights and, therefore, of the increase in the applicant’s hotel OTA market share, which is, in any event, minimal.

276The Commission disputes the applicant’s arguments.

277As a preliminary point, it should be recalled that the Commission stated that it was not in the applicant’s strategic interest to depend on third parties in order to be able to offer flights. It is apparent from the applicant’s internal documentation that, after finding that the development of its own flight solution would take longer than expected and might not even be of the same quality as that of the intervener, it decided to pursue the transaction (recitals 677 to 679 of the contested decision), since the acquisition of the intervener’s flight solution was the applicant’s preferred option (recitals 721 to 727 of the contested decision). Accordingly, from a strategic point of view, the transaction would enable the applicant to optimise its flight business in the fastest and most efficient way.

278In qualitative terms, the Commission describes, in recitals 828 to 839 and 904 to 916 of the contested decision, how the transaction would enable the applicant to optimise its flight business and accelerate its growth.

279In that context, the Commission explains that it regards the intervener as a best-in-class flight OTA (recitals 828 to 839 of the contested decision). The feedback from the Commission’s market investigation and certain internal documentation of the applicant and the intervener demonstrate, in particular, that the intervener is generally seen as a strong player in the flight OTA market (recitals 828, 829, 831 and 837 of the contested decision).

280Next, the Commission sets out certain synergies which the applicant could achieve by incorporating the intervener’s flight solution into its platform.

281First, the Commission states that, for a flight OTA, price is the most important parameter of competition, in particular on MSS channels. Nevertheless, [confidential] (recitals 904 to 911 of the contested decision).

282Second, the Commission states that a considerable proportion of the applicant’s customers search exclusively on its platform and purchase without consulting other platforms beforehand, which also strengthens its ability to sell flights (recitals 912 to 916 of the contested decision).

283From that, the Commission infers that the transaction would enable the applicant to create a travel ecosystem taking advantage of its brand strength and customer inertia, which, moreover, is difficult for its competitors to replicate (recitals 906, 939 and 940 of the contested decision).

284In quantitative terms, the Commission is of the view that the intervener was already on a growth path prior to the transaction (recitals 801 to 827 of the contested decision) and that that growth would accelerate post-transaction (recitals 840 to 903 of the contested decision).

285First, the Commission observes that the intervener was able to increase its market share over the period from 2019 to 2022, with the result that, in 2022, it was, together with the applicant, the second largest player on the flight OTA market with a combined market share of [confidential]%, the other main players being eDreams Odigeo [20-30]%; Trip.com [10-20]%; Lastminute [5-10]%; Kiwi [5-10]%; Travelgenio [5-10]% and Expedia [0-5]% (recitals 824 to 826 of the contested decision).

286Second, the Commission is of the view, based on estimates of the size of the flight OTA market in the EEA for the period from 2023 to 2026 and the TTV projections provided by the applicant and the intervener, that the combined market share of the two parties to the transaction would increase to [confidential]% in 2026 (recitals 844 to 903 of the contested decision).

287In the first place, as regards the arguments by which the applicant disputes the intervener’s uniqueness (see paragraphs 271 and 272 above), it should be noted that the Commission bases the classification of the intervener as a leading flight OTA on a number of statements made (i) by a number of other flight OTAs in the market investigation and (ii) by the applicant and the intervener themselves in various internal documents (recitals 829 and 838 of the contested decision).

288Some competitors highlighted, in particular, the intervener’s ‘strength’, its ‘leadership position’ and ‘wide network of direct relationships with airlines’ and the fact that it ‘scores high’ on various parameters, especially price, but also the breadth of its offering and its technical capabilities (recitals 829 and 830 of the contested decision).

289Furthermore, in the internal documentation mentioned above, the intervener describes itself as ‘#1 flight platform’, while also referring to its ‘pricing capabilities’ providing it with ‘#1 position’ in ‘Meta traffic’, its ‘best content’ and its ‘winning business model’, whereas the applicant describes the intervener as being ‘best-in-class at combining flights and ancillary content from multiple suppliers’, which is far superior to the applicant’s own flight platform [confidential] (recitals 831 to 839 of the contested decision).

290As the Commission correctly points out, the applicant has not disputed that evidence in its written pleadings.

291In addition, in response to the applicant’s argument based on the results of the market investigation carried out (see paragraph 272 above), the Commission stated in the defence, without being challenged by the applicant, that the applicant had omitted the replies of four parties to which questions had been put and which had initially requested confidential treatment and that, taking those responses into account, the intervener was ranked first as regards price, it ranked second as regards technical capabilities as well as back-end and third as regards the breadth of the offering.

292Lastly, in response to the applicant’s argument based on the intervener’s share of the flight OTA market, demonstrating that the intervener is not the main operator on that market (see paragraphs 270 and 271 above), the Commission stated at the hearing that it had not taken a static assessment, namely the intervener’s current market share, as its basis, but rather a prospective analysis. According to the Commission, the transaction would enable the applicant to accelerate the growth of its flight OTA business in order to become the main operator in that field by 2026 (recitals 841 to 868 of the contested decision), which the applicant did not dispute in its written pleadings.

293In the second place, as regards the applicant’s argument that, in essence, its travel ecosystem is no different from those which competitors already have or could create (see paragraph 273 above), it should be noted that the objective of creating such an ecosystem, which would be difficult for competitors to replicate, is referred to in a number of the applicant’s internal documents and emails, [confidential] (recitals 906, 940 and 941 of the contested decision).

294In addition, at the hearing, the Commission stated that it was the very strong hotel component which made the applicant’s travel ecosystem difficult for other OTAs to replicate. By contrast, since the travel ecosystems of the competitors referred to by the applicant in that regard were mainly flight OTAs and only hotel OTAs as a secondary activity, their flight component was relatively significant and their hotel component relatively weak. While the Commission established, without being challenged by the applicant, that the applicant would be able to expand its position on the flight OTA market to the same level, or even beyond, that of its competitors (see paragraphs 284 to 292 above), it was unlikely that those competitors could do so on the hotel OTA market given the applicant’s dominant position on that market.

295Therefore, it is necessary to reject the argument by which the applicant disputes that its travel ecosystem would be difficult for its competitors to replicate.

296In the third place, as regards the applicant’s argument that competing hotel OTA brands are just as strong as its own (see paragraph 274 above), it should be noted that the Commission referred only to the strength of the applicant’s brand in order to highlight a synergy which the applicant could achieve by integrating the intervener’s flight solution into its platform. Accordingly, the strength of the applicant’s trade mark enables it to increase sales of airline tickets, [confidential] (see paragraph 281 above). The potential strength of the other brands is not capable of calling that finding into question.

297In the fourth place, as regards the applicant’s argument that the importance for it of acquiring the intervener’s flight solution ought to be reflected in the growth in the number of hotel customers and, therefore, in the increase in its share of the hotel OTA market (see paragraph 275 above), it should be recalled that it is common ground that the transaction would enable the applicant to optimise its flight business and accelerate its growth.

298Indeed, even if, in the counterfactual scenario, at least in the short term, the applicant were also to use the intervener’s flight solution based on a commercial affiliated agreement (see paragraphs 120 to 122 above), it should be recalled that the Phase 2 Agreement contains a series of restrictions which limit the applicant’s ability to exploit the full potential of its flight offering, but which would no longer exist if the transaction took place. Under that agreement, [confidential] (see recitals 681 to 684 of the contested decision).

299Accordingly, it must be held that the transaction would enable the applicant to optimise and increase its flight business in the fastest and most efficient way by acquiring its preferred flight solution, and to promote its hotel business by stimulating cross-sales of hotel rooms from flights, while increasing the loyalty of its hotel customers, including high-value customers.

300Therefore, it is necessary to reject the arguments of the applicant and the intervener by which they dispute the unique attributes of the intervener and the travel ecosystem which the applicant could create by acquiring the intervener.

3. The third complaint: the applicant disputes the calculations relating to the increment in its share of the hotel OTA market post-transaction

301In the second part of the third plea (paragraphs 147 to 155 and 165 to 177 of the application), the applicant notes a number of errors allegedly made by the Commission in its calculations of the increment in the applicant’s share of the hotel OTA market post-transaction. More specifically, whilst that increment could, according to the Commission, be up to [0-5]% (see paragraph 52 above), the applicant’s view is that it is limited to a few tenths of a per cent.

302In that context, the applicant refers to a number of errors allegedly made by the Commission in applying the modified TVM methodology and the Figure 123 methodology. More specifically, it disputes, first, the fact that the gains resulting from combined sales of hotel rooms and non-flight travel services were taken into account in the calculations, second, the calculation of the number of additional overnight stays it would acquire post-transaction, third, the conversion of that number into euro and the division of the resulting amount by the size of the hotel OTA market and, fourth, the use of the ‘zero-flights’ counterfactual scenario as a starting point for the calculations.

(a)The first alleged error: gains from combined sales of accommodation and non-flight travel services were taken into account

303The applicant submits that the Commission could not take into account instances in which it sold hotel rooms alongside another travel service, such as a taxi transport service or car rental, since those services are not offered by the intervener and the transaction therefore did not concern them. For years, the applicant has made such combined sales and it will continue to do so independently of whether or not it offers flights.

304The Commission disputes the applicant’s arguments.

305In the first place, the Commission submits that, since the development of a flight offering is the cornerstone of the connected trip strategy, optimising that strategy is the main object of the transaction. Given that, without such an offering, there would be no connected trip, hotel bookings and other travel services would not be increased by the concomitant ‘Value Leadership’ effects and ‘Halo Effects’. Accordingly, the addition of flights would enable those effects to influence all hotel bookings of the applicant resulting from the connected trip strategy, by increasing customer loyalty and the number of those bookings.

306The Commission concludes that flights are, directly or indirectly, related to all increases in hotel accommodation sales by the applicant resulting from its connected trip strategy, even combined sales of hotel accommodation and non-flight travel services.

307In the second place, the Commission maintains that flights have a greater effect on hotel bookings than other travel services, such as taxi or car rental services. Indeed, such travel services have only limited potential to attract customers to a hotel OTA platform, unlike flights, which represent the first product in the travel booking process for [confidential]% of travellers, are the travel service most likely to be booked online for the same trip as accommodation and have the highest opportunity to be booked on the same website as accommodation.

308In the third place, the Commission maintains that the applicant has provided no evidence to prove that the sudden increases in sales of hotel rooms in 2023 and 2025 were the product of the gains resulting from sales of hotel rooms combined with non-flight travel services, considering that sales by the applicant of hotel rooms combined with taxi or car rental services having occurred well before the transaction.

309In that regard, in the first place, it should be recalled that the connected trip, that is to say the combined sale of a number of travel services, is intended, inter alia, to increase the number of hotel bookings on the applicant’s platform. It must be held that the Commission’s line of argument is based on the incorrect premiss that, absent the transaction, the applicant would not offer any connected trips.

310First, as has been established in paragraphs 119 to 121 above, it is common ground between the parties that, whatever the counterfactual scenario, the applicant would also have a flight offering playing a central role in its connected trip strategy.

311Second, it must be noted that the applicant indicated that, even before launching its flight offering, it was already making sales of hotel rooms combined with non-flight travel services and that those sales would continue irrespective of whether or not it offered flights, which the Commission has not disputed.

312In that context, it must be observed that the fact that flights play a central role in the applicant’s connected trip strategy and that taxi transport and car rental services also form part of that strategy does not mean that each booking of a hotel room combined with such a service may be attributed to there being a flight offering on the applicant’s platform. As the applicant has stated, without being challenged by the Commission, if a user simply reserves a hotel room combined with a taxi or a rental car, that user has booked a connected trip without a flight.

313Lastly, even if, as the Commission maintains, other travel services, such as taxi transport or car rental, have only limited potential to attract customers to a hotel OTA platform, that does not prove that every customer who has booked such a service combined with a hotel room was attracted to the applicant’s platform because of the flight offering on that platform.

314Therefore, it is necessary to reject the Commission’s argument that all hotel bookings forming part of a connected trip, even those which do not include a flight, must be attributed to the transaction.

315In the second place, it should be observed that, as the applicant has maintained without being challenged by the Commission, the impact of the applicant’s connected trip strategy on its hotel business is not reflected solely by cross-sales, that is to say, by the sale of a travel service, such as a hotel room, to a customer who has just purchased another travel service, such as a flight, on the same platform. Indeed, that impact may also be found in ‘Value Leadership’, which refers to the increased likelihood that a customer will book accommodation where that accommodation is offered with a discount on any other travel service offered on the applicant’s platform, and in ‘Halo Effects’, which relate, in that context, to the estimated reduction in the cancellation rate and the increase in the rebook rate for a customer who has booked various services for a given trip.

316First, the applicant stated, without being challenged by the Commission, that the ‘Value Leadership’ effect occurs when a customer on its platform who intends to book a hotel room is offered another travel service, such as taxi transport or car rental, at a reduced price or even free of charge, in order to encourage that customer to make the booking. In such a case, the customer was attracted by the hotel room rather than by the flight offering.

317In that regard, it should be observed that it is apparent from Tables 20 and 21 of the contested decision that the vast majority, namely [confidential]%, of sales of hotel rooms attributed to ‘Value Leadership’ effects relates to sales of hotel rooms combined with non-flight travel services.

318Indeed, as the applicant has stated without being challenged by the Commission, taxi transport or car rental has clear tangible value with limited commitment (because it is not perceived as an important decision for the trip), and the applicant has a sufficient margin on its hotel room product, which is the first product booked, to fund the taxi or car rental cost.

319By contrast, as the applicant has argued without being challenged by the Commission, the number of flights sold as a result of the ‘Value Leadership’ effect is very limited, since customers typically book flights at a different time from when they book hotel rooms, they often do so through a different channel and they view flights as a much more critical and expensive purchase, warranting careful comparison of offers from different providers.

320Second, as regards the ‘Halo Effects’, the applicant explained, without being challenged by the Commission, that those effects referred to the greater ‘stickiness’ of the platform, because a customer who, on that platform, has purchased various services for a given trip, such as a flight and a hotel room, or a hotel room and taxi transport or car rental, was less likely to cancel the booking and more likely to make a new booking on the same platform.

321In that regard, it is apparent from Tables Nos 20 and 21 of the contested decision that the vast majority, namely [confidential]%, of sales of hotel rooms attributed to ‘Halo Effects’ related to sales of hotel rooms combined with non-flight travel services, with the result that the ‘Halo Effects’ which may be attributed to flights are limited.

322Therefore, the Commission erred in regarding all the hotel room bookings attributed to Value Leadership effects and Halo Effects as themselves attributable to the flight offering on the applicant’s platform and, thereby, to the transaction.

323In the third place, it is necessary to reject the Commission’s argument that the simultaneous nature of the applicant deploying its flight offering and the sudden growth in the number of hotel bookings attributable to Value Leadership effects and Halo Effects suggest a causal link between the two.

324Indeed, the applicant stated at the hearing, without being challenged by the Commission, that that growth was mainly due to the market recovering after the COVID-19 pandemic and to an increase in the number of taxi and car rental bookings associated with a hotel room booking, as reflected in the figures relating to ‘Value Leadership’ effects and ‘Halo Effects’ which the Commission reproduced in Tables 20 and 21 of the contested decision.

325In the light of the foregoing, it must be concluded that the arguments put forward by the Commission are not capable of establishing a causal link between the transaction and the additional hotel bookings which are not directly linked to flights, with the result that they must be excluded from the calculations of the increments in the applicant’s share of the hotel OTA market.

326It is therefore necessary to disregard the percentages calculated by the Commission which take into account the gains resulting from combined sales of hotels and non-flight travel services, namely the [0-5]% used in the modified TVM methodology and the [0-5]% used in the Figure 123 methodology.

(b)The second alleged error: the calculation of the number of overnight stays which the applicant would acquire post-transaction

327In that regard, it should be recalled that, in the context of the modified TVM methodology, the Commission calculated the increment in the applicant’s share of the hotel OTA market resulting from sales of hotel rooms combined with flights in eight steps, the first six of which concerned the calculation of the number of overnight stays which the applicant would acquire post-transaction and the last two of which consisted of converting that number into euro (TTV) and dividing the resulting amount by the size of the market, also expressed in euro (TTV).

328First, as a starting point, the Commission extracted from the TVM the number of flight transactions subject to cross-selling, estimated for 2025 at [confidential], of which [confidential] concerned the EEA (Table 19 in recital 1043 of the contested decision).

329However, given that (i) that figure dates from November 2021, the date when the TVM was drawn up, and (ii) the file contains a more recent estimate dating from May 2022, the Commission adjusted that figure to reflect that later estimate. To that end, the Commission applied an increase of [confidential] to the figure of [confidential], resulting in an estimated number of flight transactions available for cross-selling in the EEA of [confidential] (recital 1043(iii) of the contested decision).

330Second, an attach rate was applied to the number resulting from the first step. As stated in footnote 794 to the contested decision at page 143, the attach rate reflects the share of customers who, after purchasing one product (for example, a flight ticket), purchase (or ‘attach’) an additional product (for example, a hotel room). To that end, the Commission applied the rate of [confidential] used by the applicant in the TVM (recital 1046 of the contested decision), so that the estimated number of cross-sales of accommodation from flights in the EEA is [confidential].

331Third, a cannibalisation rate was applied to the number resulting from the second step, reflecting the number of users who book accommodation via a cross-sale from flights, but who would also have booked that accommodation through the applicant even if it did not offer flights, with the result that those bookings ‘cannibalised’ part of the bookings which would in any event have been made through the applicant. To that end, the Commission applied the rate of [confidential]% used by the applicant in the TVM (recital 1047 of the contested decision), so that the estimated number of cross-sales of accommodation from flights was [confidential], that is to say, [confidential]% of [confidential].

332Fourth, the number resulting from the third step was multiplied by the average number of overnight stays per transaction (that is to say, the number of nights for which accommodation is booked in a single transaction), in order to calculate the total number of nights in accommodation which the applicant would be able to sell to flight customers. To that end, the Commission used the figure of [confidential], used by the applicant in the TVM, such that the number of additional overnight stays resulting from cross-selling from flights amounted to a total of [confidential] (recital 1049 of the contested decision).

333Fifth, the Commission added the part of the ‘Value Leadership’ effects and ‘Halo Effects’ directly linked to flights, representing [confidential] nights booked in accommodation to the number resulting from the fourth step. That gave a total number of overnight stays booked in accommodation of [confidential] (recitals 1050 and 1058 of the contested decision).

334Sixth, given that the number of overnight stays resulting from the fifth step also included types of accommodation other than hotels, such as private accommodation, that number was reduced. To that end, the Commission assumed that [confidential]% of the overnight stays in question concerned hotels, which resulted in a total number of [confidential] hotel room nights (recitals 1051 and 1059 of the contested decision).

335In that regard, the applicant disputes, first, in the context of the modified TVM methodology, the [confidential]% increase in the estimated number of flight transactions available for cross-selling, second, in the context of that methodology, the use of a cannibalisation rate of [confidential]% and, third, in the context of the Figure 123 methodology, the alleged failure to apply any cannibalisation rate.

(1)The [confidential]% increase in the number of flight transactions available for cross-selling under the modified TVM methodology

336The applicant disputes the [confidential]% increase which the Commission applied under the modified TVM methodology to the estimated number of flight transactions available for cross-selling taken from the TVM (see paragraph 329 above).

337The applicant submits, more specifically, that the projections in one of its internal documents cited in Figure 119 of the contested decision (recitals 1043 and 1044), on which the [confidential]% increase is based, [confidential]. That was confirmed by the applicant’s 2023 Flights Budget, which was reviewed and approved by the applicant’s senior management and board of directors and which contained much more realistic estimates of flight transactions compared with those in Figure 119. By contrast, the projections on which the [confidential]% increase was based were not used in the applicant’s 2023 Flights Budget.

338The Commission disputes the applicant’s arguments.

339In that context, it should be observed, first, that, in its written pleadings, the applicant has not proved or further substantiated the alleged lack of internal consensus and approval of those projections by its management team and its board of directors and, second, that that lack of consensus and approval is not sufficient to call into question the credibility of those projections.

340In addition, it should be noted that the applicant’s reference to its 2023 Flights Budget lacks precision and does not make it possible to identify which elements of that budget contradict the projections set out in Figure 119 of the contested decision. Indeed, two of the four items of evidence submitted in that context by the applicant, namely the response to the Commission’s request for information No 6 [paragraph 31.1(ii)], and the response to the Commission’s request for information No 17 (paragraphs 3.1 to 3.3), refer to [confidential], whereas the other two items of evidence, namely the response dated 23 June 2023 to the statement of objections (paragraph 3.3.4.3), and the response dated 7 August 2023 to the second letter of facts (page 45), contain only general assertions by the applicant, which are similar to those made in the application and referred to in paragraph 337 above, and lack precision.

341Furthermore, the Commission stated, without being challenged by the applicant, that the projections which formed the basis of the [confidential]% increase were not theoretical, but were part of the ‘Go-to-Market model’ (‘the GTM model’), a marketing model which the applicant had drawn up in May 2022 with a view to the integration of the intervener’s activities into its own (recitals 889 to 896 of the contested decision).

342Lastly, as the Commission correctly submits, it follows from the document cited in Figure 119 that the projections referred to therein [confidential]. The only reference in that document to [confidential].

343Therefore, it must be concluded that the applicant has failed to prove that the Commission erred in applying, in the context of the modified TVM methodology, a [confidential]% increase in the estimated number of flight transactions available for cross-selling, with the result that the arguments raised by the applicant in that regard are not capable of calling into question the percentage of [0-5]% referred to in paragraph 52 above, which the Commission established in accordance with that method.

(2)The use of a cannibalisation rate of [confidential]% in the modified TVM methodology

344The applicant disputes the use of a cannibalisation rate of [confidential]% in the modified TVM methodology (see paragraph 331 above).

345In that regard, the applicant explains that the cannibalisation rate must reflect its share of the hotel OTA market, which, according to the applicant, was [confidential]%. In the applicant’s view, if it held a certain percentage of that market, it would be reasonable to assume that it would in any event have obtained the same percentage of bookings via cross-sales. Since, according to the Commission, that market share was [confidential]%, the cannibalisation rate ought also to have been [confidential]%.

346The Commission disputes that the cannibalisation rate corresponds to the applicant’s market share.

347In that context, it should be noted, as the Commission did, that the applicant has provided no evidence in support of the allegation that the cannibalisation rate of [confidential]%, which it used in the TVM, is actually related to its share of the hotel OTA market. Indeed, in that context, the applicant and the Commission refer to a number of documents referred to in recitals 1047 and 1048 of the contested decision, citing passages which relate to an attach rate, and not to a cannibalisation rate. Each of those documents refers to a cannibalisation rate of [confidential]%.

348Therefore, it must be concluded that the Commission did not err in applying a cannibalisation rate of [confidential]% in the context of the modified TVM methodology, with the result that the arguments raised by the applicant in that regard are not capable of calling into question the [0-5]% established by the Commission under that methodology.

(3)Failure to apply any cannibalisation rate in the Figure 123 methodology

349The applicant submits that, in applying the Figure 123 methodology, the Commission, without justification, failed to take into account the ‘cannibalisation’ effect, thereby considerably inflating the increment in its share of the hotel OTA market attributable to the transaction. The applicant adds that, since the Commission considered, in general, that it was appropriate to apply a cannibalisation rate, it ought to have also done so in the context of the Figure 123 methodology.

350The Commission denies that it failed to apply a cannibalisation rate in that context and maintains that it merely followed the applicant’s own methodology to calculate the number of accommodation transactions which were incremental to cross-sales from flights.

351In that regard, it should be noted that, in the Figure 123 methodology, the Commission took the number of accommodation bookings resulting from cross-sales from flights post-transaction from a slide in an internal presentation of the applicant, [confidential] (recital 1061 of the contested decision).

352The Commission relied, more specifically, on a passage in that slide in which it is stated that [confidential].

353Next, the Commission based its calculation on that number of [confidential] accommodation transactions, without applying a cannibalisation rate (recitals 1062 to 1064 of the contested decision).

354However, the Commission maintains that the indicated number of [confidential] accommodation transactions already takes account of the ‘cannibalisation’ effect.

355In that context, it must be stated that the references to the slide referred to in paragraph 351 above to [confidential] and to the verb [confidential] suggest that the number of [confidential] reflects [confidential].

356In that regard, it should be noted that, as is apparent from recitals 1043, 1046 and 1047 of the contested decision (see paragraphs 328 to 334 above), the attach rate is, in principle, applied before the cannibalisation rate is applied. Indeed, that calculation begins with the number of flight transactions available for cross-selling, to which an attach rate is applied, in order to estimate the number of accommodations actually cross-sold, to which, next, a cannibalisation rate is applied, in order to eliminate from the calculation the sales of hotel rooms which would have been made even if the applicant’s platform had no flight offering (see paragraph 264 above).

357Therefore, contrary to the Commission’s assertions in paragraph 213 of its defence, there is nothing in Figure 123 to indicate that the number of transactions set out therein is calculated taking into account the ‘cannibalisation’ effect.

358Lastly, when questioned on that point by the Court at the hearing, the Commission was unable to explain why there was a significant discrepancy between the [0-5]% calculated using the modified TVM methodology, to which a cannibalisation rate had been applied (recital 1047 of the contested decision), and the equivalent [0-5]% calculated using the Figure 123 methodology, if a cannibalisation rate had also been applied to the latter percentage, considering that both percentages were based on two internal documents which the applicant had prepared at virtually the same time, that is to say, during May 2022 (Figures 121 and 123 and recitals 1043 and 1061 of the contested decision).

359It follows that the slide displayed in Figure 123 is not capable of substantiating the [0-5]% established by the Commission under the Figure 123 methodology, with the result that that percentage must be disregarded.

(c)The third alleged error: the conversion into euro of the number of additional overnight stays acquired by the applicant post-transaction and the division of the resulting amount by the size of the hotel OTA market

360It should be recalled that the last two steps of the calculation of the increment in the applicant’s share of the hotel OTA market, resulting from sales of hotel rooms combined with flights, consist, first, in converting into euro the number of additional overnight stays that the applicant would acquire as a result of the transaction and, second, in dividing the resulting amount by the size of the market also expressed in euro (see paragraph 302 above).

361In that context, the applicant criticises the Commission’s use of the total average room night value which it sold in 2022 (EUR [confidential]) in order to convert into euro the number of overnight stays it would acquire as a result of the transaction, whereas it would have been more appropriate to use the average value of hotel rooms cross-sold from flights in 2021 (EUR [confidential]), which it used in the TVM. In that regard, the applicant’s view is that the Commission’s assumption that those two values would converge is speculative and that the use of the total average room night value cannot be justified by high inflation across the economy as a whole, including the hotel sector.

362In addition, the applicant submits that, in order to calculate the increase in its share of the hotel OTA market in 2025, the Commission used the size of that market it had estimated for 2022, which meant that that market would not grow from 2022 to 2025. According to the applicant, that assumption is not realistic, however, in the context of the continued growth of the online travel market, including following the recovery from the COVID-19 pandemic, and data provided by Euromonitor, which anticipates the 2025 market size to be 53% larger than the 2022 market size.

363The Commission disputes those arguments.

364First, as regards the size of the hotel OTA market in 2025, it should be noted that, contrary to the Commission’s allegations, the applicant did indeed dispute, during the administrative procedure, the assumption that that market would not grow after 2022 (footnote 1369 to the contested decision).

365In addition, it should be noted that, contrary to the Commission’s allegations, the fact that, in 2022, the market recovered better than Euromonitor had anticipated does not mean that that market would no longer grow after that year. In that context, it should be added that the Commission itself rejected Euromonitor’s estimates on a number of occasions as being unreliable.

366Furthermore, it is apparent from the contested decision that the flight OTA market grew considerably between 2023 and 2025 (Tables 15 and 16 of the contested decision) and that there is a link between that market and the hotel OTA market, in that flights often constitute the beginning of the booking journey and are the service with the highest opportunity to be booked on the same website as accommodation (recitals 587, 753, 756 and 928 to 930 of the contested decision).

367Lastly, as the applicant correctly states, the Commission did not adjust the market size for 2022 expressed in euro (TTV) to take account of inflation up until 2025, whilst the contested decision refers to high inflation in the economy as a whole including the hotel sector, to assumed annual inflation rates of 5.4% for 2023 and 3% for 2024 to 2026 and to average annual price increases per overnight hotel room stay of [confidential]% since 2020 (recitals 854, 1052 and 1054 of the contested decision).

368It follows that the Commission underestimated the size of the hotel OTA market in 2025 expressed in euro (TTV), by assuming that there would be no growth or inflation from 2022, with the result that the [0-5]% established by the Commission under the modified TVM methodology must be regarded as an overestimate.

369Second, as regards the Commission’s use of the total average room night value rather than the average value of hotel rooms cross-sold through flights, it is necessary to reject the applicant’s argument that the Commission’s assumption that those two values would converge was speculative.

370Indeed, even if the Commission did not dispute that the average value of hotel rooms booked via cross-selling through flights is generally lower than the total average room night value, in particular due to flight customers’ price sensitivity, it is apparent from the contested decision that, by means of its connected trip strategy which the transaction aims to strengthen, the applicant is seeking, inter alia, to acquire and retain high-value customers who generally spend more per night than the average customer (recital 936 of the contested decision). That fact is sufficient to justify the Commission’s use of the total average room night value, assuming that the size of the 2025 hotel OTA market reflects inflation since 2022.

(d)The fourth alleged error: the zero-flights counterfactual scenario as the starting point for the Commission’s calculations

371In that regard, it should be recalled that in Section 6.7.2.4 of the contested decision the Commission calculated the increment in the applicant’s share of the hotel OTA market which the transaction would bring about in 2025 based on a ‘zero-flights’ counterfactual scenario (see paragraph 124 above), that is to say, disregarding the number of hotel rooms which the applicant could also sell together with a flight in the counterfactual scenario.

372In recitals 727, 728, 1007 and 1008 of the contested decision, it is stated that, despite a request from the Commission to that effect, the applicant had not provided evidence making it possible to estimate the number of flights it would be able to sell in a counterfactual scenario envisaging lesser cooperation with the intervener or the development of its own flight platform, with the result that the Commission had no choice but to use the ‘zero-flights’ counterfactual scenario as the starting point for calculating the increment in the applicant’s share of the hotel OTA market.

373The applicant and the intervener contend, in essence, that it is for the Commission to gather the necessary evidence. Therefore, if the Commission wished to know the number of flights sold by the applicant in such a counterfactual scenario, it ought to have made an information request to that effect.

374It should be noted that the increment in the applicant’s share of the hotel OTA market attributable to the transaction would logically be lower than the percentage calculated by the Commission in Section 6.7.2.4 of the contested decision due to the fact that the applicant would also make a number of cross-sales of hotel rooms from flights in the appropriate counterfactual scenario (see paragraphs 119 to 121 above), which the Commission acknowledged at the hearing. It follows that, given that that percentage is marginal, taking account of the errors found above, the question whether the Commission could criticise the applicant for the lack of evidence concerning the number of flights which the applicant would achieve in the counterfactual scenario has no bearing on the outcome of the present action. Therefore, the arguments raised by the applicant in that regard must be rejected as being ineffective.

(e)Conclusion

375In the light of the foregoing, it is necessary, first, to disregard the percentages calculated by the Commission which take into account the gains resulting from sales of hotel rooms combined with non-flight travel services, namely the [confidential]% used in the modified TVM methodology and the [confidential]% used in the Figure 123 methodology (see paragraph 326 above), second, to disregard the [0-5]% established by the Commission under the Figure 123 methodology on account of the failure to apply a cannibalisation rate (see paragraph 359 above) and, third, to conclude that the [0-5]% established by the Commission under the modified TVM methodology is an overestimate on account of the underestimation of the size of the 2025 hotel OTA market (see paragraph 368 above) and because it reflects all those sales even if only part of them can be attributed to the transaction, given that the applicant would also have a flight offering in the counterfactual scenario (see paragraph 374 above).

4. The fourth complaint: the applicant disputes the increase in barriers to entry and expansion in the hotel OTA market

376In the second part of the third plea (paragraphs 180 to 198 of the application), the applicant disputes the Commission’s statement that the transaction would increase barriers to entry and expansion in the hotel OTA market for its existing and potential competitors, as set out in Section 6.7.2.3 of the contested decision.

377In that section, which is entitled ‘The Commission concludes that the Transaction is likely to increase barriers to entry/expansion for rival hotel OTAs and strengthen network effects’, the Commission recalls, in the first place, certain features of the hotel OTA market, namely the applicant’s significant market share and its significant strength in paid advertising (recitals 922 to 925 of the contested decision), the limited ability of other OTAs to ‘contest’ the applicant’s position (recitals 920 to 925 of the contested decision), the difficulties for them to create their own flight offering (recitals 944 to 947 of the contested decision) and the lack of alternative channels for acquiring hotel customers (recitals 973 and 988 to 994 of the contested decision).

378In the second place, the Commission recalls the usefulness of having a flight offering as a hotel customer acquisition channel (recitals 927 to 932 of the contested decision) and the additional traffic which a flight offering could attract to a hotel OTA platform (recitals 972, 976 to 981 and 986 of the contested decision), including by strengthening customer loyalty (recitals 933 to 938 of the contested decision).

379In the third place, the Commission reiterates that the objective of the connected trip strategy is to consolidate the applicant’s position on the hotel OTA market (recitals 939 to 943 of the contested decision) and explains how the applicant would integrate the intervener’s business into its own (recitals 956 to 965 of the contested decision).

380In the fourth place, the Commission indicates that the growth of the applicant’s hotel customer base post-transaction would result in increased network effects, taking into account also the limited extent of multi-homing (recitals 926 and 948 to 955 of the contested decision), and increased difficulties for other hotel OTAs to build up a customer base which would make it attractive for hotels to list their rooms on their platforms (recitals 971 to 973 and 982 to 987 of the contested decision).

381The applicant disputes only the last two factors referred to in paragraph 380 above, raising, in essence, three arguments which will be examined below, namely (i) the fact that the number of additional customers it could acquire as a result of the transaction would be too low to have an impact on network effects, (ii) the fact that the high prevalence of multi-homing on the hotel OTA market would preclude hotels from ending their existing relationships with rival hotel OTAs on account of the minimal decline in sales by those OTAs post-transaction and (iii) the fact that that decline is not sufficient to threaten the commercial viability of those rivals.

(a)Strengthening network effects

382In the first place, the applicant submits that the maximum [confidential]% increment in its share of the hotel OTA market entailed by the transaction is not sufficient to conclude that network effects would increase and, therefore, that barriers to entry or expansion in that market would be raised. The applicant is of the view that, as it has already been offering flights on its platform since the Phase 2 Agreement was implemented, any shift linked specifically to it offering flights ought already to have happened.

383In that context, the Commission cannot maintain that the extent of the increment in the applicant’s share of the hotel OTA market is irrelevant, since that increase reflects the number of additional hotel customers it could acquire post-transaction, which is decisive for the increase in the network effects underlying the theory of harm. Accordingly, the fact that the market share increment is negligible is of particular importance to determining whether the Commission has demonstrated that the transaction would result in such an increase.

384In the second place, the applicant submits that an increase in network effects ought to result in a change in hotels’ behaviour vis-à-vis hotel OTAs post-transaction, which the Commission has not demonstrated. However, the change in the number of hotel bookings on its platform is too low to affect hotels’ commercial decisions.

385In that context, the applicant refers to the results of the market investigation from which it is apparent that 63% of the hotels which replied to the Commission’s questionnaire during the administrative procedure indicated that, post-transaction, they would not change their inventory offered on the applicant’s platform, while 5% of hotels which replied indicated that they would reduce that inventory, 4% would increase it and 27% would wait to see how the market developed.

386The Commission disputes the applicant’s arguments.

387As a preliminary point, it should be recalled that network effects occur when the value of a product for a customer increases when the number of other customers also using it increases (footnote 64 to the Non-Horizontal Merger Guidelines and, see, to that effect, judgment of 17 July 2024, Bytedance v Commission, T‑1077/23, under appeal, EU:T:2024:478, paragraph 153).

388In the first place, as regards the applicant’s argument that the increment in its share of the hotel OTA market post-transaction would be so low that network effects could not increase, it must be observed that, in the contested decision and the Commission’s written pleadings, a number of other qualitative and quantitative factors are put forward which, according to the Commission, would result in that increase.

389As such, from a qualitative point of view, the Commission recalls the usefulness of having a flight offering as a hotel customer acquisition channel, in that it would give the applicant access to data enabling it to target potential hotel customers, and the fact that acquiring such an offering would enable it to create a travel ecosystem which would be difficult for other OTAs to replicate (recitals 928 to 930 and 939 to 946 of the contested decision), factors which the applicant has not succeeded in calling into question (see paragraphs 259, 260 and 293 to 295 above).

390In addition, from a quantitative point of view, the Commission refers to (i) the growth which the applicant might see post-transaction in flights, referring to the anticipated increase in its share of the flight OTA market and the number of visits to its platform by customers looking for a flight (recitals 931, 932 and 936 of the contested decision), (ii) the growth in the number of hotel customers post-transaction (recitals 926 and 955 of the contested decision) and (iii) the ‘Halo Effects’ caused by the transaction, which would increase customer loyalty and inertia, including of high-value customers (recitals 926, 933 to 938 and 955 of the contested decision).

391First, as regards the applicant’s alleged growth in flights, it should be recalled that the Commission’s theory of harm relates only to the hotel OTA market, in the absence of competition concerns regarding the flight OTA market.

392Moreover, as the applicant stated at the hearing and the Commission itself acknowledged (see paragraph 266 above), traffic data cannot demonstrate the potential for cross-sales of hotel rooms from flights, in that the websites which customers view or the offers on which they click when searching for a travel service do not reflect the actual purchases they make. Moreover, it should be noted that the Commission provided only figures for the number of views by customers searching for flights on the applicant’s platform, without presenting any data on the number of customers who searched for hotel rooms there.

393That said, it must be stated, as the Commission did, that an increase in potential flight customer traffic and, therefore, in sales of airline tickets logically results in a certain increase in potential customer traffic for hotels and, subsequently, in a number of additional hotel accommodation bookings, which the applicant does not dispute.

394Second, as regards the expected growth in the number of hotel customers, it must be noted that even if, apart from the calculations of the increment in the applicant’s hotel OTA market share, the Commission provided no data indicating, even approximately, the number of additional customers which the applicant could acquire post-transaction, it is sufficient to note that, in paragraph 223 of the defence and in recitals 926 and 955 of the contested decision, the Commission pointed out that there was a link between the increase in the alleged network effects and the growth in the number of hotel customers post-transaction, which is not disputed by the applicant.

395Third, as regards the ‘Halo Effects’ brought about by the transaction, it should be noted, as the Commission correctly states in paragraph 110 of the rejoinder, that those effects are reflected in the increment in the applicant’s share of the hotel OTA market. That increase is calculated by reference to the revenue (TTV), and not to the number of customers or even to the number of hotel bookings, with the result that it also reflects the additional revenues obtained as a result of the increased loyalty and inertia of the applicant’s existing customers, that is to say, the ‘Halo Effects’, and as a result of the higher spend by newly acquired and existing high-value customers, to which the transaction would give rise.

396It must be concluded, as the applicant submits, that the increase in the applicant’s share of the hotel OTA market is, of all the quantitative indicators put forward by the Commission, the most accurate one, in that its percentage reflects not only the number of additional hotel customers which that platform would reach post-transaction, but also the additional hotel bookings, the ‘Halo Effects’ and the higher spend by newly acquired and existing high-value customers to which the transaction would give rise. In that context, it should be recalled that, although that increase represents only a modest percentage (see paragraph 375 above), the possibilities available to the applicant to cross-sell hotel rooms arise not only when the flight is sold, but also before or after such a sale, as a result of the data which the OTA possesses (see paragraphs 29 to 32 above). Therefore, the actual impact of the transaction on the existing network effects on the hotel OTA market in favour of the applicant is likely to exceed what that modest percentage suggests.

397Furthermore, it has been established that the hotel OTA market was characterised by a low level of competition, by significant barriers to entry and expansion, in particular on account of there being network effects, by the lack of entry of new players or of expansion by existing players and by the dominant position of the applicant, which acts independently with regard to competing hotel OTAs and hotels (see paragraphs 159 to 241 above).

398In that context, the contested decision refers, first, to network effects and to the fact that the inter-dependency between hotels and end customers results in a self-reinforcing dynamic which is difficult for smaller online travel agencies and new entrants to break (recital 221 of the contested decision) and, second, to the significant gap between the applicant’s market share and that of its main competitor (Expedia), as well as to the applicant’s expectation that it would continue to grow in the near future (sections 6.4.4 and 6.4.5 of the contested decision).

399In the light of such market conditions, it was open to the Commission to conclude that even a relatively small increase, in quantitative terms, in the applicant’s share of the hotel OTA market would be likely to strengthen existing network effects, in that not only would that increase the applicant’s customer base, but it would also prevent rival OTAs from expanding their customer base, which could have a chilling effect on competitive dynamics.

400In the second place, as regards the applicant’s argument that an increase in network effects ought to result in a change in hotels’ behaviour towards hotel OTAs post-transaction (see paragraph 384 above), it should be noted that the Commission’s starting premiss is that the applicant is already an unavoidable trading partner for hotels, given the size of its customer base and its hotel portfolio (recitals 406 and 407 of the contested decision) and that hotels depend on the applicant for a significantly larger proportion of their sales than on other OTAs, given that for most hotels between 61% and 100% of their total bookings via the OTA channel are made via the applicant (recital 424 of the contested decision).

401In recitals 926 and 955 of the contested decision, the Commission states that, post-transaction, the applicant’s platform would become even more attractive to hotels because of the additional customers it could acquire, which would guarantee it the broadest and best hotel content, with the result that it would attract even more end customers to its platform, which would increase existing network effects on the hotel OTA market even further.

402In that regard, it should be observed that, even if the number of additional customers which the applicant could acquire post-transaction were not large enough to cause those hotels to transfer an additional part of their hotel room inventory to its platform, that fact supports the applicant’s dominance on the hotel OTA market rather than its dominant position post-transaction being strengthened only to an insignificant degree.

403Indeed, in recital 950 of the contested decision, the Commission observes that, even before the transaction, 88% of hotels were present on the applicant’s platform and that, of those, a large number of hotels made their inventory fully available to the applicant, with the result that there remained relatively few hotels which could expand their inventory in any way, as the Commission correctly states. Similarly, in recital 407 of the contested decision, the Commission refers to internal documents of the applicant which state that in 2019 the applicant had already maximised the size of its hotel portfolio, so that ‘the … [return on investment] of capturing the remainder appears low’.

404It follows that the lack of evidence that, post-transaction, hotel room inventory of competing hotel OTA platforms would be transferred to the applicant’s platform does not mean that that transaction is incapable of strengthening existing network effects, but above all reflects the applicant’s current dominant position as an unavoidable partner for hotels which would be consolidated by that transaction.

(b)Multi-homing by hotels

405The applicant submits that the high prevalence of multi-homing on the hotel OTA market means that the alleged minor increase in its share of the hotel OTA market cannot significantly reinforce network effects in its favour.

406The Commission disputes the applicant’s arguments.

407In that regard, it submits that the very existence of multi-homing (see paragraph 231 above) is not disputed in the contested decision, but that the decision states that most hotels listed their rooms both on the applicant’s platform and on Expedia, but not on any other hotel OTAs.

408According to the Commission, multi-homing is limited since it is costly for hotels, and the applicant provides no evidence to support its statement to the contrary. In that context, the Commission maintains that the evidence submitted by the applicant in support of its argument concerning the use of channel managers is inadmissible in that it was not put forward during the administrative procedure.

409Therefore, hotels’ incentives to multi-home depend on the bookings which they expect to make via each hotel OTA. According to the Commission, if post-transaction, the number of bookings on alternative hotel OTAs is likely to diminish or, at least, risks not increasing as much as would have been possible absent the transaction, hotels will, post-transaction, be more likely to terminate their commercial relationships with the applicant’s competitors, while joining or remaining on the applicant’s platform in order to access its increasing end-customer base (recitals 230, 952 and 953 of the contested decision).

410In that regard, it should be noted that it is stated in recitals 950 and 951 of the contested decision that most independent hotels were present on both the applicant’s platform (88% of those hotels) and on Expedia’s platform (61% of those hotels), but not on the platform of any other hotel OTA. In particular, 28% of those hotels use only one OTA, 39% use two OTAs, 17% use three OTAs and only 15% are present on more than three OTAs. Those figures were not disputed by the applicant in its written pleadings.

411From that, the Commission infers, in recital 952 of the contested decision, that hotels can generally manage relationships with only one or two hotel OTAs in parallel, such that they generally opt for the OTAs with the largest customer base, in order to reach the greatest number of potential end customers.

412However, in paragraph 187 of the application, the applicant submits that [confidential]% of chain hotels and [confidential]% of independent hotels on its platform use channel managers, which allow them to upload their hotel room inventory once, after which the inventory is distributed to numerous different OTAs. According to the applicant, the significant use of channel managers demonstrates the ease with which hotels manage relationships with multiple hotel OTAs.

413In the defence, the Commission does not dispute the validity of that argument, but pleads that the evidence which the applicant submitted in support of it, namely a study dated 4 December 2023, which post-dates the contested decision and was therefore not submitted to it during the administrative procedure, is inadmissible.

414Even though it is true that that study was issued after the contested decision, such that it must be held, as the Commission did, that it is inadmissible, it is apparent from recitals 431 and 1081 of the contested decision that the applicant did indeed rely on the widespread use of channel managers during the administrative procedure. More specifically, in Section 5.D of the response to the statement of objections, the applicant expressly referred to such use raising, in broad terms, the same line of argument as it had put forward in the application (see paragraph 412 above), with the difference, however, that in paragraph 4.68 (iv)(a) of Section 4.B.6 of that response, it indicated a percentage of [confidential]% of hotels on its platform’s active listings in 2022, instead of the respective percentages of [confidential]% and [confidential]% of chain hotels and independent hotels which it had indicated in the application.

415It follows, first, that the applicant’s argument based on the widespread use of channel managers is admissible and, second, that the Commission failed to examine that argument, which, if it were established, could call into question its statement that hotels can generally manage only a limited number of hotel OTAs.

416In that regard, it should be recalled that where an institution has a broad discretion, such as the Commission in the control of concentrations, observance of procedural guarantees is of fundamental importance, including the obligation for the institution to examine carefully and impartially all the relevant aspects of the situation in question (see judgment of 26 June 2025, EVH and Others v Commission, C‑464/23 P, C‑465/23 P, C‑467/23 P, C‑468/23 P and C‑470/23 P, EU:C:2025:478, paragraph 211 and the case-law cited).

417It must be held that the Commission did not comply with that obligation when it failed to examine the applicant’s argument concerning the widespread use of channel managers. Therefore, the evidence provided by the Commission is not capable of substantiating its conclusion that the applicant’s acquisition of additional customers post-transaction would be likely to result in hotels terminating their commercial relationships with other hotel OTAs. Indeed, if the use of a channel manager were to allow hotels to maintain relationships with a number of OTAs without substantial additional costs or workload, the relatively limited loss of customers suffered by OTAs other than the applicant post-transaction ought not to encourage hotels to reconsider their relations with those OTAs. However, that finding does not detract from the conclusion that the applicant is already an unavoidable partner for hotels, which the transaction would consolidate by chilling competitive dynamics in the hotel OTA market (see paragraphs 397 and 404 above).

(c)The increased difficulties for other hotel OTAs in building up an attractive customer base for hotels

418In the first place, the applicant submits that the Commission has not shown that the minimal loss of sales which its competitors would suffer post-transaction would affect their competitiveness.

419In the second place, the applicant submits that the contested decision contradicts itself, in that it maintains that the transaction would enable the applicant to capture sales from the hotel’s direct channel (recital 1108 of the contested decision), that all its increases in sales of hotel rooms would be at the expense of rival hotel OTAs (recital 1160 of the contested decision) and that the transaction would not generate new demand for hotel OTA services (recital 1103 of the contested decision).

420In that regard, it is the applicant’s view that if the Commission had examined the impact on rival hotel OTAs of an increase in its market share, it would have found that, since direct hotel sales represent [confidential]% of all online hotel bookings in Europe, any increase in the applicant’s sales would likely come in large part from users who would not otherwise have booked with a rival hotel OTA.

421In the third place, the applicant submits that it is not necessary for an OTA to offer flights in order to be active on the hotel OTA market and that OTAs which wish to do so, may conclude commercial affiliated agreements with flight OTAs, such as Expedia, Kiwi or Hopper.

422The Commission disputes the applicant’s arguments.

423In that regard, it should be recalled that the Commission is of the view that the transaction would affect the ability of hotel OTAs competing with the applicant to develop a sufficiently attractive customer base for hotels to be willing to contract with them (recitals 971 to 973 and 982 to 987 of the contested decision).

424According to the Commission, one of the ways for OTAs already active on the hotel OTA market and for potential entrants to expand their position on that market is to develop a hotel customer base through flights. Some hotel OTAs would, in particular, use their flight offerings to attract customers to whom they could sell hotel rooms, while others could do the same by means of a commercial affiliated agreement with a flight OTA (recitals 971, 974 and 993 of the contested decision). That is all the more so since flights are one of the few customer acquisition channels still available for hotel OTAs (recitals 988 to 992 of the contested decision).

425The Commission therefore considers it reasonable to assume that the growth which the applicant would have to achieve on the flight OTA market post-transaction would be to the detriment of competitors, who would likely see their market shares decreased, which would reduce their access to hotel customers (recitals 972, 973, 986 and 987 of the contested decision).

426In the first place, it should be recalled that it is possible that a merger will result in a significant impediment to effective competition by reducing competitors’ sales prospects to a level below minimum viable scale, with the result that they are discouraged from expanding their position or entering the relevant market (paragraph 112 of the Non-Horizontal Merger Guidelines), provided that foreclosed competitors play a sufficiently important role in the competitive process on the relevant market, which may be the case despite a relatively small market share (paragraph 48 of the Non-Horizontal Merger Guidelines).

427The applicant submits that (i) the minimal loss of sales of hotel rooms post-transaction would not affect the competitiveness of other hotel OTAs, since the transaction would allow it to increase only marginally its share of the hotel OTA market and (ii) part of that increase would be to the detriment of the hotels’ direct channel, and not of other hotel OTAs.

428In that regard, it should be recalled that it is true that the increase in the share of the hotel OTA market which the applicant could achieve as a result of the transaction is in any event modest (see paragraph 375 above), but, as the Commission correctly submits, the effect of a concentration must be measured by reference to the level of competition which would exist in the absence of the transaction, so that even a relatively small reduction, in quantitative terms, in the potential customer base of the applicant’s rival OTAs could have a chilling effect on the pre-existing low level of competition (see paragraph 399 above).

429In the second place, contrary to the applicant’s assertions, it should be observed that, in the contested decision, the Commission did not conclude that the transaction would have a significant impact on the hotels’ direct channel.

430In the contested decision, the Commission merely found that the majority of hotels which expressed a view during the market investigation were of the opinion that the transaction would enable the applicant to acquire at least 5% of sales via their direct channel, through which approximately [confidential]% of all online hotel bookings in Europe are made (recitals 468 and 1108 to 1110 of the contested decision).

431In the third place, contrary to the applicant’s assertions, it is not apparent from the contested decision that the Commission’s view is that having a flight OTA offering is necessary to compete effectively in the hotel OTA market. Rather, the Commission’s view is that flights are one of the few channels still available to OTAs to attract hotel customers (recital 971 of the contested decision).

432In that regard, it should be noted that, besides the applicant and its main competitor Expedia, the market for hotel OTAs can broadly be divided into two categories of players, namely (i) a number of OTAs which have their own hotel content and (ii) a number of OTAs which, whilst primarily active in the flight sector, also offer hotel rooms as a secondary activity, mainly through a commercial affiliated agreement with another hotel OTA (recital 982 of the contested decision).

433As regards, first, OTAs with their own hotel content, such as HRS, Weekendesk and Travelminit, the Commission maintains, in recitals 974, 993 and 1003 of the contested decision, that (i) it is unlikely that, post-transaction, the applicant would maintain or conclude commercial affiliated agreements with them which would enable them to offer flights on their platforms and (ii) the other OTAs which would be ready to do so, might not be at the level of the intervener. In that context, it is sufficient to observe that the applicant has not disputed those assertions by the Commission in its written pleadings.

434As regards, second, OTAs which do not have their own hotel content, such as eDreams Odigeo, Kiwi and Tix, or only to a limited extent, such as Trip.com, Lastminute and TUI (recitals 983 to 985 of the contested decision), it should be recalled that they source their hotel content, either fully or in part, from other OTAs and are therefore active at the B2B level of the hotel OTA market only on a modest scale, with market shares of approximately 1% or less (recital 216 of the contested decision).

435Nevertheless, in recitals 983 to 986 of the contested decision, the Commission states that a number of those OTAs, such as Trip.com, Lastminute, TUI, eDreams Odigeo, Kiwi and Tix, have the potential to expand their activities on the hotel OTA market, using their flight platforms to develop a hotel customer base likely to purchase own hotel content, gain scale and become more attractive partners for hotels.

436By contrast, in Section 6.4.6.4, entitled ‘No evidence of likely entry or expansion that would effectively constrain [the applicant]’, the Commission states that its market investigation ‘has not indicated that there will be likely entry or expansion into the hotel OTA market in the EEA that would effectively constrain [the applicant]’ (recitals 473 to 479 of the contested decision).

437In recitals 480 and 481 of the contested decision, it is in particular indicated that, in essence, the potential expansion of OTAs already active on the hotel OTA market cannot be regarded as effective expansion which would challenge the applicant’s dominant position on that market, given their low market shares and the fact that they relied heavily on the hotel content of other OTAs.

438In that context, in response to a question put by the Court by document dated 10 June 2025, in order to clarify whether there is any contradiction between the assertions referred to above in the light of the obligation under the second paragraph of Article 296 TFEU to state reasons, the Commission clarified at the hearing that, in Section 6.4.6.4 of the contested decision, it had observed that actual or potential competitors were not likely to enter or expand on the hotel OTA market so as to exert real competitive pressure on the applicant. By contrast, in recitals 983 to 985 of the contested decision, it considered that it was possible for certain flight OTAs to expand their activities on the hotel OTA market, although some of them were better placed than others to do so since they had a certain amount of proprietary hotel inventory. More specifically, the Commission explained that its theory of harm was based primarily on the objective of protecting the ability of those OTAs to expand and, consequently, the contestability of the applicant’s position on that market in the future.

439In that regard, in relation to the role which the OTAs referred to in recitals 983 to 986 of the contested decision could play in the competitive process on the hotel OTA market, as referred to in paragraph 48 of the Non-Horizontal Merger Guidelines, it should be observed, first, that the applicant expressly mentions Trip.com, Lastminute, TUI and eDreams Odigeo in paragraphs 200 and 201 of the application as being among its ‘main rivals’ and having well-recognised brands and, second, that the widespread use of channel managers, relied on by the applicant during the administrative procedure and in the application (see paragraphs 412 and 414 above), enables OTAs to have access to hotels, despite their relatively small size on the hotel OTA market. On that market, the applicant is subject to weak competitive constraints, since its market share is much larger than rival OTAs’ shares. In addition, network effects are significant and the size of the customer base, which is characterised by a high degree of inertia, plays a major role in the competitiveness of hotel OTAs (recitals 264 to 268 and 349 to 352 of the contested decision). In those circumstances, the residual competition exerted by the few remaining competitors is of particular importance in a market with low levels of competition. Indeed, the effectiveness of merger control and the Commission’s ability to oppose the strengthening and consolidation of existing dominant positions would be undermined if it were subject to a particularly high burden of proof to demonstrate the likely growth of a dominant undertaking’s competitors, since that dominance makes such development more difficult to prove a priori, or at least more difficult to predict.

440Lastly, as regards the impact on the growth potential of OTAs referred to in recitals 983 to 986 of the contested decision, it must be stated that, despite the modest increment in its share of the hotel OTA market post-transaction, the applicant would achieve such increment via strong growth in one of the few channels still available for acquiring hotel customers, namely flights (recitals 973 and 988 to 994). Therefore, since those OTAs, which are mainly flight OTAs, depend heavily on that channel in order to develop their hotel customer bases, the transaction would have a disproportionate effect on their competitiveness. Indeed, post-transaction, the applicant would not only be the dominant hotel OTA, but would also become the leader in the flight OTA market, thereby creating a travel ecosystem which would be difficult for other OTAs to replicate (see paragraph 294 above), including for those referred to in recitals 983 to 986 of the contested decision.

441It follows that the arguments raised by the applicant are not capable of calling into question the Commission’s conclusion that the transaction would reduce other hotel OTAs’ scope for expansion.

5. The fifth complaint: the applicant disputes the impact which the transaction would have on hotels and consumers

442In the second part of the third plea (paragraphs 220 to 229 of the application), the applicant challenges the Commission’s statement, as set out in Sections 6.7.2.5 and 6.7.2.6 of the contested decision, that hotels and their customers would likely be harmed, as they would likely be required to pay higher commission rates and hotel room prices, respectively, as a result of the transaction.

443In the first place, the applicant submits that the Commission erred in concluding that commission rates would likely be higher as a result of the transaction and that it cannot suggest that maintaining commission rates at their level post-transaction is evidence of an anticompetitive effect, without establishing that it was likely that, absent the transaction, those commission rates would have decreased. Therefore, the Commission has not proved that the applicant would have less incentive to reduce commissions and offer more beneficial terms post-transaction.

444In the second place, the applicant maintains that it is not an expensive hotel OTA. Indeed, the Commission did not assert that it is an expensive hotel OTA, but simply stated that it is not always the cheapest channel for consumers (recital 1135 of the contested decision). The applicant’s view is that such an observation does not make it the most expensive, or even an expensive, channel.

445The Commission disputes those arguments.

446In that regard, in the first place, it should be recalled that the expression ‘increased prices’ is often used as a shorthand for the various ways in which a merger may result in competitive harm. In particular, such harm includes not only an undertaking’s ability to increase prices profitably, but also covers situations where prices are decreased less, or are less likely to decrease, than they otherwise would have without the merger and where prices are increased more or are more likely to increase, than they otherwise would have without the merger (paragraph 8 of and footnote 7 to the Horizontal Merger Guidelines and footnote 8 to the Non-Horizontal Merger Guidelines).

447It should be noted that the Commission does not state that the transaction is likely to cause such competitive harm, since it does not argue that it is likely that commission rates and hotel prices would increase post-transaction or that they would have decreased absent the transaction.

448Indeed, in recital 1097 of the contested decision, the Commission simply stated that, while it cannot be excluded that the applicant would increase commission rates post-transaction, the evidence on file indicated that it is likely that, as its position becomes more difficult to contest and its bargaining position vis-à-vis hotels further increases, its incentives to lower commission rates and to provide more beneficial terms and conditions to hotels will likely decrease (recital 1097 of the contested decision).

449In the second place, it should be recalled, first, that the transaction would enable the applicant to increase its sales of hotel rooms, in particular by means of cross-sales from flights or through ‘Halo Effects’ in so far as they are linked to flights (see paragraphs 390 and 394 to 397 above), second, that, absent the transaction, those additional sales could, at least in part, have been made by other hotel OTAs or via the hotels’ direct channel (see paragraph 430 above), third, that the applicant has failed to rebut the Commission’s statement that its commissions are higher than those of other hotel OTAs (see paragraphs 192 to 206 above) and, fourth, that the hotels’ direct channel is less expensive for hotels than hotel OTAs, on account of the commissions charged by those OTAs (recital 1102 and Table 24 of the contested decision). It follows that the applicant’s arguments are not capable of calling into question the plausibility of the Commission’s argument that at least some hotels would face higher costs post-transaction.

450In the third place, it must be observed that the Commission does not state that the applicant is the most expensive hotel OTA, or even an expensive OTA, for consumers. However, it indicated, without being challenged by the applicant, that the prices displayed by hotels on its platform were, in general, higher than the prices which hotels displayed on their own websites (recital 1135 of the contested decision). Therefore, in so far as the transaction would be likely to divert at least some bookings from the hotels’ direct channel to the applicant’s platform, it was open to the Commission to conclude that the customers concerned could pay a higher price for their hotel room.

451Consequently, it is necessary to reject the arguments by which the applicant disputes the fact that hotels and their customers are liable to be harmed.

6. The Court’s overall assessment

452In the light of the examination of the substance of the various arguments put forward by the applicant in the context of the second part of the third plea against the theory of harm, it is necessary to make an overall assessment of whether all the relevant factors on which the Commission relied and which may be regarded as established are sufficient to demonstrate that there is a significant impediment to effective competition within the meaning of Article 2(3) of the EC Merger Regulation, and also to examine the first part of the third plea, by which the applicant disputes the legal standard applied by the Commission in that context (see paragraph 157 above).

(a)The legal standard

453In the context of the first part of the third plea, the applicant, supported by the intervener, submits that, in the contested decision, the Commission applied an incorrect legal standard in concluding that the transaction would give rise to a significant impediment to effective competition within the meaning of Article 2(3) of the EC Merger Regulation, on the hotel OTA market.

454In that regard, the applicant and the intervener submit that the Commission addressed the present case from two alternative approaches, in that, first, it maintained, in essence, that the transaction would have the effect of increasing the applicant’s share of the hotel OTA market and, second, it appears to assert that, irrespective of the precise increment in the applicant’s share of the hotel OTA market, the transaction would nevertheless strengthen its position on that market simply because such an increment, however negligible, existed.

455The applicant and the intervener are of the view that the second approach, whereby the Commission dispenses with the need to examine the likely effects of the transaction on the basis of a presumption that even a marginal increase in its position would necessarily significantly impede effective competition, is contrary to Article 2(3) of the EC Merger Regulation and fails to have due regard to the case-law, in particular the judgments of 21 February 1973, Europemballage and Continental Can v Commission (6/72, EU:C:1973:22); of 16 March 2023, Towercast (C‑449/21, EU:C:2023:207); and of 25 October 2002, Tetra Laval v Commission (T‑5/02, EU:T:2002:264). The applicant is also of the view that, in accordance with that case-law, the Commission ought to have demonstrated that the transaction would affect rivals’ ability to compete and to constrain the applicant to such an extent that it substantially fetters competition. In that context, the Commission cannot ignore the fact that the increment in its share of the hotel OTA market post-transaction would be minimal, by relying on ‘qualitative’ elements in order to conclude that there was a significant impediment to effective competition.

456In that regard, it must be stated that it may be difficult, if not impossible, to distinguish between ‘quantitative’ and ‘qualitative’ arguments or evidence. An argument of a ‘qualitative’ nature is often supported by figures, with the result that it may appear artificial to separate one from the other and to accept the relevance of the quantitative element alone where it is in fact intended to support an argument of a qualitative nature (see, to that effect, judgment of 17 July 2024, Bytedance v Commission, T‑1077/23, under appeal, EU:T:2024:478, paragraph 40).

457In the area of merger control, the Commission must base its decision on a sufficiently cogent and consistent body of evidence, having regard to the essential function of the evidence, which is to establish convincingly the merits of an argument (see, to that effect, judgment of 13 July 2023, Commission v CK Telecoms UK Investments, C‑376/20 P, EU:C:2023:561, paragraphs 77 and 87 and the case-law cited), and in the light of the principle of the unfettered evaluation of evidence, from which it follows, inter alia, that the only relevant criterion for the purpose of assessing evidence is its credibility (see, to that effect, judgment of 19 December 2013, Siemens and Others v Commission, C‑239/11 P, C‑489/11 P and C‑498/11 P, not published, EU:C:2013:866, paragraph 128 and the case-law cited).

458It follows that it cannot be ruled out that the Commission may base the assessments it has made in the context of Article 2(2) and (3) of the EC Merger Regulation on qualitative elements, whether or not they are supported by quantitative elements, provided that the evidence adduced by the Commission is sufficiently cogent and consistent, that it contains all the relevant evidence which must be taken into consideration and that it is capable of substantiating the conclusions drawn from it.

(b)The strengthening of the applicant’s dominant position

459In that context, it should be stated that the applicant has not succeeded in calling into question the Commission’s conclusion that it has a dominant position on the hotel OTA market, in the light of its share of that market, the fact that it receives from hotels a higher commission than rival OTAs, that it is an unavoidable trading partner for hotels, that it is not subject to sufficient pressure from rival OTAs or sufficient out-of-market constraints and that it outperforms its competitors on each parameter determining the competitiveness of a hotel OTA. Furthermore, it has been established that the hotel OTA market was characterised by there being high barriers to entry and expansion, on account of significant network effects, customer loyalty and inertia, such that entry or expansion capable of effectively constraining the applicant was not likely (see paragraphs 160 and 240 above).

460In addition, it has been established that the transaction was part of the connected trip strategy for which a flight offering was of critical importance and which was implemented by the applicant in order to [confidential] (see paragraphs 251 to 254 and 299 above). Indeed, flights are often ([confidential]%) the ‘entry point’ of the trip, in so far as they are generally booked before accommodation and can be used by a hotel OTA to generate additional customer traffic and obtain access to customer data (see paragraph 259 above), given that [confidential]% to [confidential]% of hotel OTA customers also require flight OTA services (see paragraph 263 above).

461Therefore, the transaction would enable the applicant to optimise its flight business in the fastest and most efficient way by acquiring its preferred flight solution and to achieve growth in its flight business, which would result in growth in its hotel business as a result of cross-sales of hotel rooms from flights.

462It follows that the evidence provided by the Commission is capable of substantiating its conclusion that the transaction is likely to strengthen the applicant’s dominant position on the hotel OTA market.

(c)The significance of the impediment to effective competition resulting from the strengthening of the applicant’s dominant position

463In that regard, as the applicant correctly states, it is apparent from the case-law that the mere fact that the acquiring undertaking already holds a dominant position on the relevant market may constitute an important factor, but does not in itself suffice to justify a finding that a reduction in the potential competition which that undertaking must face constitutes a strengthening of its position (see, to that effect, judgment of 25 October 2002, Tetra Laval v Commission, T‑5/02, EU:T:2002:264, paragraph 312).

464Indeed, the fact that a concentration would create or strengthen a dominant position is not, in itself, sufficient for that concentration automatically to give rise to a significant impediment to effective competition, in that there is no automatic link between the test of dominance and the test of a significant impediment to effective competition (see, to that effect, judgments of 13 November 2024, NetCologne v Commission, T‑58/20, EU:T:2024:813, paragraph 108; of 13 November 2024, Deutsche Telekom v Commission, T‑64/20, EU:T:2024:815, paragraph 193; and of 13 November 2024, Tele Columbus v Commission, T‑69/20, EU:T:2024:816, paragraph 223).

465In that context, it must be noted, first, that the Commission’s analysis of the increment in the merged entity’s share of the hotel OTA market is vitiated by a number of errors, with the result that it cannot therefore be ruled out that that increment would be limited to a few tenths of a per cent (see paragraph 375 above), and that it has also not been established that the growth which the applicant could achieve, post-transaction, would lead hotels to transfer an additional part of their hotel room inventory to the applicant’s platform (see paragraph 402 above) or to terminate their commercial relationship with other hotel OTAs (see paragraph 417 above), or that such growth would enable the applicant to increase the commissions received from hotels (see paragraph 447 above).

466Second, a number of recitals of the contested decision refer to the fact that the transaction would make the applicant’s dominant position on the hotel OTA market less contestable (recitals 982 to 987 of the contested decision). Similarly, at the hearing, the Commission emphasised that the primary intention of its theory of harm was to avoid the expansion of the applicant’s existing competitors or the entry of potential competitors on that market from becoming increasingly difficult because of the gradual increase in barriers to entry and expansion to that market, with the result that those competitors would, in the future, be less and less able to challenge the applicant’s position thereon.

467In that regard, it must be observed that the EC Merger Regulation does not preclude the Commission from relying on a theory of harm in which it does not allege that the concentration at issue would necessarily reduce the already low level of competition existing on the relevant market, but considers that the concentration would sustain that level of competition, to the benefit of the parties to the concentration and to the detriment of existing and potential competitors and consumers.

468In such a case, the significance of the impediment to effective competition on the relevant market found by the Commission does not necessarily follow from a significant reduction, post-transaction, of the competitive pressure on the parties to the concentration, but from the low level of competition on that market becoming consolidated and perpetuated. Such an effect on the relevant market must be regarded, first, as significant for the purposes of Article 2(3) of the EC Merger Regulation and paragraph 10 of the Non-Horizontal Merger Guidelines and, second, as affecting a sufficiently large part of the relevant market within the meaning of paragraphs 112 and 113 of those guidelines.

469It must be observed that the Commission put forward two specific circumstances which demonstrated that, in the present case, the strengthening of the applicant’s dominant position had characteristics which meant it also had to be regarded as capable of significantly impeding effective competition on the hotel OTA market.

470First, it must be stated that, in the specific situation of a market characterised by strong network effects and a significant gap between the dominant leader and its main competitors, such as the hotel OTA market in the present case, it is open to the Commission to conclude that even a relatively small increase, in quantitative terms, in the market share of the dominant undertaking, in the present case the applicant, would be capable of strengthening the existing network effects. Not only would that expand the dominant undertaking’s customer base, but also prevent competitors from expanding their customer bases, which could, given the importance of the size of its customer base for a hotel OTA due to network effects, have a chilling effect on the already weak competitive dynamics and result in the consolidation of the leader’s existing dominant position, making that position even less contestable (see paragraphs 396 to 399 above).

471Second, it should be recalled that, in the present case, the increment in the applicant’s market share would be achieved through high growth in one of the few channels for acquiring hotel customers which the applicant does not yet dominate, namely flights, thereby creating a travel ecosystem which would be difficult for other OTAs to replicate, since the applicant would be the leader on both the hotel OTA market and the flight OTA market (see paragraph 440 above). That growth is, in addition, likely to affect in particular certain OTAs which the Commission expressly identified in the contested decision and which, despite their currently modest position on the hotel OTA market, must be regarded as playing a sufficiently important role in the competitive process on that market that the impact on their competitiveness can be regarded as capable of having a negative impact on the dynamics of competition (see paragraph 439 above).

472In that context, it should be recalled that the intervener is generally regarded as a leading OTA which has a large network of direct relations with airlines and which scores highly on various parameters, in particular on price, but also on the breadth of its offering and its technical capabilities. It is also common ground that the transaction would enable the applicant to accelerate the growth it has already achieved on the flight OTA market by means of the Phase 2 Agreement, which could only benefit its hotel business, both by cross-sales of hotel rooms and by increased loyalty and inertia of its hotel customers.

473It follows that, on the basis of an overall assessment of the evidence in support of the theory of harm which can be considered to be established (see paragraph 156 above) and in view of the Commission’s discretion with regard to economic matters (see paragraph 158 above), the Court considers that the Commission has proved, to the requisite legal standard, that the transaction would give rise to a significant impediment to effective competition within the meaning of Article 2(3) of the EC Merger Regulation, with the result that the first and second parts of the third plea must be rejected.

D. The third part of the third plea: the contested decision errs in finding that the transaction’s efficiencies would not outweigh any competition concerns

474In the third part of the third plea, the applicant submits that the transaction brings pro-competitive advantages for flight OTA and hotel OTA customers. In Section 7 of the contested decision, the Commission failed to assess correctly those efficiencies, in that it considerably overstated the potential disadvantages of the transaction without recognising its obvious benefits.

475The applicant relies, more specifically, on a number of efficiencies which would arise, first, to the benefit of hotel customers and hotels themselves and, second, to the benefit of flight customers.

476The Commission submits that the applicant’s line of argument is in part inadmissible and in part unfounded.

477As a preliminary point, it should be recalled that efficiencies may counteract the adverse effects of the merger on competition (paragraphs 76 and 77 of the Horizontal Merger Guidelines, which apply to non-horizontal mergers pursuant to paragraph 6 of the Non-Horizontal Merger Guidelines), with the result that the likely efficiencies demonstrated by the undertakings concerned must be taken into account in order to determine the impact of a concentration on competition (recital 29 of the EC Merger Regulation). To that end, efficiencies must, first, benefit consumers, second, be merger-specific and, third, be verifiable (paragraph 78 of the Horizontal Merger Guidelines).

478In that regard, it is apparent from paragraph 86 of the Horizontal Merger Guidelines that the purpose of the condition requiring efficiencies to be verifiable is to enable the Commission to ‘be reasonably certain that the efficiencies are likely to materialise, and be substantial enough to counteract a merger’s potential harm to consumers.’ In that regard, that paragraph states that, where reasonably possible, efficiencies and the resulting benefit to consumers should therefore be ‘quantified’ and that, ‘when the necessary data are not available to allow for a precise quantitative analysis, it must be possible to foresee a clearly identifiable positive impact on consumers, not a marginal one’.

479Lastly, it is apparent from the case-law that it is for the party which has notified a concentration, first, to demonstrate that the efficiencies are likely to counteract any adverse effects on competition that might otherwise result from the merger (judgments of 6 July 2010, Ryanair v Commission, T‑342/07, EU:T:2010:280, paragraph 427, and of 9 March 2015, Deutsche Börse v Commission, T‑175/12, not published, EU:T:2015:148, paragraphs 262 and 361) and, second, to put forward the arguments and evidence relating thereto during the administrative procedure, since the legality of a decision on the compatibility of a concentration with the internal market must be assessed in the light of the information available to the Commission when the decision was adopted (see, to that effect, judgment of 20 October 2021, Polskie Linie Lotnicze ‘LOT’ v Commission, T‑296/18, EU:T:2021:724, paragraph 55 and the case-law cited).

1. The alleged efficiencies benefiting hotel customers and hotels themselves

480In that regard, it should be noted that, in the application, the applicant relies on various efficiency gains which would arise to the benefit of hotel customers and hotels themselves, namely (i) an advantage for travellers related to the creation of a one-stop shop for the sale of flights and hotel rooms, which would enable them to purchase those travel services on the same platform, thereby improving the connected trip experience, (ii) an advantage for hotels as a result of an increase in the number of hotel room bookings on account of the applicant’s connected trip offering being more attractive and (iii) an advantage for hotel customers and hotels themselves on account of the fact that, post-transaction, some customers would book a hotel room on the applicant’s platform rather than on the platforms of rival hotel OTAs and the fact that the applicant is [confidential] than those OTAs.

481The Commission disputes the admissibility of the arguments raised by the applicant concerning efficiencies, in that they were not raised during the administrative procedure.

482First, as regards the efficiency gain related to the one-stop shop, the applicant refers, in paragraph 241 of the application, to an expert opinion dated 4 December 2023. In so far as that document, having regard to its date, post-dates the contested decision and could not be examined by the Commission during the administrative procedure, it is necessary to reject the applicant’s arguments in so far as they are based on that expert opinion.

483Moreover, in paragraph 128 of the reply, the applicant refers to paragraph 2.7 of the parties’ response to the decision under Article 6(1)(c) EC Merger Regulation, which describes the implications of the connected trip seeking to offer value and convenience to customers, some of whom might choose to purchase a number of travel services from a single provider offering greater flexibility, optionality or a better user experience.

484In the same vein, the applicant refers to paragraphs 1.19-1.20, 1.28(i) and 3.10 of the Form CO which it submitted in relation to the transaction, in which it is stated that the one-stop shop would enable consumers to benefit from better prices, a better service and better choice. Consequently, in so far as some consumers choose to purchase a number of travel services from the same provider, competition between the various suppliers offering various travel services would intensify, to the benefit of those consumers.

485In that regard, it should be noted that it is not apparent from the documents mentioned above that, during the administrative procedure, the applicant explicitly put forward the one-stop shop as an efficiency. Indeed, it is not mentioned in the sections of the parties’ response to the decision under Article 6(1)(c) EC Merger Regulation or the Form CO dealing with efficiencies. Rather, in those documents, the applicant simply provides general descriptions of certain consumer benefits, in particular the connected trip, without the Commission having clearly established that that was an efficiency as such being relied on.

486It follows that the applicant’s arguments based on the efficiency related to the creation of a one-stop shop are inadmissible.

487Second, as regards the other efficiencies which the transaction would provide to hotel customers and hotels themselves, the applicant refers, in paragraphs 243 to 245 of the application, to the document entitled ‘Analysis of accommodation pricing (RPD data)’, dated 20 January 2023, to Annex 2 to the parties’ response to the Commission’s request for information No 20, Figure 2, p. 19, to the document entitled ‘Efficiency benefits and claimed hotel OTA harm’, dated 4 December 2023, and to the document entitled ‘Merger-specific efficiencies’, dated 16 March 2023.

488In that regard, even though, in view of their dates, that evidence pre-dates the contested decision, with the exception of the document entitled ‘Efficiency benefits and claimed hotel OTA harm’, it must be stated that, in paragraph 282 of the defence, the Commission indicated, without being challenged by the applicant, that, during the administrative procedure, the applicant had neither made those claims nor substantiated the benefits of them, in particular the alleged benefits to hotel customers, on the one hand, and to hotels themselves, on the other, since those benefits had also not been quantified in the application.

489It follows that the applicant’s arguments based on various efficiencies for hotel customers and hotels themselves are inadmissible.

2. The alleged efficiencies benefiting flight customers

490In that regard, it should be noted, first, that the benefit to the applicant’s flight customers stems from the ‘Cournot effect’, as referred to in paragraph 117 of the Non-Horizontal Merger Guidelines, in a ‘zero-flights’ counterfactual scenario and from the internalisation of double mark-ups as referred to in paragraph 55 of those guidelines, in a Phase 2 Agreement counterfactual scenario and, second, that, subsequently, other flight OTAs would react by also lowering their prices, which would benefit their customers.

491In a ‘zero-flights’ counterfactual scenario, the transaction would create a link between the additional hotel OTA and flight OTA services, which would give the applicant a commercial incentive to reduce its prices compared to the prices of products supplied separately, with the result that passengers would benefit.

492In addition, in a Phase 2 Agreement counterfactual scenario, there would be a vertical relationship between the merging parties on account of the intervener providing flight OTA services to the applicant. In return, the applicant would pay the intervener a per-transaction fee, which it would incur as a marginal cost and which it would pass on to its flight customers. However, after the transaction, the applicant would no longer incur that cost, which would incentivise it to reduce its airline ticket prices.

493Subsequently, according to the applicant, other flight OTAs would react by also lowering their prices, which would benefit their customers.

494However, according to the applicant, the Commission dismissed the efficiencies relying on three reasons, which are erroneous, based on the fact that (i) the applicant would not pass on the benefits to consumers (recitals 1155 and 1165 of the contested decision), (ii) the efficiencies quantified by the applicant and the intervener factor in the behaviour of the applicant’s rivals (recital 1167 of the contested decision) and (iii) the alleged harm to hotels and hotel customers cannot be offset by the benefits conferred on air passengers.

(a)The first reason: the applicant would not pass on the benefits to flight customers

495The applicant submits, first, that the fact that consumers respond to prices and that profit-maximising undertakings therefore have an incentive to pass on marginal cost reductions are fundamental principles of economics and, second, that it submitted internal documents which referred precisely to the expectation of lower flight prices post-transaction.

496In the view of the applicant, contrary to the Commission’s assertion that there is no evidence that the applicant would have an incentive to pass on efficiencies because ‘consumers [confidential] (recitals 1165 to 1173 of the contested decision), the flight OTA market is characterised by a high degree of customer price sensitivity which would create an incentive for flight OTAs to pass on cost reductions to consumers because such price reductions will have a significant impact on sales volumes.

497The applicant states that evidence of that sensitivity is available. First, according to data from Kayak, an MSS operated by the applicant, the price differential between the lowest and second-lowest price between November 2022 and October 2023 was only [confidential], which confirms the high price sensitivity and the strong incentive to pass on a price reduction of EUR [confidential], as there is considerable potential to increase flight sales by selling at the lowest price. Second, flight customers’ price-sensitivity is demonstrated by internal data from the intervener’s own MSS (flygresor.se) showing that 87.3% of clicks from an MSS select the lowest fare for their itinerary. Third, those dynamics are supported by the price elasticity analysis conducted by the applicant and the intervener.

498The Commission disputes those arguments.

499As a preliminary point, it should be recalled that under paragraph 77 of the Horizontal Merger Guidelines, the notifying party is required to prove that the efficiencies claimed are likely to ‘enhance the ability and incentive of the merged entity’ to pass on the benefits of the transaction to consumers.

500In that regard, it should be noted that the Commission found, in the contested decision, that the applicant’s internal documents contained no evidence demonstrating that it had actually considered passing on the cost savings to its customers (recital 1155 of the contested decision). Moreover, the discount on airline ticket prices, that is to say, the pass-on to end customers estimated by the applicant, is so low that it is unlikely that customers would regard it as significant or be guided by it in their purchasing choices (recitals 1165 and 1173 of the contested decision).

501Consequently, the Commission concluded that it was also unlikely that the applicant would have the incentive to pass on the alleged cost savings to customers, as a discount of that small size was unlikely appreciably to affect those customers’ behaviour and to increase sales of airline tickets by the applicant (see recital 1165 of the contested decision).

502It is clear that none of the arguments put forward by the applicant can call into question the Commission’s assessment.

503First, as regards the applicant’s argument that the transaction would lead to lower flight prices (see paragraph 496 above), it must be held that the Commission was entitled to conclude that the applicant had not proved that it intended to lower flight prices post-transaction.

504It should be noted that, in that context, the applicant simply refers to general remarks found in a presentation to its board of directors, entitled ‘Project Bahamas Board Materials’, dated 23 November 2021, in which reference is made to [confidential], and in Annex 8 to its response to the statement of objections, in which it is stated that it is apparent from its internal documents [confidential], but those references are not supported by evidence describing the level of those discounts.

505In addition, in response to the applicant’s argument that the flight OTA market is characterised by a high degree of consumer price sensitivity (see paragraph 497 above), the Commission submits, first, that, during the administrative procedure, the applicant did not provide any evidence in support of its claim that the price differential on Kayak between the lowest price and the second lowest price was EUR [confidential], which the applicant did not dispute in its written pleadings. That argument by the applicant is therefore inadmissible.

506Second, as the Commission correctly points out, proposals on which customers click when searching for flights on an MSS are not indicative of what they purchase, since they may click on a number of options before making their purchase decision. It follows that the fact that 87.3% of clicks from an MSS select the lowest fare does not make it possible to know whether customers would switch flight OTA in response to a price reduction of EUR [confidential].

507Third, it should be noted, as the Commission did, that the theoretical price elasticity analysis for flights drawn up by the applicant in the context of the administrative procedure and on which the applicant relies in that context is of lesser evidentiary value than the pre-existing internal document to which the Commission refers in recitals 1165 to 1166 of the contested decision, which reflects the applicant’s experience of customers’ actual behaviour and from which it is apparent that those customers [confidential].

508It therefore follows from all of those considerations that the Commission was correct in not taking into account the efficiencies relied on by the applicant, since it had not been established that the applicant would pass them on to consumers.

(b)The second reason: the efficiencies quantified by the parties take into consideration rivals’ responses

509In the second place, the applicant submits that the Commission erred in failing to consider rivals’ responses (recitals 1167 and 1174 of the contested decision), because efficiencies must ‘relate to the conduct of the merged entity’ and not to that of its competitors.

510Given that an analysis of potential anticompetitive unilateral effects typically accounts for rivals’ responses to the merged entity’s conduct, it would be inconsistent for the Commission to ignore that reasonably predictable conduct when it is put forward as a justification of a pro-competitive efficiency.

511The applicant states that contrary to the Commission’s suggestions in recitals 1167 and 1174 of the contested decision, that decision includes no analysis of rivals’ responses, whereas the applicant did provide evidence demonstrating rivals’ likely responses, in particular the level of competition on the flight OTA market and the importance of price sensitivity in that regard. In that context, the applicant conservatively estimated that competitors would reduce their prices by only [confidential]% compared with the reduction which would be made by the merged entity.

512The Commission disputes those arguments.

513In that regard, it is necessary to examine whether the alleged benefit to customers of other flight OTAs, relied on by the applicant as an efficiency, may be regarded as verifiable within the meaning of paragraphs 86 to 88 of the Horizontal Merger Guidelines, in view of the fact that it is for the notifying party to adduce precise and compelling evidence enabling, as far as possible, the expected efficiencies to be quantified (judgment of 23 February 2022, United Parcel Service v Commission, T‑834/17, EU:T:2022:84, paragraph 138).

514In that context, the applicant refers to a presentation of 23 November 2021, entitled ‘Project Bahamas Board Materials’ used [confidential]. It should be noted that it is not apparent from that presentation which ‘potential responses’ were envisaged, let alone whether there was evidence as to the likelihood of any ‘response’.

515In that regard, it must also be stated that, as the Commission correctly points out, the applicant did not quantify the savings for consumers, but simply indicated that the assumption that competitors would respond by lowering their prices by [confidential]% of the price reduction to be applied by the merged entity has been ‘conservatively’ estimated. In that regard, the applicant refers solely to a report dated 16 March 2023 prepared by its economic adviser which considers the ‘assumption of a [confidential]% rival response’ to be ‘reasonable’, in view of the ‘price sensitivity of flight consumers’ and the expectation that ‘rivals will face considerable competitive pressure to respond to a reduction in prices by [the applicant]’.

516In the absence of any evidence of a correlation between (i) that price sensitivity and that competitive pressure and (ii) the [confidential]%, the Commission did not err in finding that the applicant had not substantiated that its argument was verifiable within the meaning of the Horizontal Merger Guidelines.

517Accordingly, it must be concluded that the Commission was justified in concluding that rivals’ responses could not counteract the potentially harmful effects of the transaction.

(c)The third reason: the alleged harm to hotels and hotel customers cannot be offset by the benefits for flight customers

518The applicant states that, in the contested decision, the Commission itself concedes that there is a commonality between end-users of hotel OTAs and of flight OTAs (see footnote 1502 to the contested decision), which would be strengthened by the increase in cross-sales of hotel rooms to flight customers post-transaction. According to the applicant, if users are assumed to be harmed in one market but to benefit in another, it is illogical to suggest that the benefit is irrelevant because it does not occur in the market where the alleged harm arose, since from a consumer’s perspective what will matter is that they are better off overall.

519The Commission disputes those arguments.

520In that regard, it should be noted that paragraph 79 of the Horizontal Merger Guidelines provides that the efficiencies claimed should benefit consumers in those relevant markets where it is otherwise likely that competition concerns would occur.

521Similarly, according to case-law, the assessment of the claimed efficiencies is intended to ascertain, on the markets giving rise to a significant impediment to effective competition, whether those efficiencies are capable of counterbalancing the anticompetitive effects which the concentration is likely to produce (see, to that effect, judgment of 23 February 2022, United Parcel Service v Commission, T‑834/17, EU:T:2022:84, paragraph 273).

522It is true that it follows from the judgment of 11 September 2014, MasterCard and Others v Commission (C‑382/12 P, EU:C:2014:2201, paragraphs 237 and 242), that consideration of efficiencies outside the market in which the harm occurs could be accepted, provided, inter alia, that the group of consumers affected by the restriction and the group of consumers benefiting from the efficiencies are substantially the same.

523While the contested decision recognises commonality between end-users of hotel OTAs and of flight OTAs, and that the transaction may lead to an increase in cross-sales, as the applicant correctly submits, those circumstances do not mean that the customers active on the hotel OTA market and customers on the flight OTA market are substantially the same. On the contrary, according to the Commission, there is a limited commonality between the end-users of hotel OTAs and end-users of flight OTAs (footnote 1502 to the contested decision), whilst the applicant estimates that between only [confidential]% and [confidential]% of hotel customers also purchase tickets from flight OTAs (recitals 242 and 748 of the contested decision).

524In the light of the foregoing, it must be concluded that the Commission was justified in concluding that the alleged harm to hotels and to hotel customers cannot be offset by the benefits for flight customers, with the result that the third part of the third plea must be rejected. Since all the pleas in law put forward in the application have been rejected, the action must be dismissed in its entirety.

V. Costs

525Under Article 134(1) of the Rules of Procedure of the General Court, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party’s pleadings.

526Since the applicant has been unsuccessful, it must be ordered to bear its own costs and to pay those incurred by the Commission, in accordance with the form of order sought by the Commission.

527In addition, in accordance with Article 138(3) of the Rules of Procedure, the intervener must bear its own costs.

Operative part

On those grounds,THE GENERAL COURT (Tenth Chamber, sitting with five Judges)hereby:

1.Dismisses the action;

2.Orders Booking Holdings Inc. to bear its own costs and to pay those of the European Commission;

3.Orders Etraveli Group AB to bear its own costs.

Delivered in open court in Luxembourg on 9 September 2026.