TABLE OF CONTENTS

1.

Introduction

3

1.1.

Purpose of the Guidelines

3

1.2.

Scope and structure of the Guidelines

5

1.2.1.

Scope

5

1.2.2.

Structure

6

2.

General principles applicable to the assessment of dominance

7

2.1.

Introduction

7

2.2.

Single dominance

8

2.2.1.

Market position of the undertaking concerned and of its competitors

8

2.2.2.

Barriers to entry or expansion

10

2.2.3.

Countervailing buyer power

12

2.2.4.

After-markets

13

2.3.

Collective dominance

14

2.3.1.

Collective dominance based on structural, contractual, or other links between the undertakings concerned

14

2.3.2.

Collective dominance based on tacit coordination

15

2.3.2.1.

Reaching terms of coordination

15

2.3.2.2.

Ability to monitor adherence to the terms of coordination

16

2.3.2.3.

Existence of a credible deterrence mechanism

16

2.3.2.4.

External stability – lack of constraints exerted by actual or potential competitors and lack of countervailing power by customers

16

3.

General principles to determine if conduct by a dominant undertaking distorts effective competition

16

3.1.

Introduction

16

3.2.

Conduct departing from competition on the merits

19

3.2.1.

The concept of departure from competition on the merits

19

3.2.2.

Relevant factors to establish that conduct departs from competition on the merits

19

3.3.

Capability to produce exclusionary effects

21

3.3.1.

Notion of exclusionary effects

21

3.3.2.

Type of exclusionary effects analysis to be applied

22

3.3.2.1.

Pricing conduct

23

3.3.2.2.

Non-pricing conduct

24

3.3.2.3.

Multi-faceted conduct

25

3.3.3.

Causation

25

3.3.4.

Elements that may be relevant to the assessment of a conduct’s capability to produce exclusionary effects

27

4.

Principles to determine whether specific types of conduct distort effective competition

29

4.1.

Introduction

29

4.2.

Predatory pricing

29

4.3.

Margin squeeze

33

4.4.

Rebates that are not conditional on exclusivity

36

4.5.

Exclusive dealing

39

4.6.

Tying and bundling

43

4.7.

Access restrictions

47

4.8.

Refusal to supply

48

4.9.

More favourable treatment of the dominant undertaking’s own products (so-called self-preferencing)

50

4.10.

Conduct that is by its very nature harmful to competition

52

5.

General principles applicable to the assessment of objective justifications

53

5.1.

Introduction

53

5.2.

Objective necessity defence

54

5.3.

Efficiency defence

56

5.3.1.

General principles

56

5.3.2.

Verifiable efficiencies

58

5.3.3.

Fair share for consumers

60

5.3.4.

Indispensability of the dominant undertaking’s conduct

61

5.3.5.

No elimination of effective competition

62

6.

Final remarks

63

1.   INTRODUCTION

1.1.   Purpose of the Guidelines

1.

The European Union rules on competition protect genuine, undistorted and effective competition in the internal market (‘effective competition’). Effective competition drives market players to deliver the best market outcomes for consumers in terms of product (1) prices, choice, quality and innovation. It ensures that markets remain open and dynamic and provides a level playing field for innovative players, including start-ups and small and medium-sized enterprises (‘SMEs’) so they can seize new opportunities and challenge established players. It also spurs innovation and ensures that resources are efficiently allocated, thereby contributing to sustainable development and making it possible to have strong and diversified supply chains, all of which contributes to the European Union’s resilience and long-term prosperity.

2.

Article 102 of the Treaty on the Functioning of the European Union (‘TFEU’) does not prevent an undertaking from acquiring a dominant position on a given market based on its own merits, in particular its skills and abilities (2). However, that provision seeks to prevent competition from being distorted to the detriment of the public interest, other market players, and ultimately consumers (3). The competitive harm caused by dominant undertakings’ abusive conduct can take various forms, such as higher prices, a deterioration in the quality of products (4), a reduction in innovation (5) or a limitation of consumer choice (6).

3.

It is to address such forms of competitive harm that Article 102 TFEU places a special responsibility on dominant undertakings not to distort effective competition (7).

4.

In view of growing market concentration in various industries and of the digitalisation of the European Union economy, which makes strong network effects (8) and ‘winner-takes-all’ dynamics (9) increasingly common, it is important that Article 102 TFEU is applied vigorously and effectively (10). It is equally important that Article 102 TFEU is applied in a predictable and transparent manner so that companies can operate freely in the internal market, within the limits laid down in European Union legislation, considering also the decentralised enforcement of Article 102 TFEU (11).

5.

Pursuant to the case-law of the European Union courts (12), Article 102 TFEU applies to all practices by dominant undertakings that may directly or indirectly harm the welfare of intermediate and final consumers (13), including practices that may harm consumers by undermining an effective structure of competition (14).

6.

In particular, dominant undertakings can harm the welfare of consumers by hindering, through recourse to methods different from those governing normal competition, the maintenance of the degree of competition existing in a market or the growth of that competition (15). Such behaviour, if not objectively justified, is referred to as an ‘exclusionary abuse’ (16).

7.

Against this background, these Guidelines set out principles and operational guidance to assess whether conduct by dominant undertakings constitutes an exclusionary abuse under Article 102 TFEU, in the light of the case-law of the European Union courts and the Commission’s own decisional practice based on the case-law. The Commission seeks thereby to enhance legal certainty and help undertakings self-assess whether their conduct constitutes an exclusionary abuse.

8.

In general, these Guidelines outline the current state of the case law of the European Union Courts. However, the Commission also seeks to explain its position with regard to issues that have not been dealt with in the case law, or that are subject to interpretation. Such explanations are without prejudice to the interpretation that the European Union Courts may give to those provisions in the future. The Commission will further develop its practice taking into account developments in the case-law (17), economic theory, the evolution of markets and competition dynamics. Although not binding on them, these Guidelines may also guide national courts, without prejudice to their power to refer any issue of interpretation for a preliminary ruling under Article 267 TFEU, and the national competition authorities of the Member States in their application of Article 102 TFEU.

1.2.   Scope and structure of the Guidelines

1.2.1.   Scope

9.

The general principles set out in these Guidelines need to be applied to the particular facts and circumstances of each case. The practices listed in the text of Article 102 TFEU do not exhaust the methods of abusing a dominant position prohibited by European Union law (18). Given the large number of possible types of exclusionary abuses by dominant undertakings and the wide range of market contexts in which they may occur, it is not possible to provide specific guidance for every possible scenario.

10.

While these Guidelines concern exclusionary abuses only, the principles applicable to the assessment of dominance (Section 2) and the assessment of objective justifications (Section 5) are also relevant for the assessment of other forms of abusive conduct, including exploitative conduct by dominant undertakings that harms consumers directly through the imposition of unfair trading conditions or excessive prices. For the avoidance of doubt, the same conduct by a dominant undertaking may have both exclusionary and exploitative effects.

11.

These Guidelines are without prejudice to the application of other provisions of European Union competition law to the same facts, in particular Article 101 TFEU and the rules on its application (19).

12.

Article 102 TFEU may also apply to conduct of dominant undertakings that falls within the scope of other European Union or national regulations (20). The fact that the conduct of a dominant undertaking has been found to have infringed other legislation does not preclude the possibility that, under certain conditions, the same undertaking may be sanctioned for an infringement of Article 102 TFEU for the same conduct (21). Similarly, the fact that an undertaking’s conduct has been found to comply with other legislation does not preclude the possibility that the same undertaking may be sanctioned for infringing Article 102 TFEU through that conduct (22).

1.2.2.   Structure

13.

To determine whether an undertaking has infringed Article 102 TFEU, it is necessary to carry out the following steps. First, it is necessary to assess whether the undertaking concerned holds a dominant position in the relevant market(s). Second, it is necessary to assess whether the conduct of the dominant undertaking distorts effective competition. Third, it may be necessary to assess whether the conduct is objectively justified.

14.

Accordingly, these Guidelines are structured as follows: — Section 2 describes the general principles applicable to the assessment of dominance. — Section 3 describes the general principles used to determine if conduct by dominant undertakings distorts effective competition. — Section 4 describes the principles used to determine whether specific types of conduct by dominant undertakings distort effective competition. — Section 5 describes the general principles applicable to the assessment of objective justifications, which under certain conditions may justify or counterbalance the effects of conduct that distorts effective competition.

15.

Finally, for completeness, it should be borne in mind that conduct not capable of appreciably affecting trade between Member States falls outside the scope of Article 102 TFEU. The Commission has provided guidance on the assessment of effect on trade in the Guidelines on the effect on trade concept (23).

2.   GENERAL PRINCIPLES APPLICABLE TO THE ASSESSMENT OF DOMINANCE

2.1.   Introduction

16.

A dominant position relates to a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market (24), by giving it the power to behave to an appreciable extent independently of its competitors, of its customers and ultimately of consumers (25). Both suppliers and buyers of a product can hold a dominant position. Article 102 TFEU applies not only to undertakings that hold a dominant position in relation to their customers (i.e. on the supplier side), but also to those that hold a dominant position in relation to their suppliers (i.e. on the buyer side) (26).

17.

The existence of competition on a particular market does not preclude a finding of dominance (27). The fact that a certain degree of competition may exist on a market is a relevant but not a decisive factor for determining whether a dominant position exists (28).

18.

To establish dominance, it is, as a general rule, a prerequisite to define the relevant market (29). The Market Definition Notice provides guidance on the rules, criteria and evidence that the Commission uses when defining markets (30). Market definition is a tool to structure and facilitate the competitive assessment, notably in cases where it is relevant to assess market power. It involves identifying in a systematic way the effective and immediate competitive constraints exerted on the undertakings concerned when they offer products in a certain area. When defining the relevant market, the Commission may need to take into account that the undertakings concerned already exert market power (31). The definition of the relevant market and the assessment of whether the undertakings concerned hold a dominant position within that market are therefore interrelated. The approach to market definition may also need to be adapted depending on the competitive concern under consideration (32).

19.

Dominance and market power (33) are interrelated concepts, but they are not synonyms. Dominance is a binary concept, meaning that undertakings either are or are not dominant in a given market at a given point in time. By contrast, market power exists on a spectrum, meaning that undertakings (both dominant and non-dominant) can have varying degrees of market power. The assessment of dominance is carried out based on the factors set out in Sections 2.2 and 2.3. Once dominance has been established, the exact degree of market power of the dominant undertaking is not decisive to determine the applicability of Article 102 TFEU. The degree of market power may however be relevant, among other factors, for the purpose of analysing whether the conduct of the undertaking concerned is capable of having exclusionary effects (34).

20.

A dominant position may be held by one undertaking (single dominance) or by two or more undertakings (collective dominance), as explained below.

2.2.   Single dominance

21.

Single dominance concerns a situation where a single undertaking has the power to behave to an appreciable extent independently of its competitors, its customers and ultimately of consumers on the relevant market (35).

22.

In general, the existence of a dominant position derives from a combination of several factors which, taken separately, are not necessarily decisive (36). The following sub-sections illustrate some of these factors, which are neither exhaustive nor cumulative (37). Other factors may be relevant for the assessment of dominance, depending on the specific circumstances of each case (38).

2.2.1.   Market position of the undertaking concerned and of its competitors

23.

The analysis of the market position of the undertaking concerned and of its competitors during the time period considered (39) provides insights into the constraints that the undertaking faces from actual competition in the relevant product and geographic market.

24.

One important factor is the market share of the undertaking concerned. The existence of a very large market share over a sustained period of time (40) is by itself – save in exceptional circumstances – evidence of the existence of a dominant position (41). This is the case in particular where an undertaking holds a market share of 50 % or more (42). Dominance may also be found in cases where an undertaking has a market share below 50 % (43). In such cases, factors other than the market share of the undertaking concerned, such as the strength and number of its competitors are particularly relevant (44). Based on the Commission’s experience, dominance is generally unlikely if the undertaking concerned holds a market share below 40 %. However, dominance can still be found in case of market shares below 40 %, for example, where customers are generally dependent on the undertaking concerned or where competing undertakings face serious capacity limitations (45).

25.

The Commission calculates market shares based on the indicator or indicators that best reflect the market power of the undertakings in the relevant market and over a reference period that reflects the characteristics of the relevant market (46). For instance, the value or volume of sales or purchases can provide useful information in this respect, but, depending on the circumstances of the case, other indicators may better reflect the market power of undertakings (47).

26.

It is also important to compare the market share of the undertaking concerned and the market shares of its competitors (48). For example, where the undertaking concerned holds a market share that is much larger than that of its competitors, this is a relevant factor for the assessment of whether the undertaking concerned holds a dominant position (49).

27.

Market shares should be interpreted in light of the relevant market conditions, the dynamics of the market and the extent to which products are differentiated. In particular, it is often appropriate to take into account the evolution of market shares over time (50). In fast-growing markets with short innovation cycles, high market shares may be a less useful indicator of dominance because those shares may turn out to be ephemeral (51). However, high market shares that remain stable over time may still be a reliable indicator of dominance even in this type of markets (52).

28.

In markets where there are frequent and significant investments in research and development (‘R&D’), the level of R&D expenditure or the number of patents or patent citations may be used as metrics to assess the market position of the undertaking concerned (53).

2.2.2.   Barriers to entry or expansion

29.

The second relevant factor for establishing dominance is the existence of barriers to market entry or expansion that hinder potential competitors from gaining access to the market or that hinder actual competitors from expanding their activities on the market (54). Easy entry into or expansion in a market limits the ability of an undertaking in that market to behave independently. For example, setting prices above the competitive level would attract new entry or expansion by rivals. Conversely, the existence of barriers to entry or expansion increases the ability of the undertaking concerned to behave independently and exert market power.

30.

Barriers to entry or expansion may exist due to various factors. Legal and regulatory barriers may include, for example, tariffs or quotas, planning regulations (55), licensing and authorisation requirements (56), statutory monopolies (57) and intellectual property rights (58). Other barriers to entry or expansion (59) may derive from certain advantages, such as: (i) an established distribution and sales network (60), (ii) economies of scale (61) or scope (62), (iii) vertical integration and exclusive or preferential access to inputs or customers (63), (iv) access to critical raw materials (64), (v) the inertia of customers or other market players (65), (vi) brand image and brand effects (66), (vii) the existence of a first mover advantage (67), (viii) talent recruitment and retention (68), and (ix) capital strength and the ability to acquire competitors and innovators (69).

31.

Data-driven advantages (70) may also create barriers to entry and expansion. These advantages may relate, for example, to access to unique or non-replicable data, to data accumulation and economies of scale and scope relating to data, such as data-driven network effects. Data-driven advantages may also arise out of the capability to process and utilize data effectively. This can, for example, be relevant in the context of the development of artificial intelligence, where access to large, high-quality datasets and the computational power to analyse them are crucial.

32.

Other factors that may create barriers to entry or expansion are significant upfront investment requirements and high sunk costs (71), lock-in effects (72), as well as other costs and impediments faced by customers when switching to a rival (73), including behavioural biases (74).

33.

For instance, in the online platform economy and other industries such as payment systems, network effects can create barriers to entry or expansion, as they may make it more difficult for competitors to attract a critical mass of users. In the case of direct network effects (75), the users’ willingness to adopt a new product depends on the willingness of other users of the same group to do the same. In the case of indirect network effects, the willingness of one group of users to adopt a new product depends on the willingness of another group of users to do the same. This can put a market entrant in the difficult position of having to attract a sufficient number of users in one or both group(s) (76). Entry barriers resulting from network effects may be higher when consumers single-home than when they multi-home, since a new entrant must persuade users to switch entirely to a new platform (77).

34.

Network effects can also occur in the context of digital ecosystems of interlinked products, services or platforms (78). In addition to network effects, such ecosystems may also feature other barriers to entry or expansion resulting from lock-in of users, high switching costs or data-driven advantages that may relate to multiple products (79). Rival entry or expansion may therefore be particularly difficult when an undertaking has gained one or more dominant positions in markets that are part of an ecosystem.

35.

Where the undertaking concerned holds a high market share over a prolonged period, this itself may indicate the existence of barriers to entry or expansion (80).

2.2.3.   Countervailing buyer power

36.

Competitive constraints may be exerted not only by actual or potential competitors of the undertaking concerned, but also by customers with countervailing buyer power (81). Countervailing buyer power can prevent even an undertaking with a high market share from acting to an appreciable extent independently of its customers (82).

37.

Buyer power may result from the customers’ size or their commercial significance for the undertaking concerned. Countervailing buyer power differs from general bargaining or negotiation power, which refers to the ability to favourably influence the outcome of a negotiation (83). Countervailing buyer power refers to the ability of customers to switch quickly to competing suppliers, to promote new entry or to vertically integrate, or at least their ability to credibly threaten to do so. If countervailing buyer power is sufficiently strong, it may deter or defeat an attempt by the undertaking concerned to exercise market power. However, buyer power that only shields a particular or limited segment of customers from the market power of the undertaking concerned may not be considered a sufficiently effective constraint to rule out dominance (84). Countervailing buyer power is less likely to be present when the undertaking concerned faces a large number of dispersed buyers (85), when switching away from the undertaking is subject to significant difficulties (86) or when buyers would risk losing significant business if they were to stop carrying the products of the undertaking concerned (87).

2.2.4.   After-markets

38.

After-markets are markets for the supply of products used in connection with what is typically a relatively durable product that has already been acquired. This latter product is referred to as the ‘primary product’ (and its market is called the ‘primary market’). The complementary products used in connection with the primary product are referred to as ‘secondary products’ (and their market is called the ‘after-market’) (88). An undertaking may be dominant on a primary market, on an after-market or on both.

39.

When assessing dominance in after-markets, it needs to be considered whether there are specific competitive constraints that can impact the assessment. This may notably be the case where the after-market consists of secondary products associated with a particular brand of the primary product due, among other things, to a lack of compatibility with other primary products (89). In that scenario, the producer of the primary product often holds a strong position on the market for the secondary product used in conjunction with its primary product. Despite that, in certain circumstances, effective competition on the primary market may discipline the possible market power of the producer of the primary product in the after-market. In such cases, it is therefore generally appropriate to analyse the interdependence of the primary market and the after-market (90). To reach the conclusion that the primary market and the after-market are interdependent and effective competition on the primary market disciplines market power on the after-market (and excludes dominance on the after-market), the following four conditions need to be cumulatively met (91): (a) Customers can make an informed choice, taking into account lifecycle-pricing, between the various suppliers in the primary market. (b) Customers are likely to make such an informed choice. (c) If an apparent policy of exploitation (92) (such as a significant price increase or another form of deterioration in the conditions of supply) were to be pursued in the after-market, a sufficient number of customers would adapt their purchasing behaviour in the primary market. (d) Customers would adapt their purchasing behaviour within a reasonable amount of time.

40.

If any of the four conditions is not met, it follows that competition in the primary market does not sufficiently constrain market power in the after-market. However, this does not necessarily mean that the producer of the primary product is dominant in the after-market. Whether dominance can be established in the after-market in such a scenario requires an assessment based on the other factors set out in Section 2.2.

2.3.   Collective dominance

41.

Collective dominance refers to a situation in which two or more legally independent undertakings present themselves or act together from an economic point of view on a particular market and are able to behave to an appreciable extent independently of competitors, customers and ultimately of consumers (93).

42.

Collective dominance does not necessarily require that competition between the undertakings concerned be completely eliminated or that the undertakings concerned adopt identical conduct on the market in all respects (94). It follows that the alleged abuse does not necessarily need to be carried out by all the undertakings holding a dominant position. It is sufficient that the conduct amounting to an abuse can be identified as one of the manifestations of such a joint dominant position (95).

43.

To establish collective dominance, it is necessary to examine the economic links or factors giving rise to a connection between the undertakings concerned that enable them to act together and adopt a common policy independently of their competitors, their customers and consumers (96).

44.

The existence of an agreement or of structural links between undertakings is not indispensable for establishing collective dominance (97). Collective dominance may also arise from tacit coordination in oligopolistic markets.

2.3.1.   Collective dominance based on structural, contractual, or other links between the undertakings concerned

45.

Collective dominance may arise because of structural, contractual, or other links between the undertakings concerned that enable them to present themselves or act together on a particular market as a collective entity vis-à-vis competitors, customers and ultimately of consumers (98).

46.

The required connection between the undertakings concerned may result from the nature and terms of an agreement between them or from the implementation of such an agreement, or it may result from structural or other links (such as personal ties), provided that those links lead the undertakings to present themselves or act together as a collective entity (99). This could be the case if undertakings have concluded cooperation agreements that lead them to coordinate their activities on the market, or if cross-shareholdings, participation in joint ventures, interlocking directorships (100) or other links in law lead the undertakings concerned to coordinate.

47.

Once it has been established that the undertakings concerned present themselves or act together as a collective entity, the assessment of dominance is based on essentially the same factors as the assessment of single dominance (101). Collective dominance based on structural or other links between the undertakings concerned can exist both horizontally between undertakings active at the same level of the supply or distribution chain or vertically between undertakings at different levels (102).

2.3.2.   Collective dominance based on tacit coordination

48.

Collective dominance may also result from other connecting factors, requiring an economic assessment of the structure and characteristics of the market in question (103) and the way the undertakings in question interact on the market. This means that, where the characteristics of the market facilitate the adoption of a common policy by the undertakings concerned, collective dominance can be established without there being an agreement or structural links (104). Article 102 TFEU has been enforced only rarely against undertakings that hold a collective dominant position based on tacit coordination, but enforcement may still be warranted in certain situations (105).

49.

The following sub-sections describe the four factors that are relevant to establish collective dominance on the basis of (tacit) coordination between the undertakings concerned (106). The analysis of these factors should not be undertaken mechanically or in an isolated and abstract manner but should take into account the overall mechanism of hypothetical tacit coordination, including on the allegedly abusive conduct itself (107).

2.3.2.1.   Reaching terms of coordination

50.

Tacit coordination is more likely to emerge if competitors can easily arrive at a common perception of how the coordination should work, and, in particular, of the parameters that lend themselves to being a focal point of the coordination (108). The less complex and more stable the economic environment, the easier it is for undertakings to reach a common understanding on the terms of coordination. Whether a market is conducive to coordination depends on the characteristics of the relevant market (109). Coordination may be easier between undertakings with a high degree of similarity in terms of, for example, market shares (110), cost structures, production capacities, product offering (for example, price or quality), market positioning (for example, degree of brand recognition) and degree of vertical integration (111). However, depending on the market characteristics, tacit coordination may also be possible under more complex economic conditions. For example, firms without such similarities may still reach an understanding on the terms of coordination where the mechanism of coordination involves following the market leader (112). Moreover, algorithms provided by online platforms or third parties may facilitate tacit coordination even in markets characterised by asymmetries between the undertakings concerned.

2.3.2.2.   Ability to monitor adherence to the terms of coordination

51.

The undertakings concerned must have the means to know whether the other undertaking(s) are adopting and maintaining the same strategy (113). Therefore, there must be a sufficient degree of observability for the undertakings concerned to be aware of changes in the market conduct of the other undertakings concerned. This ensures that any deviation from the focal point can be easily identified and lead to a reaction by the other undertaking(s) concerned. If markets are transparent, observability is high. However, the use of artificial intelligence and algorithms may also improve market observability for markets that would otherwise be less transparent.

2.3.2.3.   Existence of a credible deterrence mechanism

52.

In order for coordination to be sustainable over time, there must be an incentive for each undertaking concerned not to depart from such coordination (114). This may be the case where each undertaking is aware that any competitive action on its part designed to increase its market share would provoke similar actions by the other(s), such that it would derive no benefit from its initiative (115). Coordination cannot be sustainable unless the consequences of deviation are sufficiently credible and timely to convince the coordinating undertakings that it is in their best interest not to deviate. The mere existence of a credible deterrence mechanism suffices; there is no need to adduce proof of either a threat or the actual use of a retaliatory mechanism (116). Moreover, the prospect of reverting to normal competitive conditions may be a sufficient deterrent decreasing the incentives of the undertakings concerned to deviate from the terms of coordination (117).

2.3.2.4.   External stability – lack of constraints exerted by actual or potential competitors and lack of countervailing power by customers

53.

For coordination to be successful, the actions of actual or potential competitors or customers should not be able to jeopardise the intended outcome of the coordination (118). In other words, competitive pressure exerted by competitors or countervailing buyer power exerted by customers must not be of such a magnitude that it would render the coordination by the undertakings concerned unsuccessful.

3.   GENERAL PRINCIPLES TO DETERMINE IF CONDUCT BY A DOMINANT UNDERTAKING DISTORTS EFFECTIVE COMPETITION

3.1.   Introduction

54.

This section provides guidance on the general principles applicable to the assessment of whether conduct by dominant undertakings distorts effective competition (119).

55.

Holding a dominant position is not in itself unlawful and dominant undertakings are entitled to compete on the merits. However, dominant undertakings have a special responsibility not to engage, directly or through the actions of third parties (120), in conduct that distorts effective competition (121) and is thus capable of harming consumers. The concept of abuse is an objective one, so it is not necessary to show that the undertaking concerned intended to distort effective competition in order to establish an abuse of a dominant position (122).

56.

The assessment of whether conduct distorts effective competition must be made in the light of all the relevant factual circumstances, irrespective of whether they concern the conduct itself, the market (s) in question or the functioning of competition on that market or those markets (123).

57.

In the context of the identification of the factors that are relevant for the assessment, the Commission considers it helpful to set out, for explanatory purposes, the economic mechanism through which the conduct is capable of directly or indirectly harming consumers, in its specific economic and legal context. This is commonly referred to as a ‘theory of harm’. A theory of harm can be articulated by reference to a conduct’s potential impact on the competitive market structure and process, including by raising barriers to entry or expansion, and by reference to the conduct’s potential impact on any of the relevant parameters of competition, including increased prices, a deterioration in the quality of goods, reduced innovation and diminished consumer choice.

58.

The more a given conduct is considered generally likely to distort effective competition, the less case-specific evidence is required to prove that this is the case, and the other way around (124). In some circumstances, the demonstration of certain factual elements may lead to the application of a presumption that the conduct distorts effective competition. Such presumptions, whose strength and scope may vary depending on the specific analytical framework and the specific evidence adduced in its support, shift the evidentiary burden from the Commission onto the dominant undertaking. The dominant undertaking can rebut the application of the presumption by showing with sufficient evidence, during the administrative procedure, that the conduct does not distort effective competition in light of the factual circumstances of the case (125).

59.

At a conceptual level, conduct by a dominant undertaking distorts effective competition when it: (1) departs from competition on the merits (see Section 3.2 below) and (2) is capable of having exclusionary effects (see Section 3.3 below) (126). There is no legal requirement to demonstrate direct harm to consumers to establish that conduct distorts effective competition (127).

60.

The assessment of whether the conduct departs from competition on the merits is in principle different from the assessment of whether the conduct is capable of having exclusionary effects. However, the way and order in which each of these assessments is carried out in concrete cases will depend on the circumstances of the case, the conduct and the competition concerns at hand.

61.

For instance, in some cases it may be appropriate to start the analysis by assessing whether the conduct departs from competition on the merits, whereas in other cases it may be appropriate to start by assessing the capability of the conduct to have exclusionary effects. It may also be appropriate, in a specific case, to assess one of these limbs in more detail than the other. Finally, certain pieces of evidence and elements of analysis may be relevant for the assessment of both aspects (128).

62.

There are scenarios where establishing that conduct by a dominant undertaking distorts effective competition does not require analysing separately whether the conduct departs from competition on the merits and whether it is capable of having exclusionary effects.

63.

First, this is the case for certain types of conduct for which the general principles of Article 102 TFEU have been translated into specific analytical frameworks (129) that have been recognised by the European Union courts (130).

64.

Second, this is the case when it is established that the conduct is capable of excluding a hypothetical equally efficient competitor (131).

65.

Third, this is the case when it is established that the conduct is by its very nature harmful to competition, for instance, because it has no economic interest for the dominant undertaking other than that of restricting competition (132).

66.

Finally, where it is demonstrated that the conduct of a dominant undertaking distorts effective competition, it remains possible for the undertaking to show that the conduct is either objectively necessary to achieve a legitimate aim and proportionate to that aim, or that the distortion of effective competition is counterbalanced or even outweighed by advantages in terms of efficiency that also benefit consumers (133). The framework used for that assessment is set out in Section 5.

3.2.   Conduct departing from competition on the merits

3.2.1.   The concept of departure from competition on the merits

67.

The concept of competition on the merits refers to conduct within the scope of normal competition on the basis of the performance of economic operators (134). This notion covers, in principle, a competitive situation in which consumers benefit from lower prices, better quality and a wider choice of new or improved products (135), both in the short and long term. Thus, conduct that has the effect of broadening consumer choice by putting new goods on the market or by increasing the quantity or quality of the goods on offer generally falls within the scope of competition on the merits (136).

68.

Article 102 TFEU does not preclude the departure from the market or the marginalisation, as a result of competition on the merits, of competitors that are less efficient than the dominant undertaking and thus less attractive to consumers from the point of view of, among other things, price, choice, quality or innovation (137).

69.

However, dominant undertakings have a special responsibility not to allow their conduct to distort effective competition on the internal market. While an undertaking in a dominant position may protect its commercial interests by means of reasonable and proportionate steps (138), any defence of those interests must stay within the scope of competition on the merits (139).

3.2.2.   Relevant factors to establish that conduct departs from competition on the merits

70.

The assessment of whether conduct departs from competition on the merits (140) takes into account all relevant factual circumstances, including those that concern the conduct itself and those that concern its legal and economic context, such as the market(s) in question and the functioning of competition on that or those market(s) (141).

71.

As indicated in paragraphs 62-65 above, there are scenarios where it can be concluded that a given conduct distorts effective competition without the need to analyse in separate steps whether it departs from competition on the merits and it is capable of having exclusionary effects.

72.

In other cases where such analysis is needed, there is a strong indication that conduct departs from competition on the merits where it consists of: (i) providing misleading information, particularly to administrative or judicial authorities or other public bodies (142), (ii) misusing regulatory procedures or legal proceedings to prevent or make it more difficult for competitors to enter the market (143), or (iii) violating rules in other areas of law (for instance, relating to data protection or sustainability) and thereby negatively affecting a relevant parameter of competition, such as price, choice, quality or innovation (144).

73.

A finding that the dominant undertaking’s conduct relies on the use of resources or means inherent to the holding of the dominant position plays an important role in establishing that conduct departs from competition on the merits (145).

74.

Based on the case-law of the European Union courts and the Commission’s enforcement experience, other factors may be relevant for the purpose of assessing whether conduct departs from competition on the merits, depending on the specific circumstances of the case (146). These may include whether the dominant undertaking prevents consumers from exercising their choice based on the merits of the products, including product quality (147), or whether the dominant undertaking’s conduct consists of, or enables, biased or discriminatory treatment that favours itself over its competitors (148).

75.

The European Union courts have also highlighted that a dominant undertaking’s intention to compete on the merits, even if established, is not sufficient to prove the absence of an abuse (149). On the other hand, evidence of intent to restrict competition is a relevant element to establish that conduct departs from competition on the merits (150).

76.

Moreover, in view of its special responsibility a dominant undertaking may have to refrain from engaging in certain practices that may be permitted for undertakings that do not hold a dominant position. The mere circumstance that the conduct is also implemented by non-dominant undertakings in the market does not preclude a finding that it departs from competition on the merits, when implemented by a dominant undertaking (151).

77.

A dominant undertaking may submit evidence which either (a) seeks to show that its conduct did not depart from competition on the merits or (b) seeks to show that there are benefits resulting from its conduct which counterbalance, or even outweigh, its distortion of effective competition because it pursues a legitimate objective or brings about efficiencies.

78.

The first type of evidence under (a) is normally intended to seek to challenge the factors or evidence relied upon by the Commission in its findings that the conduct departed from competition on the merits. The second type of evidence under (b) is analysed as part of the assessment of the objective justifications (see Section 5 below) (152).

3.3.   Capability to produce exclusionary effects

79.

This section outlines the elements that are relevant to assess exclusionary effects under Article 102 TFEU. Conduct for which the European Union courts have developed a specific analytical framework is assessed by applying that framework (see Section 4), taking into account the elements and principles outlined in this section (153).

3.3.1.   Notion of exclusionary effects

80.

As a rule, to conclude that conduct distorts effective competition, it is necessary to demonstrate that the conduct is capable of having exclusionary effects (154).

81.

The notion of exclusionary effects covers a broad range of possible hindrances to the normal competitive process (155), including the total exclusion or marginalisation of competitors, an increase in barriers to entry or expansion (156), the hampering or elimination of effective market access or access to parts of a market (157), the imposition of constraints on the potential growth of competitors (158), the elimination or reduction of competitors’ ability or incentive to exert a competitive constraint on the dominant undertaking (159) and more generally making it more difficult for competitors to enter or remain on the market (160). Both static and dynamic elements are relevant in the assessment of exclusionary effects (161).

82.

Establishing that a conduct is capable of having exclusionary effects requires the effects in question to be more than hypothetical (162). This does not require proof that the conduct has produced actual exclusionary effects (163) or that such conduct would be profitable (164).

83.

Moreover, while Article 102 TFEU is not concerned with conduct whose actual or potential effects are insignificant (165), there is no de minimis threshold for the purposes of determining whether a conduct infringes Article 102 TFEU (166). Further, a finding that a conduct by a dominant undertaking is capable of having exclusionary effects cannot be called into question by proving that the remaining part of the market that is not affected by the conduct is still sufficient to accommodate a limited number of competitors (167).

3.3.2.   Type of exclusionary effects analysis to be applied

84.

The conduct and its legal and economic context determine the type of effects analysis to be applied in each specific case.

85.

In this regard, the European Union courts have held that, as a rule, it is necessary to show that the conduct of the dominant undertaking is capable of excluding or hampering the growth of equally efficient competitors (168). In practice, evidence of the capability to exclude a hypothetical equally efficient competitor has predominantly been considered relevant by the European Union courts in relation to pricing abuses (169).

86.

The European Union courts have also held that conduct may be characterised as abusive where it has been proven to have the actual or potential effect – or even the object – of impeding at an earlier stage competitors from even entering the market(s) in question, through the placing of obstacles to entry or the use of other blocking measures or other means different from those which govern competition on the merits, and, in so doing, preventing the growth of competition therein to the detriment of consumers, by limiting production, product or alternative service development or innovation (170). The European Union courts have more generally held that not every finding of an exclusionary abuse of dominance is subject to proof that the conduct concerned is capable of having exclusionary effects in relation to an equally efficient competitor (171).

87.

Against this background, the Commission considers that the analysis of exclusionary effects in pricing, non-pricing and multi-faceted conduct should follow the principles set out in the following sections.

3.3.2.1.   Pricing conduct

88.

In these Guidelines, the Commission considers the expression ‘pricing conduct’ as referring to unilateral conduct consisting of the application of pricing policies. As such, predatory pricing (Section 4.2), margin squeeze (Section 4.3) and conditional rebates (Section 4.4) are considered pricing conduct (172).

89.

The assessment of pricing conduct is typically based on determining whether the conduct is capable of excluding hypothetical equally efficient competitors (in which case it is not necessary to prove separately that the conduct departs from competition on the merits; see paragraph 64 above). A price-cost test is normally (173) applied to help determine whether this is the case, within the limits and conditions set out in Section 4 (174).

90.

In the interest of legal certainty, the price-cost test is typically carried out by reference to price and cost data from the dominant undertaking at the time of the abuse in question (175). Nonetheless, where relevant and appropriate, dynamic considerations, such as the need for equally efficient competitors in innovation-driven industries to recover product-related development costs, can be taken into account (176).

3.3.2.2.   Non-pricing conduct

91.

In these Guidelines, the Commission regards the expression ‘non-pricing conduct’ as referring to unilateral conduct that is different from pricing conduct.

92.

The assessment of non-pricing conduct may be based either on the conduct’s capability to exclude a hypothetical equally efficient competitor (in which case it is not necessary to prove separately that the conduct departs from competition on the merits, see paragraph 64 above), or on other types of exclusionary effects in relation to which the concept of a hypothetical equally efficient competitor is not relevant (177). In practice, in cases concerning a variety of non-pricing conduct, the European Union courts have not required evidence that the conduct was capable of excluding a hypothetical equally efficient competitor (178).

93.

In scenarios where the focus of the assessment is on the exclusion of a hypothetical equally efficient competitor, the notion of ‘efficiency’ may also relate to parameters of competition other than price, such as quality, innovation and choice (179). In general, a price-cost test will not be informative in such scenarios because it is difficult for such a test to capture these non-price parameters of competition (180). Other types of evidence, including qualitative evidence and actual market developments, will be relevant for the assessment of such exclusionary effects (181).

94.

In other scenarios, the concept of a hypothetical equally efficient competitor is not relevant for the analysis of exclusionary effects. Notably, conduct may impede entry by potentially competing undertakings at an earlier stage, thereby preventing the growth of competition in the market to the detriment of consumers (182). This may be the case for example in digital markets and ecosystems, where features such as innovation, access to data, multi-sidedness, user behaviour or network effects play a decisive role (183). In this context, it may be relevant to take into account dynamic considerations, for instance whether the conduct prevents competitors, actual or potential, from enjoying demand-related advantages, such as network and learning effects, and exerting a stronger competitive constraint in the future.

95.

There may also be other situations where the emergence of an equally efficient competitor would be practically impossible, for instance due to the dominant undertaking’s very large market share, the presence of significant barriers to entry or expansion, or the existence of regulatory constraints (184).

96.

In these settings, even a less efficient competitor may exert a genuine constraint on the dominant undertaking, thereby preventing it from raising prices or reducing output, quality, choice or innovation to the detriment of consumers (185).

3.3.2.3.   Multi-faceted conduct

97.

In the case of multi-faceted conduct (i.e. conduct involving a combination or succession of pricing and non-pricing behaviour), the type of analysis to be carried out will be determined on a case-by-case basis taking into account the particular legal and economic circumstances. For instance, if a conduct predominantly consists of non-pricing behaviour, it may not be necessary to apply a price-cost test to assess the price-based component of the conduct.

98.

Depending on the conduct in question, the Commission may apply a specific analytical framework to assess that conduct, if there is such a framework among those developed by the European Union courts that sufficiently captures the essence of the conduct, or it may apply the general principles set out above in the assessment of specific types of conduct, as appropriate (see Section 4).

3.3.3.   Causation

99.

The exclusionary effects must be attributable to the conduct in question (186). It is sufficient to establish that the conduct increases the likelihood of the exclusionary effects materialising on the market (187). Effects can be actual or potential, so the requirement of causation is not to be understood as meaning that the conduct needs to have caused actual effects (see paragraphs 82 above and 106(g) below). Moreover, the conduct does not need to be the sole cause of those exclusionary effects (188). Finally, the assessment of whether the conduct in question is capable of producing exclusionary effects does not require proof that the conduct is enabled by the dominant position (189).

100.

Conceptually, the analysis of the capability of the conduct to produce exclusionary effects entails a comparison of the situation where the conduct in question was implemented with the situation absent that conduct. Given that the hypothetical situation absent the conduct cannot be directly observed, the comparison is often done by comparing the market situation before the conduct was implemented with the market situation after the implementation of the conduct (190).

101.

In certain cases, it may be appropriate to use as a basis for the comparison an alternative hypothetical scenario where the conduct would be absent and where certain developments in the market, sufficiently foreseeable at the time of the conduct, are also taken into account. This may be appropriate, for example, in cases where there have been developments in the market that would probably have occurred independently of the actions of the dominant undertaking (191). However, developing reliable assumptions for an alternative hypothetical scenario, which take into account the various possible combinations of events that could have arisen absent the conduct, or would artificially isolate parts of a wider strategy, may be particularly challenging, if not impossible (192). As such, it may be sufficient to establish only one credible alternative hypothetical scenario (193).

102.

In any event, the Commission may rely on a range of evidence, without being required systematically to use any single methodology, in particular an alternative hypothetical scenario, to prove the existence of a causal link (194). A comparison with an alternative hypothetical scenario may not be required in particular where the conduct of the undertaking has made it very difficult or impossible to ascertain the objective causes of observed market developments (195).

103.

Assessing whether a conduct is capable of having exclusionary effects is based on the facts and circumstances existing at the time when the conduct was implemented (196). For instance, when the conduct removes the commercial uncertainty relating to the entry or expansion of competitors, the conduct in question needs to be assessed with reference to the uncertainty that existed at the time of the conduct’s implementation (197). Moreover, where it is established that a conduct is capable of having exclusionary effects (198), this cannot be called into question by the actual reaction of third parties (199), or by the alleged ability of competitors on the market concerned to mitigate such exclusionary effects (200).

3.3.4.   Elements that may be relevant to the assessment of a conduct’s capability to produce exclusionary effects

104.

While conduct may take place and produce exclusionary effects either on dominated markets or on non-dominated markets (201), the substantive legal standard to prove the exclusionary effects is the same (202). When assessing exclusionary effects in a dominated market, the fact that in that market competition is already weakened due to the very presence of the dominant undertaking is a factor that must be taken into account (203).

105.

The assessment of whether a conduct is capable of having exclusionary effects must be based on specific, concrete elements of analysis and evidence. That assessment must take into account all relevant facts and circumstances (204). Conduct may prove to be capable of producing exclusionary effects in certain specific circumstances when that might not be the case were that conduct to have taken place in a different context. In particular, considering a factual element as a relevant contextual factor does not depend on whether that element amounts to a behaviour that in itself is characterised as abusive (205).

106.

The following factors may be generally relevant, while bearing in mind that the scope of analysis of the different factors, and their weight in the overall assessment, may be different depending on the type of case and its context, and accordingly this list should not be mechanically applied: (a) The market position of the dominant undertaking. In general, the greater the market power of the dominant undertaking, the more likely it is that its conduct is capable of having exclusionary effects (206). (b) The conditions on the relevant market. This notably includes the conditions of entry and expansion, such as the existence of economies of scale or scope and network effects (207). Where economies of scale are present, competitors are less likely to enter or stay in the market if the dominant undertaking holds a significant share of the relevant market. Similarly, the conduct may allow the dominant undertaking to ‘tip’ a market characterised by network effects in its favour, or to further entrench its position on such a market (208). (c) The position of the dominant undertaking’s competitors. This includes the importance of actual or potential competitors for the maintenance of effective competition. A specific competitor may play a significant competitive role, even if it holds only a small market share compared to other competitors or the dominant undertaking. Despite holding a small market share, a competitor may, for example: (i) be a close competitor of the dominant undertaking; (ii) be a particularly innovative competitor; or (iii) have in the past undercut prices or have the reputation of systematically undercutting prices. (d) The extent of the allegedly abusive conduct. In general, the higher the share of total sales in the relevant market affected by the conduct and the longer the duration of the conduct, the greater is the capability of the conduct to produce exclusionary effects (209). However, even conduct affecting a small share of total sales in the relevant market can be capable of having exclusionary effects, for instance where the customers or the market segment targeted by the conduct are of strategic importance for entry or expansion (see point (e) below) (210). (e) The position of customers or input suppliers. The dominant undertaking may apply the conduct only to selected customers or input suppliers who may be of particular importance for the entry or expansion of competitors (211). Those customers or suppliers: (i) may correspond to a particular means of distributing the product that would be suitable for a new entrant; (ii) may be situated in a geographic area that is well suited to new entry, or (iii) may be likely to influence the behaviour of other customers. In the case of customers, they may, for example, be the customers most able and likely to sponsor entry (or expansion) to the market by actual or potential competitors of the dominant undertaking. In the case of input suppliers, they may be the input suppliers most able and likely to sponsor entry or expansion by competitors of the dominant undertaking in a downstream market, or they may produce a variety of the product – or produce at a specific location – that is particularly suitable for a new entrant. (f) Evidence of an exclusionary strategy. Although the concept of abuse of dominance is an objective concept, for which it is generally not necessary to establish an intent to exclude competition (see paragraph 55 above) (212), evidence of such intent may still be relevant for the purposes of establishing an abuse (213). Such evidence may include, for example: (i) internal documents indicating a plan to eliminate or reduce competition; or (ii) concrete threats of exclusionary action (214). (g) Evidence relating to actual market developments: Actual developments in the relevant markets following the implementation of the conduct may provide evidence of the conduct’s capability to produce exclusionary effects, notably if the conduct has been in place for a sufficiently long period of time (215). Examples of such developments are (i) the market share of the dominant undertaking may have increased; (ii) actual competitors may have been marginalised or may have exited; or (iii) potential competitors may have tried to enter the market and failed (216). However, the absence of actual developments consistent with actual exclusionary effects cannot in itself disprove the capability of the conduct in question of having exclusionary effects. It may simply constitute indicia that the conduct was incapable of producing the alleged exclusionary effects (217). The dominant undertaking must supplement such indicia by evidence showing that the absence of actual effects was indeed the consequence of its conduct being unable to produce such effects (218). The absence of actual exclusionary effects could be due to a variety of causes, such as changes in the relevant market that are unrelated to the conduct, the fact that the undertaking in a dominant position was unable to fully implement the strategy underpinning the conduct (219), or the fact that third parties did not react as expected (220).

4.   PRINCIPLES TO DETERMINE WHETHER SPECIFIC TYPES OF CONDUCT DISTORT EFFECTIVE COMPETITION

4.1.   Introduction

107.

This section provides guidance on types of conduct for which the European Union courts have developed a specific analytical framework to establish whether the conduct in question distorts effective competition (predatory pricing, margin squeeze, exclusive dealing, tying and bundling, refusal to supply) (221) or have provided guidance on how to apply the general principles set out in Section 3 to specific scenarios (access restrictions, rebates that are not conditional on exclusivity, and more favourable treatment of the dominant undertaking’s products). Finally, this section provides guidance on conduct that is by its very nature harmful for competition.

4.2.   Predatory pricing

108.

Predatory pricing is a below-cost pricing practice by a dominant undertaking that aims at marginalising competitors or otherwise reducing competition.

109.

The dominant undertaking may engage in predatory pricing on the expectation to increase market power and, thereby, be in a position to benefit from the sacrifice after the end of the conduct (222). However, for the purposes of establishing an abuse, it is irrelevant whether the dominant undertaking can subsequently recoup the losses that it incurs as a result of its below-cost pricing (223).

110.

Predatory pricing can be applied to the whole relevant market in which the undertaking is dominant; it can also be applied to a particular market segment (224), for example with the aim of preventing actual or potential competitors from getting a solid foothold in the market (225), or it can be applied selectively to specific customers (226). In fact, pricing practices that target specific markets, market segments, or specific customers can be an effective means of predation from the point of view of the dominant undertaking. This is because such ‘selective predation’ may allow the dominant undertaking to limit the negative impact of the below-cost pricing on its profits, as compared with a general policy of low prices applied to a given output as a whole (227).

111.

Predatory pricing can also distort effective competition in a market other than the dominated market, which has not been strictly defined (228).

112.

The assessment of whether pricing conduct is predatory is carried out by means of a price-cost test, i.e. an analysis based on a comparison of the average prices applied and the average costs incurred by the dominant undertaking in relation to the products concerned (229).

113.

Based on the price-cost test, the following scenarios can be distinguished: (a) If prices are below average variable costs (‘AVC’), the pricing conduct is to be considered predatory. A dominant undertaking is deemed to pursue no economic objective other than eliminating its competitors by applying such prices, as each sale generates a loss (230). (b) If prices are below average total costs (‘ATC’) but above AVC, the pricing conduct is to be regarded as predatory if it is part of a plan to eliminate or reduce competition (231). The objective of that plan can be the elimination or marginalisation of one or more specific competitors (232) or the elimination or reduction of competition as such (233). To demonstrate the existence of such a plan, reference can be made to direct evidence (234), indirect evidence (235), or both, insofar as that evidence is sound and consistent (236). (c) If prices are above ATC, the pricing conduct cannot be regarded as predatory (237).

114.

Establishing that the dominant undertaking has charged predatory prices within the meaning of paragraph 113(a) or (b) is sufficient to demonstrate that the conduct distorts effective competition (238). In this respect, the Commission is not required to examine whether the share of the market covered by the contested pricing practice is of sufficient magnitude for the practice to distort competition (239).

The application of a price-cost test in predatory pricing cases

115.

The following considerations are relevant for the application of the price-cost test.

(a)   Cost benchmarks

116.

The cost benchmarks identified by the European Union courts for assessing whether pricing conduct is predatory are AVC and ATC (see paragraph 113 above), with AVC defined as costs that vary depending on the quantities produced, and ATC defined as the sum of an undertaking’s fixed and variable costs (240). To assess, in practical terms, whether pricing conduct is predatory in a specific case, it may be appropriate to use as cost benchmarks average avoidable costs (‘AAC’) or long-run average incremental costs (‘LRAIC’) respectively (241).

117.

AAC are the average of the costs that could have been avoided (242) or recovered (243) if the undertaking had not produced the discrete amount of output which is the subject of the conduct. In other words, AAC are the average of the costs that vary with the output that is related to the conduct over the relevant time horizon for the assessment. AAC can hence generally represent an appropriate benchmark for determining whether the prices of the dominant undertaking are such that the sales that are related to the conduct generate a loss.

118.

LRAIC are the average of all costs that are directly attributable to the production of the product in question during its lifecycle (i.e., all costs that are incremental to that product in the long run). In addition to AAC, LRAIC includes product-specific fixed and sunk costs incurred before the period in which the allegedly abusive conduct took place, including costs of capital. LRAIC are hence the appropriate benchmark for assessing whether the dominant undertaking’s prices are such that it earns a positive return on the product-specific assets. In contrast to ATC, LRAIC do not include costs that are not incremental to the production of the product in question, notably common costs. However, the mere fact that a certain cost category is recorded as common to several products in an undertaking’s accounts does not necessarily imply that these costs are not incremental. It may be necessary to assess whether such costs would have been incurred, partially or totally (244), if the undertaking had decided not to provide the product in question (245). To the extent that costs that are recorded as common are found to be incremental to the production of the product in question, they will be attributed to the LRAIC for that product (246).

(b)   Price and cost data to be considered and possibly adjusted

119.

The price-cost test is generally carried out on the basis of the price and cost data of the dominant undertaking itself, rather than the data of actual or potential competitors (247). This is in line with the principle of legal certainty to enable dominant undertakings to assess the lawfulness of their conduct (248).

120.

For this purpose, it is appropriate to consider the data in the dominant undertaking’s accounts (249). However, where these data are not available or are not sufficiently reliable (250), proxies and other pertinent information can be used, such as data from customers of the dominant undertaking that show the prices quoted and the discounts given (251). Depending on the circumstances of the case, it may be necessary to adjust the dominant undertaking’s data to calculate the effective prices paid (for example to calculate the effective prices net of discounts) or the costs incurred under the applicable cost standard (for example by spreading costs over a certain period in line with the principle of the depreciation of assets) (252). Furthermore, it may be appropriate to account for the opportunity costs of the dominant undertaking (253). In the case of two-sided markets, it may also be appropriate to include in the assessment revenues and costs incurred on both sides (254).

(c)   Scope and reference period

121.

When carrying out the price-cost test to assess the prices that are applied by the dominant undertaking to its products or vis-à-vis its customers, and that are suspected to be predatory, the relevant scope of the test and the reference period depend on the circumstances of the case. As regards the scope, this notably means that the price-cost test can be carried out for the products or customers together, or separately for each product or customer concerned (255).

122.

As regards the reference period over which the price-cost test is carried out, relevant factors to consider may include, for example, the timing of the competitive interaction and the price-setting intervals for each product to which the suspected predation relates (256).

4.3.   Margin squeeze

123.

Margin squeeze refers to a situation where an undertaking that is active in an upstream input market and in an associated downstream market (257) sets its upstream and downstream prices at a level that prevents downstream competitors that rely on that input from operating profitably on a lasting basis (258). By doing so, a dominant undertaking leverages its control over an input – through a combination of its upstream and downstream pricing – in a way that is capable of excluding rivals from the downstream market.

124.

For conduct to amount to a margin squeeze in breach of Article 102 TFEU, the undertaking concerned must be vertically integrated and dominant on the upstream market (259). This condition requires that the vertically integrated undertaking, regardless of the actual form of the integration (260), sells a product, for which it is dominant, as an input to undertakings in a downstream market and competes with those same undertakings on the downstream market. There is no need for the undertaking to also be dominant on the downstream market for an abusive margin squeeze to exist (261).

125.

In addition, the margin of an equally efficient competitor in the downstream market that relies on the dominant undertaking’s input must be negative (assuming it charges the same downstream price as the dominant undertaking). This is the case when the difference between the dominant undertaking’s downstream and upstream price (i.e. the ‘spread’) does not allow such an equally efficient competitor to cover the (incremental) costs of supplying the product in the downstream market (262), including costs of capital. The assessment of whether the margin is negative is done by means of a price-cost test, as set out in this section.

126.

When the spread between the dominant undertaking’s downstream and upstream price is negative, it is not necessary to consider the dominant undertaking’s downstream costs, as it is clear in that case that an equally efficient competitor using the dominant undertaking’s input will earn a negative margin and hence cannot operate profitably on a lasting basis on the downstream market.

127.

Establishing a margin squeeze does not require the upstream prices for the input to be in themselves excessive or the downstream prices to be in themselves predatory (263). Furthermore, it is irrelevant whether the dominant undertaking can subsequently recoup the losses that it incurred, if any, as a result of the margin squeeze (264).

128.

Given that a margin squeeze leads to an equally efficient competitor’s margin to be negative, it is probable that such conduct has exclusionary effects (265). Therefore, absent any evidence to the contrary (266), the Commission may conclude that the margin squeeze distorts effective competition.

129.

In this regard, it is not necessary to establish that the upstream input is indispensable for rivals to compete downstream (267). On the other hand, the more important the upstream input is to effectively compete downstream, the more likely it is that the conduct is capable of having exclusionary effects (268).

The application of a price-cost test in margin squeeze cases

130.

The price-cost test in the context of margin squeeze serves to determine whether the downstream arm of the dominant undertaking could operate profitably on the basis of the upstream price charged to its downstream competitors and the downstream price charged by the downstream arm of the dominant undertaking (269).

131.

Similarly to the application of the price-cost test to assess whether prices are predatory, to carry out a price-cost test in the context of a margin squeeze, it is necessary to determine the relevant cost benchmark, the price and cost data, as well as the scope and reference period. These parameters are determined on a case-by-case basis depending notably on the characteristics of the markets concerned, the competitive conditions and the other circumstances of the case.

132.

The price-cost test is usually based on the difference between, on the one hand, the effective downstream price charged by the dominant undertaking to customers and, on the other hand, the sum of the upstream price charged by the dominant undertaking to its downstream competitors and the LRAIC (270) of the dominant undertaking’s downstream arm.

133.

As a general rule, the price-cost test is based on the prices and costs of the dominant undertaking, to establish whether that undertaking would itself be able to offer its downstream products profitably if it had to pay its own upstream prices (271). In particular circumstances, where it is not possible to refer to the prices and costs of the dominant undertaking (e.g. if such data is not available or not sufficiently reliable), the prices and costs of competitors can be taken into account (272).

134.

The application of the price-cost test in margin squeeze cases does not imply that the competitors of the dominant undertaking would in fact be able to replicate its upstream assets. The price-cost test is applied from the perspective of a hypothetical equally efficient competitor, namely a competitor that uses the upstream product of the dominant undertaking, is in competition with the dominant undertaking on the downstream market, and whose costs on that market are the same as those of the dominant undertaking (273).

135.

The guidance set out in paragraphs 121 and 122 on the scope of the price-cost test (274) and the reference period is also relevant in the context of a margin squeeze.

4.4.   Rebates that are not conditional on exclusivity

136.

Conditional rebates refer to a system of rebates or other advantages (275), whether monetary or not, granted by a dominant undertaking to its customers to reward them for a particular form of purchasing behaviour.

137.

The rebate may be conditional on the purchase of a certain quantity of one product over a reference period (‘single product rebates’) or on the purchase of two or more different products together (‘multi-product rebates’). Multi-product rebates are also known as ‘mixed bundling’ or ‘bundled rebates’.

138.

This section focuses on single-product conditional rebates. However, it is also generally relevant for the assessment of multi-product rebates, with the necessary adaptations for the specific context.

139.

This section does not cover exclusivity rebates (i.e. rebates that are conditional on the purchase of all or most of a customer’s requirements of a product from the dominant undertaking). Exclusivity rebates are subject to the specific analytical framework set out in Section 4.5. This section is also without prejudice to the possibility that certain forms of conditional rebates may amount to a different type of abusive practice (276).

140.

Conditional rebates are a common business practice. Undertakings may offer such rebates to attract more demand and, as a result, they may stimulate demand and benefit consumers. However, when rebates are granted by a dominant undertaking, they may infringe Article 102 TFEU, depending on the circumstances of the case. In particular, conditional rebates may be used by the dominant undertaking to leverage its market power in a way that distorts effective competition.

141.

Whether conditional rebates by a dominant undertaking depart from competition on the merits and are capable of producing exclusionary effects in the specific circumstances of the case is assessed under the general principles set out in Section 3. However, the European Union courts have provided specific guidance on this or similar type of behaviour and the Commission considers that some specific elements of the case-law on exclusivity rebates and predatory pricing may also be used as point of reference by analogy and to the extent relevant.

142.

On this basis, the following factors may be typically relevant for the assessment of whether the conditional rebate scheme is capable of producing exclusionary effects (277): (a) The position of the dominant undertaking on the relevant market, namely the degree of market power held by that undertaking (278). (b) The extent of the conduct, in terms of duration and share of the market that is affected by the conduct. When market coverage is low, the conduct may nevertheless achieve exclusionary effects when the customers targeted by the conduct have strategic importance for entry or expansion (279). (c) The part of the demand of customers for which the dominant undertaking is an unavoidable trading partner (280). (d) The size of the rebate as a share of the total price, or the value of the non-price advantages, and the threshold triggering the rebate (281). Whether the conditions governing the rebate create uncertainty, due to, for example, lack of transparency, is also a relevant consideration (282). (e) The retroactivity of the rebates. Compared to rebates that apply only to volumes that exceed the threshold to qualify for the rebate (‘incremental rebates’), rebates that apply retroactively to all volumes purchased (‘retroactive rebates’) generally have a higher capability to produce exclusionary effects, as it may be less attractive for customers to switch small amounts of demand to an alternative supplier if they risk losing the whole retroactive rebate (283). (f) The individualised nature of the rebate. Rebates that are individualised for each customer (or type of customer) are in general more capable of having exclusionary effects, because they allow the dominant undertaking to target the rebate thresholds to each customer’s requirements, thereby increasing the loyalty effect (284). (g) The length of the reference period. In general terms, the longer the reference period, the higher the pressure on the customer to reach the purchase threshold required to qualify for the rebate or to avoid suffering the loss of the rebate. (h) That the conduct is part of a plan a to eliminate or reduce competition (285).

143.

As indicated in paragraph 64, in situations where it is demonstrated that a given conditional rebates scheme is capable of excluding a hypothetical equally efficient competitor, it is not necessary to demonstrate separately that such conduct departs from competition on the merits.

144.

In this regard, when the inducements offered by the dominant undertaking through the rebate are monetary or can easily and reliably be converted into a quantified monetary amount, the assessment of whether a conditional rebate scheme is capable of excluding a hypothetical equally efficient competitor is conducted, as a rule, by means of a price-cost test which compares the effective prices for the relevant units implied by the dominant undertaking’s rebate scheme to the costs for the relevant units (see sub-section below) (286). The price-cost test provides a quantitative assessment relating to some of the factors listed in paragraph 142 (287). In cases where a price-cost test is used, its results are interpreted as follows: (a) When the price-cost test shows that (i) the effective price is below AAC, or (ii) the effective price is below LRAIC but above AAC and the conduct forms part of a plan to eliminate or reduce competition, it is deemed that an equally efficient competitor cannot effectively compete for the customers to whom the conditional rebate applies. Subject to an assessment of the extent of the dominant position, coverage and duration, the Commission may conclude on this basis that the conditional rebate scheme distorts effective competition (288). (b) When the effective price is above LRAIC, the conditional rebate scheme is not considered to distort effective competition.

145.

However, a price-cost test is not mandatory in all cases (289). First, there are cases in which the use of a price-cost test is not feasible, for instance, where the inducements offered by the dominant undertaking are not monetary and cannot easily and reliably be converted into a quantified monetary amount (290). Second, the conduct can be found to be capable to produce exclusionary effects without reference to its capability to exclude a hypothetical equally efficient competitor in the instances set out in paragraphs 94 to 96, which can apply to conditional rebates mutatis mutandis. In these cases, the assessment of whether the conduct distorts effective competition will be carried out on the basis of the general principles (Section 3) and the other factors referred to in paragraph 142.

The application of a price-cost test to conditional rebates

146.

The application of a price-cost test to conditional rebates must be based on effective prices for the relevant units implied by the dominant undertaking’s rebate scheme and on the costs of the dominant undertaking for the relevant units (291). Effective prices, costs and relevant units depend on the specific facts of each case, including whether the rebate is incremental or retroactive.

147.

For incremental rebates, the relevant units to which the rebate applies are the units exceeding the threshold to qualify for the rebate. The effective prices are normally the prices arrived at after deducting the rebates that are applicable to the purchase of the relevant units. The costs relevant for the price-cost test must also be based on the relevant units.

148.

For retroactive rebates, the relevant units are the ‘contestable’ units (or contestable demand), i.e. the units that a customer is realistically able and willing to switch to a hypothetical equally efficient competitor, and which such a competitor could realistically compete for and win. The rest of the customer’s demand (the non-contestable demand) are the units that the customer would in any event be likely to obtain from the dominant undertaking. The effective price then represents the price that a competitor of the dominant undertaking would have to offer in order to compensate the customer for the loss of the rebate if the customer were to switch the contestable units of its demand away from the dominant undertaking (292). The costs relevant for the price-cost test must in this case be based on the contestable units.

149.

The assessment of a customer’s contestable demand is often subject to limitations in terms of data availability (293) and precision, which may lead to various degrees of uncertainty (294). When these limitations are so serious that the price-cost test cannot deliver reliable results, the assessment of whether the conduct distorts effective competition will be carried out on the basis of the general principles (Section 3) and the other factors referred to in paragraph 142.

150.

In the case of multi-product rebates, the price-cost test consists of comparing the effective price that customers pay for each of the dominant undertaking’s products in the bundle and the cost incurred by the dominant undertaking for including that product in the bundle. The relevant units in that case correspond to the units of the product in the bundle. Calculating the effective price implies allocating the entirety of the rebate to each of the products in the bundle, and then comparing the effective price of each product with the costs of the dominant undertaking for that product, in order to determine whether a hypothetical as efficient competitor could profitably offer the product on a stand-alone basis, i.e. unbundled.

4.5.   Exclusive dealing

151.

Exclusive dealing refers to various forms of obligation, promise (295), or commitment to purchase or sell all or most (296) of a customer or a supplier’s requirements of a product from/to the dominant undertaking (‘exclusivity obligations’) or incentive schemes, such as the grant of a rebate or other advantages (297), that are conditional on a customer or a supplier purchasing or selling all or most of their requirements of a product from/to the dominant undertaking (‘exclusivity rebates’).

152.

The Commission typically applies the same analytical framework to exclusive purchasing and supply obligations. For the sake of simplicity, in the rest of this section when reference is made to principles applicable to the analysis of exclusive purchase obligations or incentive schemes, the same principles apply, by analogy, to exclusive supply obligations or incentive schemes.

153.

Exclusive dealing can result from an explicit arrangement, irrespective of its form (298), whereby the customer must purchase all or most of its total requirements of a product from the dominant undertaking or where the customer receives incentives in return for so doing (so-called de jure exclusive dealing).

154.

Exclusive dealing can also result from arrangements that do not refer explicitly to exclusivity but entail requirements to deal with the dominant undertaking which, as a matter of fact, result in exclusivity (so-called de facto exclusive dealing) (299). This may be the case, for instance, for rebates that are conditional upon the achievement of a given volume-based threshold from the dominant undertaking, which in practice accounts for all or most of the customer’s requirements of a product, or for certain stocking requirements that have the same effect (300). To ensure predictability, the Commission will, in principle, only treat such arrangements as de facto exclusive dealing where there is evidence that the dominant undertaking intended to require or incentivise the customer to purchase all or most of its requirements of a product from the dominant undertaking. In the absence of such evidence, the Commission will assess these arrangements under the principles set out in Section 3 and, where relevant, Section 4.4 of these Guidelines.

155.

Exclusive dealing may enable a dominant undertaking to leverage its market power so as to restrict customers’ sources of supply and to deny or limit other producers’ access to the market (301). While exclusive dealing does not necessarily lead to exclusionary effects in every case, it has a high potential to generate such effects (302). As a result, exclusive dealing is subject to a presumption that it distorts effective competition (303).

156.

For the presumption to apply, it is necessary to demonstrate that the dominant undertaking has engaged in exclusive dealing as defined above. It is irrelevant for the purposes of the qualification as exclusive dealing whether exclusivity is imposed by the dominant undertaking or agreed at the request of the contractual counterpart (304). The dominant undertaking can submit evidence showing that the conduct does not amount to exclusive dealing, for instance because there are no de jure conditions requiring exclusivity or requirements to deal with the dominant company leading to de facto exclusivity.

157.

During the administrative procedure, the dominant undertaking can also submit evidence seeking to rebut the presumption (305). For example, the dominant undertaking may submit: (a) That the limited coverage and duration of the exclusivity arrangements are such that they are not capable of having exclusionary effects (306). (b) That the exclusionary effects of the exclusivity arrangement are purely hypothetical because the customer would have in any event purchased the volumes covered by the exclusivity arrangement from the dominant undertaking (307). Since the capability to restrict competition must be assessed by reference to the time when the conduct is implemented (see paragraph 103), this argument may only be in principle relevant where it is established that the customers of the dominant undertaking were aware at the time of entering into the exclusivity arrangement that no competitor would have been able to supply any part of the volumes covered by the exclusivity arrangement. (c) That the exclusionary effects of the exclusivity arrangements are purely hypothetical because rivals can compete effectively for the customer’s demand, as they see fit, in spite of the exclusivity arrangements. In the Commission’s experience, this would not be the case when a part of demand is non-contestable (308), for example: when a risk-averse customer is reluctant to switch the entirety of its demand to a different supplier; where a customer needs to source that part of its requirements from the dominant undertaking due to technical or commercial reasons; where the dominant undertaking’s brand is a ‘must-stock’ product; where switching is hindered by the existence of network effects; or where competitors are capacity constrained.

158.

A dominant undertaking may also submit a price-cost test showing that the dominant undertaking’s exclusivity rebates are not leading to an effective price below cost for the contestable units (309). However, in the Commission’s experience, such test is generally not relevant for a range of exclusive dealing scenarios, such as: (i) exclusivity obligations (310); (ii) exclusivity rebates that have a non-price component that cannot easily and reliably be converted into a quantified monetary amount (311); (iii) exclusivity rebates granting price advantages where the mechanism of exclusion of a hypothetical equally efficient competitor does not rely on pricing below costs (312); and (iv) in cases where a distortion of effective competition does not need to rely on the exclusion of a hypothetical equally efficient competitor (313).

159.

The submissions put forward by the dominant undertaking during the administrative procedure give rise to an obligation for the Commission to examine the arguments and evidence submitted by that undertaking (314). The Commission’s examination can lead to the following outcomes: (a) The arguments and evidence submitted by the dominant undertaking are insufficient to rebut the application of the presumption, for instance because the evidence has insufficient probative value, because the arguments refer to theoretical assumptions rather than the actual reality of the competitive conditions on the market (315), or because the arguments and evidence are not relevant to the case at hand (316). In that case, the presumption remains applicable and the Commission may therefore conclude that the conduct distorts effective competition; or (b) The arguments and evidence submitted by the dominant undertaking are relevant and of sufficient probative value to effectively rebut the application of the presumption, and therefore the Commission cannot conclude on the basis of the presumption that the conduct distorts effective competition.

160.

When the application of presumption has been effectively rebutted, the Commission can still establish that the exclusive dealing conduct distorts effective competition on the basis of an analysis of all the relevant aspects of the individual case. To this end, the Commission must carry out an overall assessment of all the relevant circumstances, beyond those relied upon by the dominant undertaking to effectively rebut the presumption.

161.

To this end, the Commission will typically examine (317): (a) The position of the dominant undertaking on the relevant market, namely the degree of market power held by the undertaking (318) and whether a given part of the demand is non-contestable (319). (b) The share of the market that is affected by the conduct: in general terms, the larger the share of the market covered by the exclusivity arrangements (in terms of the share of customers covered and in terms of the share each customer’s demand) (320), the more likely it is that the conduct distorts effective competition (321). However, even conduct that affects a small share of the market can distort effective competition. In particular, this may be the case where the customers or the market segment targeted by the conduct have strategic importance for entry or expansion (322). (c) The conditions and arrangements of the exclusivity arrangements, such as, (i) their duration (including, as appropriate, the conditions governing the renewal of the contract and the customers’ termination rights) (323), (ii) the penalties that may be imposed if the customer does not enter into or does not comply with the exclusivity conditions (324); (iii) in the case of exclusivity rebates, the amount or value of the incentives that are granted in return for exclusivity; for instance, the granting of strategic advantages or retroactive rebates may significantly strengthen the conduct’s exclusionary effects (325); and (iv) whether the agreement directly or indirectly references volumes purchased from competitors. In this context, exclusivity arrangements that refer to a share of purchases are more likely to distort effective competition than exclusivity arrangements that refer to a minimum volume purchased from the dominant undertaking, because the former not only create an incentive to buy more from the dominant undertaking but also necessarily imply buying less from its competitors (326). (d) Whether exclusive dealing would lead to the exclusion of a hypothetical equally efficient competitor (327), if relevant in the specific circumstances of the case (328). In these circumstances, the framework for the application of a price-cost test outlined in Section 4.4 can be used for the assessment of exclusivity rebates that are monetary or that can easily and reliably be converted into a quantified monetary amount (329).

4.6.   Tying and bundling

162.

Tying consists of offering a specific product (‘tying product’) only together with another product (‘tied product’) (330).

163.

Tying can take place on a technical (331) or a contractual basis (332). Technical tying occurs, for instance, when the tying product and the tied product are physically or technically integrated. Contractual tying occurs when the customer who purchases or uses the tying product is required to also acquire or use the tied product. The legal conditions that must be met to establish that tying by a dominant undertaking is abusive are the same whether the tying is technical or contractual.

164.

Bundling occurs when two products are offered and priced jointly as a single package. Pure bundling can be distinguished from mixed bundling. In the case of pure bundling, the products are only sold jointly in fixed proportions, and they are not available for purchase on a stand-alone basis (333). In the case of mixed bundling (or ‘multi-product rebates’), the products are also available separately, but the sum of the stand-alone prices is higher than the bundled price.

165.

Since pure bundling ties two products to each other, such that neither of them can be purchased alone, the assessment of pure bundling by a dominant undertaking is subject to the same legal requirements as tying and will not be discussed separately in this section. Mixed bundling by a dominant undertaking is examined using different criteria, as set out in Section 4.4.

166.

Tying and bundling are common practices which can provide customers with better products or more cost-effective offerings (334). However, when applied by a dominant undertaking, tying and bundling may in certain circumstances be abusive.

167.

Tying may be abusive in particular where a dominant undertaking leverages its dominance in the market of the tying product by denying customers the choice to obtain the tying product without the tied product, thereby restricting competition in the market of the tied product and departing from competition on the merits (335). Furthermore, through tying and bundling, the dominant undertaking may not only restrict competition for the tied product but also protect its dominance in the market of the tying product (336).

168.

Tying distorts effective competition where the following conditions are met (337): (a) the tying and tied products are two separate products; (b) the undertaking concerned is dominant in the market for the tying product (338); (c) the undertaking concerned does not give customers the choice to obtain the tying product without the tied product (a situation referred to as ‘coercion’), and (d) the tying conduct is capable of having exclusionary effects (339).

169.

Whether these conditions are satisfied in a given case is to be assessed based on the following considerations.

170.

For the purpose of establishing that the products affected by the tie are separate, it must be assessed whether there is separate customer demand for the tied product. This is the case if, in the absence of the tying or bundling, an appreciable number of customers would purchase or would have purchased the tied product independently, that is without buying it with the tying product from the same supplier (340).

171.

Several factors may be relevant for this assessment (341), including: (i) the nature and technical features of the products concerned; (ii) the facts observed in the market (342); (iii) the history of the development of the products concerned; and (iv) the commercial practice of the dominant undertaking (343).

172.

The fact that the tying and tied product belong to separate product markets (344) (respectively, the ‘tying market’ and the ‘tied market’) is an indication that the products are separate products (345). Complementary products can constitute separate products, as customers may wish to obtain both, but from different sources (346). The technical integration of one product into another does not mean that the two products can no longer be considered as separate for the purposes of tying (347). Similarly, even when tying two products is consistent with commercial usage or when there is a natural link between the two products, they may nonetheless be separate products (348).

173.

The condition that the undertaking must be dominant in the relevant tying market is to be assessed on the basis of the principles set out in Section 2 above. There is no condition that the undertaking must also be dominant in the tied market, although this is an element that can play a role in the assessment of whether the tie is capable of having exclusionary effects in that market (349).

174.

The condition of coercion captures situations where the dominant undertaking denies customers the choice to obtain the tying product without the tied product. Customers can either be the customers of the dominant undertaking (350), or intermediate parties, who pass on such coercion to final customers (351). Coercion can still exist where the party acquiring the tied product is not charged a separate price for that product (352). As coercion only requires that customers are not given the choice to obtain the tying product without the tied product, it can still exist even if the customer, after receiving the tied product, is not forced to use it, or is not prevented from purchasing or using the equivalent product supplied by a competitor of the dominant undertaking (353).

175.

Coercion can also exist if the dominant undertaking refuses de facto to offer the tying product without the tied product (354). This may be the case where the dominant undertaking offers the tying product physically or technically integrated with the tied product, so that it can only be used with the tied product, or where the quality or usability of the tying product is seriously degraded when it is not used with the tied product. The fact that coercion only concerns a certain segment of the customers of the tying product does not exclude that the condition of coercion is met (355).

176.

As regards the condition that the tie must be capable of having exclusionary effects, depending on the relevant factual and legal circumstances, those effects may concern the tied market only (356) or both the tied market and the tying market (357).

177.

The depth of analysis required to show that the tying conduct is capable of having exclusionary effects depends on the specific circumstances of the case (358). The specific characteristics of the markets and products at hand may, in certain circumstances, facilitate a conclusion that the conduct is capable of having exclusionary effects, without a very elaborate analysis (359). In other circumstances, a closer examination of the potential exclusionary effects of the tying may be warranted, for example, where (i) the tied product is available for free and (ii) it is easy to obtain alternatives to the tied product (360). As is the case for all types of conduct (see paragraph 106(g) above), when the conduct at stake has been in place for a long period, the Commission may have a more complete evidentiary basis to assess whether it is capable of having exclusionary effects (361). In those circumstances, an examination of market developments may provide evidence of the conduct’s capability to have exclusionary effects.

178.

In addition to the factors mentioned in Section 3.3.4, the following factors, which are neither cumulative nor exhaustive, may be relevant for assessing whether tying is capable of having exclusionary effects, while bearing in mind the specific context of the case: (a) Whether the tying confers a significant competitive advantage on the dominant undertaking in the tied market, resulting from coercion rather than the quality of the tied product, where that advantage is unlikely to be offset by competitors (362); (b) The significance of the link between the tying product and the tied product. This link may derive, for example, from the complementarity of the products (363) or from the share of customers in the tied market that also purchase the tying product (364); (c) Whether the dominant undertaking also enjoys dominance or market power in the tied market (365); (d) The presence of barriers to entry or expansion in the tied market (such as the need to achieve significant economies of scale or scope (366) or the presence of network effects, for instance in digital markets) (367); and (e) The degree of consumer inertia or bias in the tied market (368).

4.7.   Access restrictions

179.

In these Guidelines, the term ‘access restrictions’ refers to various scenarios where a dominant undertaking restricts access to an input (369), for instance by denying access to the input under commercially viable conditions or otherwise hindering or delaying it.

180.

Access restrictions are assessed by applying the general principles outlined in Section 3. Therefore, at a conceptual level, access restrictions distort effective competition where they: (1) depart from competition on the merits (see Section 3.2 above) and (2) are capable of having exclusionary effects (see Section 3.3 above) (370). If the access restrictions amount to a refusal to supply as outlined in Section 4.8, they are assessed under that specific analytical framework, which was developed to take due account of the need to protect an undertaking’s freedom of contract and its incentives to invest.

181.

Access restrictions can distort effective competition even if the input at stake is not indispensable. However, the importance of the input for the party seeking access will increase the likelihood that access restrictions will lead to exclusionary effects (371).

182.

The following are examples of access restrictions that, bearing in mind the specific context of the case, may be considered as abusive: (a) Where the commercial practices of the dominant undertaking lead to the disruption of supply of existing customers who are competing with it in a downstream market, if the customers abide by regular commercial practices and the orders placed by them are in no way out of the ordinary (372); (b) Where the dominant undertaking does not develop an input solely for its own use but in order to enable third-party undertakings to use it, and nevertheless does not provide access or restricts access to the input for one or more third parties (373); (c) Where the dominant undertaking fails to comply with obligations imposed on it (or prerogatives conferred on a third party) by legislation, regulation or contract, which limit its decision-making autonomy with regard to access to the input (374); or (d) Where the dominant undertaking applies unfair or unreasonable conditions in a way that is capable of degrading, delaying or otherwise hindering access to an input. This could be the case, for instance, where the dominant undertaking (375): (i) fails to set fair and transparent terms and conditions for access, including pricing rules, or (ii) fails to provide timely responses to requests for access or delays the start of access negotiations and the signing of contracts granting access to third parties.

4.8.   Refusal to supply

183.

One specific scenario concerning access to an input is subject to a specific analytical framework rather than to the general framework outlined in Section 3. This is the case in situations where a dominant undertaking owns an input (376) that it has established or developed (377) solely for its own use and, when access is requested by another undertaking (typically, an actual or potential competitor in a downstream market), refuses to give access (378). This scenario is commonly referred to as ‘refusal to supply’.

184.

In the specific scenario of a refusal to supply, an obligation to grant access to a third party, typically vis-à-vis an actual or potential competitor, directly impinges on the freedom of contract and the right to property of the dominant undertaking, given that the dominant undertaking had developed the input solely for its own use. Such obligation may also affect the incentives for competitors to develop competing inputs and the incentives for the dominant undertaking to invest in the development of inputs (379). Consequently, the European Union courts have imposed two stricter conditions for finding that such refusals to supply distort effective competition and, therefore, that an obligation to give access can be imposed (380).

185.

First, the input must be indispensable for the undertaking requesting access to compete in a downstream market. This condition is intended to determine whether the dominant undertaking controls access to the market concerned by virtue of that input and whether, therefore, it may be appropriate to compel the dominant undertaking to grant access to that input (381).

186.

An input is considered indispensable if there is no real or potential substitute to it (382). More specifically, this means that: (a) the input cannot be duplicated realistically and in a viable way, due to physical, technical, legal or economic reasons (383); (b) an equivalent input cannot be obtained from other sources; and (c) access to the input is necessary for the requesting firm to enter or remain viably on the market (384).

187.

Should there be a real or potential substitute to the input in question, even if access to such substitute were less advantageous for the requesting undertaking, the input cannot normally be considered as indispensable (385).

188.

Second, the refusal must be capable of eliminating all effective competition on the part of the requesting undertaking (386). In this regard, the fact that the requesting undertaking (or, a fortiori, other competitors of the dominant undertaking) retains a marginal presence in certain niches of the market is not sufficient to substantiate the existence of effective competition (387).

189.

Scenarios of refusal to supply outlined in this section can also concern intellectual property rights (388). As a general rule, such refusals may be regarded as distorting effective competition if, in addition to meeting the above conditions, they also limit technical development on the market. This is the case if the refusal prevents the requesting undertaking from producing new goods or services that are not offered by the dominant undertaking and for which there is a potential consumer demand (389), even if such goods or services are in competition with those of the dominant undertaking. It is not sufficient for the undertaking seeking access to limit itself essentially to duplicating the goods or services already offered by the owner of the intellectual property right (390).

4.9.   More favourable treatment of the dominant undertaking’s own products (so-called self-preferencing)

190.

Undertakings commonly treat their own products more favourably than the products of competitors. Conduct which implements such favourable treatment, mainly by means of non-pricing behaviour, is commonly referred to as self-preferencing (391).

191.

There is no general rule that self-preferencing is problematic from a competition standpoint, even when carried out by dominant undertakings. However, there are specific situations in which self-preferencing may distort effective competition (392).

192.

In particular, self-preferencing may distort effective competition where the dominant undertaking leverages its dominance in one market (the ‘leveraging market’) to give preferential treatment to its own products in another market (the ‘leveraged market’) (393), through means that depart from competition on the merits. Leveraging can be used to strengthen the dominant undertaking’s position in the leveraged market (offensive leveraging) or to protect its dominant position in the dominated leveraging market (defensive leveraging). In the Commission’s experience, self-preferencing is more likely to raise concerns in scenarios where the dominant undertaking is able to control, in the leveraging market, the conditions of access to the leveraged market(s) or to an ecosystem of products in several markets (394).

193.

In the Commission’s experience, self-preferencing may take place where two markets are in a vertical relationship or in a conglomerate relationship. This is the case, for example, where the two markets are part of the same intermediation value chain (395) or where the competitors of the dominant undertaking in the leveraged market are actual or potential users in the leveraging market (396).

194.

Self-preferencing may take various forms. For example, it may consist of the positioning or display of the leveraged product in the leveraging market (397), the steering of consumer behaviour or choice (398), or the manipulation of selection processes (such as auctions), to the advantage of the dominant undertaking’s own business activities. Self-preferencing can also consist of a combination or succession of different practices over time (399).

195.

Self-preferencing is assessed by applying the general principles outlined in Section 3. Thus, at a conceptual level, self-preferencing distorts effective competition only where: (1) it departs from competition on the merits (see Section 3.2 above) and (2) it is capable of having exclusionary effects (see Section 3.3 above) (400).

196.

In this regard, in addition to the factors mentioned in Section 3, the following elements may be relevant for the assessment (401): (a) The preferential treatment takes place in a leveraging market that constitutes an important source of business for competitors in the leveraged market, which competitors cannot effectively replace through other means, even if it is not ‘indispensable’ within the meaning of the analytical framework for establishing an abuse consisting of a refusal to supply (402); (b) The preferential treatment amounts to an unjustified difference in treatment (403). That is the case where the preferential treatment is likely to be contrary to the underlying business rationale of the dominant undertaking’s activities in the leveraging market (404), or where in the leveraging market there is an expectation of neutrality or openness in respect of the conduct of the dominant undertaking, which subsequently acts contrary to that expectation in a non-objective and non-transparent manner (405); or (c) The preferential treatment is likely to influence the behaviour of users, irrespective of the intrinsic qualities of the leveraged product (406).

4.10.   Conduct that is by its very nature harmful to competition

197.

Certain conduct by dominant undertakings is by its very nature harmful to competition, for instance because it has no economic interest for the dominant undertaking, other than that of restricting competition. Such conduct falls manifestly outside the scope of competition on the merits and is therefore very unlikely to produce consumer benefits (407).

198.

Examples of types of conduct that, depending on the specific circumstances, can be considered by their very nature harmful to competition are: (a) payments by the dominant undertaking to customers that are conditional upon the customers not selling, or postponing the sale of, products offered by a specific competitor of the dominant undertaking or products incorporating supplies by such a competitor (408); (b) the dominant undertaking agreeing with its distributors that they will swap a competing product with its own product under the threat of withdrawing discounts benefiting the distributors (409); (c) the dominant undertaking actively dismantling an infrastructure upon which a competitor relies to effectively compete on the market (410); or (d) in a context where the dominant undertaking holds both regulatory and commercial functions and finds itself in a position of conflict of interest, the establishment and implementation of rules and sanctions that, in the absence of substantive criteria and detailed procedural rules that ensure they are transparent, objective, precise and non-discriminatory, give the dominant undertaking the power to deny at its discretion any competing undertaking access to the market (411).

199.

When it is established that a conduct is by its very nature harmful to competition, the conduct is deemed as such to distort effective competition.

200.

During the administrative proceedings, the dominant undertaking can submit evidence seeking to demonstrate that its conduct is not by its very nature harmful to competition. The European Union courts have not ruled on whether a dominant undertaking can challenge a finding that its conduct is by its very nature harmful to competition by proving that the conduct is not capable of producing exclusionary effects in the specific circumstances of the case. Pending clarification by the European Union courts, the Commission considers that such a challenge may in any event only very exceptionally be successful. Were the European Union courts to consider, on the basis of a consistent interpretation of Articles 101 and 102 TFEU (412), that conduct that is by its very nature harmful to competition under Article 102 TFEU should be subject to the same evidentiary requirements as restrictions by object under Article 101 TFEU, the Commission will adapt its case practice accordingly (413).

201.

The dominant undertaking can seek to demonstrate that conduct that is by its very nature harmful to competition is justified on the basis of objective necessity or efficiencies. However, in light of the nature of the conduct that falls in this category, the Commission’s own experience indicates that such attempts are very unlikely to be successful.

5.   GENERAL PRINCIPLES APPLICABLE TO THE ASSESSMENT OF OBJECTIVE JUSTIFICATIONS

5.1.   Introduction

202.

Conduct that distorts effective competition may escape the prohibition laid down in Article 102 TFEU where it is objectively justified (414). To be objectively justified, the conduct must either be objectively necessary (‘objective necessity defence’) or produce efficiencies that counterbalance, or even outweigh, the negative effect of the conduct on competition (‘efficiency defence’) (415).

203.

An objective necessity defence is generally based on an overriding reason that can justify the implementation of a conduct that distorts effective competition. Such reason may relate to the need to protect the dominant undertaking’s legitimate interests, or, in certain circumstances, it may relate to the pursuit of a public health, safety or other public interest objective (see Section 5.2 below). An efficiency defence, on the other hand, is based on a claim that the conduct enables the dominant undertaking to compete more effectively on price, quality, innovation or choice and thereby produce benefits for consumers which counterbalance or outweigh the distortion of effective competition caused by the conduct (see Section 5.3 below).

204.

The burden of proof for an objective necessity or efficiency defence lies with the dominant undertaking (416), in particular since the dominant undertaking is naturally well placed to disclose the existence of the objective necessity or the efficiency and to demonstrate its relevance (417).

205.

Proving an objective necessity or efficiency defence requires convincing arguments and evidence (418). Vague, general and theoretical claims, or those which rely exclusively on the dominant undertaking’s own commercial interests, are not sufficient (419). Finally, it is not relevant for the assessment of an efficiency defence whether the claimed efficiencies were intended or, on the contrary, only accidental consequences of the conduct under examination (420).

206.

Once the dominant undertaking has put forward arguments and evidence in support of an objective justification, it is for the Commission to assess whether those arguments and evidence can prevail and, consequently, whether the justification relied upon can be accepted (421).

5.2.   Objective necessity defence

207.

An objective necessity defence involves demonstrating that the behaviour of the dominant undertaking is objectively necessary to achieve a legitimate aim (422).

208.

In particular, an objective necessity may stem from legitimate commercial considerations, for example, protection of the dominant undertaking against unfair competition (423), protection against the placing of customer orders that are out of the ordinary (424), protection of the integrity or security of the dominant undertaking’s services (425), protection against customer conduct that is inconsistent with fair trade practices (426) or the need to apply below-cost price reductions in order to deplete excess and obsolete stock (427). However, conduct that distorts competition cannot be justified on the grounds that it amounts to an alignment of the dominant undertaking’s prices with those of its non-dominant competitors (428).

209.

An objective necessity may be based on technical justifications, such as the need to maintain or improve the performance of the dominant undertaking’s product (429). Where such technical justifications are invoked in relation to a refusal to ensure interoperability with an input controlled by the dominant undertaking, the lack of a necessary technical solution to allow interoperability between the input and a competing product or the difficulty of developing such a solution are not in themselves sufficient to demonstrate the existence of an objective necessity, unless the granting of such interoperability would compromise the integrity or security of the undertaking’s products, or where interoperability through the development of such a solution would be technically impossible (430). Where a dominant undertaking fails to respond to a request from an undertaking seeking interoperability, this could amount to evidence that the refusal to ensure such interoperability is not objectively justified (431).

210.

An objective necessity defence may also relate to the pursuit of a genuine and objectively defined public interest objective (432), for example, genuine and evidence-based public health objectives (433), product safety (434), network safety (435), contribution to European Union’s resilience (436), or other public interest considerations. However, the European Union courts have clarified that the pursuit of public interest objectives may not be used as a pretext for the dominant undertaking to disguise conduct that is aimed at strengthening its dominant position. Accordingly, the European Union courts have held that it is not the task of the dominant undertaking to take steps on its own initiative to eliminate products which, rightly or wrongly, it regards – based on its own perception of the qualities of such products or its own interpretation of the relevant applicable standards and regulations – as dangerous or as inferior in quality to its own products (437). More generally, it is not the task of the dominant undertaking to enforce the compliance of other undertakings with the law (438).

211.

An objective necessity defence will be accepted only if the distortion of effective competition resulting from the conduct is proportionate to the alleged aim (439). The condition of proportionality is not met where the same aim could be achieved through means that are less restrictive of competition (440). Where a conduct is found to be by its very nature harmful to competition (see paragraph 197 above), it is very unlikely that it can be justified on the basis of objective necessity (441).

5.3.   Efficiency defence

5.3.1.   General principles

212.

An efficiency defence involves demonstrating that the distortion of effective competition resulting from the dominant undertaking’s conduct is counterbalanced, or even outweighed, by advantages in terms of efficiency that also benefit consumers (442).

213.

The type of efficiencies that dominant undertakings can invoke include all objective economic efficiencies, including cost efficiencies and efficiencies of a qualitative nature.

214.

Both static and, where relevant and appropriate, dynamic considerations can be taken into account in the assessment of cost and qualitative efficiencies. Depending on the timeframe within which they come to pass, one can distinguish between short-term and longer-term cost and qualitative efficiencies. Some cost or qualitative efficiencies flow immediately from the dominant undertaking’s conduct while others take more time to materialise, as they arise, for instance, from increased investments in R&D, innovation or infrastructure.

215.

Cost efficiencies can originate from various sources, such as the development of new production technologies and methods, economies of scale and scope (443), as well as arrangements that allow for better planning or organisation of production or distribution (444).

216.

Efficiencies of a qualitative nature include, for instance, efficiencies which create value in the form of new or improved products or greater product variety (445). Qualitative efficiencies can, for example, stem from increased investment in R&D or from technical and technological advances resulting from combining physical or non-physical assets. Conduct that distorts effective competition may be objectively justified, among other reasons, because it allows the dominant undertaking to recoup or monetise investments in the development of new products (446), to secure relationship-specific investments (447), or to address free-rider or hold-up problems, or capital market imperfections (448). Qualitative efficiencies can also stem from new forms of distribution that enable the production and commercialisation of new or improved products.

217.

Sustainability benefits (449) can also take the form of qualitative efficiencies, for example when the dominant undertaking’s conduct enables the use of less raw materials or less polluting production or distribution technologies, an increase in the recyclability of products, reliance on more resilient infrastructure, a reduction in the risk of supply chain disruptions, the production of new, more sustainable products or a reduction in the time it takes to bring products to the market (450). For instance, efficiencies may result from conduct that is used to address a hold-up problem for investments that pursue sustainability objectives, such as the long-term risk of investing in a hydropower plant or wind farm (451).

218.

In certain cases, sustainability benefits stemming from the dominant undertaking’s conduct may also translate into consumer benefits in the form of cost savings, for example, where the conduct of the dominant undertaking enables more sustainable products to be produced or distributed at lower costs. As a consequence, consumers benefit not only from lower prices, but the availability of cheaper sustainable products also creates a distinct positive externality if it induces consumers to substitute less sustainable (e.g. polluting) products by more sustainable ones.

219.

Where a dominant undertaking’s conduct is found to distort effective competition, any potential efficiencies can only be examined as an objective justification (452).

220.

To establish an efficiency defence, the dominant undertaking must show that the following four cumulative conditions (453) are met: (a) the dominant undertaking’s conduct allows efficiencies to be achieved, taking into account both the actual existence and the extent of those efficiencies; (b) those efficiencies counteract any negative effects of that conduct on competition and consumers (454) on the market(s) concerned; (c) the conduct is necessary for the achievement of those efficiencies (455); and (d) the conduct does not eliminate effective competition, by removing all or most existing sources of actual or potential competition (456).

221.

Where all four conditions are met, the net effect of such conduct is to promote, on a lasting basis, the essence of the competitive process, namely to win customers by offering better products or better prices than those offered by competitors.

222.

The assessment of the four cumulative conditions follows a ‘sliding scale’. The higher the potential of the dominant undertaking’s conduct to harm competition, the less likely it is that the four conditions are fulfilled. Accordingly, while, as a matter of principle, an efficiency defence can be invoked for all types of conduct of a dominant undertaking, based on the Commission’s experience, conduct that is found to be by its very nature harmful to competition is very unlikely to fulfil the four cumulative conditions (457).

5.3.2.   Verifiable efficiencies

223.

Under the first condition, the dominant undertaking must specify the nature, the existence and the extent of the claimed efficiencies (458).

224.

All efficiencies claimed by the dominant undertaking must be substantiated and must be objective, concrete and verifiable, so that it is demonstrated with a sufficient degree of probability that the conduct makes it possible to obtain an appreciable objective advantage of such a kind as to offset the disadvantage which it entailed for competition (459).

225.

Evidence relevant to the assessment of efficiency claims includes, in particular, contemporaneous internal documents relating to the decision-making process regarding the conduct at issue, financial and accounting documents (460), historical examples of efficiencies and consumer benefits, external experts’ studies on the type and size of efficiency gains, and economic models provided that they are applicable to the facts of the case (461).

226.

Where reasonably possible, the dominant undertaking must quantify the efficiencies and the resulting benefits for consumers (462). This quantification involves estimating, as accurately as reasonably possible, the size of the efficiencies and describing in detail how the estimate was produced. The data submitted must be verifiable, so that it is demonstrated with a sufficient degree of probability that the efficiencies have materialised or are likely to materialise (463).

227.

A precise quantitative analysis may be excessively difficult or practically impossible for certain qualitative and longer-term efficiencies. In such cases, the dominant undertaking must at least describe in detail the nature of the efficiencies and why they lead to a clearly identifiable, objective and verifiable positive impact on consumers (not just a marginal one) by providing, to the extent possible, an estimate of the magnitude of the claimed benefit (464).

228.

Under this first condition, the claimed efficiencies must result from the dominant undertaking’s conduct (465). In this regard, it must be verified whether there is a sufficient causal link between the dominant undertaking’s conduct and the claimed efficiencies (466). To this end, the dominant undertaking must describe the mechanism through which the efficiencies are achieved.

229.

The causal link between the conduct and the claimed efficiencies must normally be direct. Generic claims based on indirect effects are normally too uncertain or remote to be taken into account. A direct causal link exists for instance where it can be shown that an exclusive distribution agreement allows the dominant undertaking to distribute its products at lower cost or to guarantee stable supply. Conversely, an example of efficiencies that are typically too uncertain or remote would be a case where a dominant undertaking claims that a specific conduct allows it to increase its profits, enabling it to invest more in R&D to the ultimate benefit of consumers. While there may be a link between profitability and R&D, the link is generally not sufficiently direct for the consumer benefit to be considered as an efficiency (467).

230.

In cases where the efficiencies have not yet materialised, the dominant undertaking must substantiate any projections regarding the date from which the efficiencies will become operational and start to have a significant positive impact in the market. The appropriate period for future efficiencies to be taken into account depends on the characteristics and dynamics of the market and on the theory of harm. The further into the future the efficiencies are projected to materialise, the greater the uncertainty as to whether they will occur. Efficiencies must materialise within a timeframe in which the occurrence of specific events can be predicted with a sufficient degree of certainty. When assessing efficiencies, the Commission cannot be required to carry out a prospective analysis on the basis of elements for which it is not able to envisage the long-term effects with a sufficient degree of certainty (468). When assessing the likelihood and magnitude of future efficiencies, in order to allow for an appropriate comparison of a present loss to consumers with a future gain to consumers, the value of future gains must be discounted, using a discount rate reflecting the lower value of future gains (469).

5.3.3.   Fair share for consumers

231.

The dominant undertaking must demonstrate that the efficiency gains made possible by its conduct have a positive impact on consumers (470). To this end, the dominant undertaking must demonstrate that the pass-on of the claimed efficiencies is sufficiently substantial to at least compensate affected consumers for any negative impact caused by the dominant undertaking’s conduct (471).

232.

The dominant undertaking’s incentive to pass cost efficiencies on to consumers is often related to the existence of competitive pressure from the remaining undertakings in the market or from the entry into the market of potential competitors. The greater the actual or potential negative effects on competition, the more the claimed efficiencies must be substantial, likely to be realised, and passed on, to a sufficient degree, to consumers.

233.

In the Commission’s experience, the fact that competition is already weakened due to the very presence of the dominant undertaking is a relevant factor to be taken into account when assessing whether the claimed efficiencies are likely to be passed on to consumers. Therefore, typically the greater the market power of the dominant undertaking the less likely it is that the claimed benefits will be passed on to consumers. Where the dominant undertaking’s market power is very high, or where demand is very inelastic, it is highly unlikely that the efficiency defence will apply, as it is highly unlikely that any efficiency gains will be passed on to consumers.

234.

In general, efficiencies that lead to reductions in the dominant undertaking’s variable or marginal costs are more likely to be passed on and to result in lower prices for consumers than reductions in fixed costs (472).

235.

Qualitative efficiencies can also be passed on to consumers in the form of new or improved products (473). Consumers’ valuation of higher quality can be measured by investigating their willingness to pay for improved products (474). In cases where the consumer benefits deriving from qualitative efficiencies cannot be precisely quantified, the dominant undertaking must substantiate that the increase in value stemming from such efficiencies is sufficient to compensate for any harm caused by its conduct (475).

236.

The assessment of the benefits flowing from the dominant undertaking’s conduct is, in principle, made within the boundaries of each relevant market to which the conduct relates. Negative effects on consumers in one geographic market or product market cannot normally be balanced against and compensated by positive effects for consumers in another unrelated geographic market or product market (476).

237.

However, if two markets are related, efficiencies achieved on separate markets can be taken into account, provided that the group of consumers affected by the restriction and benefiting from the efficiency gains are substantially the same (477). This may be the case, for instance, when the dominant undertaking’s conduct enables increased investments in new infrastructure or technology which has spill-over effects across various related levels of the value chain that are not part of the relevant market(s) affected by the conduct. Where there is such substantial overlap between the consumers affected by the conduct and the beneficiaries outside the relevant market, the out-of-market efficiencies that accrue to consumers in the relevant market can be taken into account as (part of) an efficiency defence. Out-of-market efficiencies may also occur when the efficiencies stemming from a dominant undertaking’s conduct create, through the internalisation of negative or positive externalities, collective benefits extending beyond the relevant market(s) adversely affected by the conduct. Such collective benefits can, for example, arise when the use of cleaner technologies or the production of more sustainable products bring about sustainability benefits for a wider section of society, independently of consumers’ individual appreciation of the technology or product (478).

238.

For out-of-market efficiencies to be taken into account, the dominant undertaking must describe the claimed benefits clearly and provide evidence that they have already occurred or are likely to occur; define clearly the beneficiaries; demonstrate that the consumers negatively affected by the conduct in the relevant market substantially overlap with the beneficiaries; demonstrate that the share of out-of-market efficiencies that accrues to the consumers affected by the conduct in the relevant market, possibly together with other (in-market) efficiencies accruing to those consumers, counterbalances or outweighs the harm suffered by those consumers as a result of its conduct (479).

5.3.4.   Indispensability of the dominant undertaking’s conduct

239.

The dominant undertaking is required to demonstrate that its conduct is necessary for the achievement of the claimed efficiencies because the alleged efficiencies cannot be achieved to a similar extent by less anti-competitive alternatives (480). If the conduct goes beyond what is necessary to attain those efficiencies, the indispensability condition is not met (481). Only those alternatives that are reasonably practical in the business situation faced by the dominant undertaking will be taken into account. The dominant undertaking must explain and demonstrate why seemingly realistic and less restrictive alternatives would be significantly less efficient. Dominant undertakings invoking efficiencies are not required to consider hypothetical or theoretical alternatives (482).

240.

The assessment of the indispensability condition is made within the actual context in which the conduct of the dominant undertaking occurs and must in particular take into account the structure of the relevant markets and the economic risks and incentives linked to the dominant undertaking’s activities. For example, if a dominant undertaking has incurred substantial sunk investments in R&D, its infrastructure, or customer or supply relationships, then certain restrictions may be necessary for it to recoup the investment, to avoid hold-ups or align the incentives of the parties to the relationship. In that case, it has to be determined whether the dominant undertaking’s conduct is proportionate to recoup the investment in question in light of the investment’s nature, size and the risks associated with it (483), as well as the duration of the period required to ensure a reasonable return on the sunk investment in question and the availability of alternative sources of revenue that compensate the dominant undertaking for its investments (484).

241.

For the dominant undertaking’s conduct to be deemed indispensable, there must be sufficient evidence that there is a direct link between that conduct and the achievement of the claimed efficiencies and that there are no less restrictive alternatives to attain the efficiencies. In the example of a decision to incur substantial sunk investments, to assess whether there were less restrictive alternatives because, for example, the dominant undertaking would have made the investments without engaging in the conduct that distorts competition, there must be evidence about the decision to finance the claimed investments (485), and it is necessary to consider evidence relating to the investment decisions of the dominant undertaking and its competitors. Evidence showing that the dominant undertaking would, in the absence of the conduct, have generated sufficient revenues or had access to alternative monetisation channels will typically indicate that the conduct was not necessary for the dominant undertaking to be able to finance its claimed investments (486). Evidence that the dominant undertaking or its competitors have made similar investments in relation to products or customers not concerned by the conduct that distorts competition will also be considered as an indication that the conduct was not necessary to ensure the dominant undertaking’s incentives to make the investments (487). Likewise, where it is shown that the dominant undertaking would have incurred the expenditure for the investments even in the absence of the conduct, for instance, to respond to changing market dynamics, the conduct cannot be considered to be necessary to recoup the investments (488).

242.

When assessing the availability of less restrictive alternatives, and still using as an example a dominant undertaking’s decision to incur substantial sunk investments, the extent to which the risks associated with the investment were contractable (i.e. the possibility to enter into and enforce a contract ensuring that the dominant undertaking is directly compensated for its investments) is a factor to be considered (489).

5.3.5.   No elimination of effective competition

243.

The dominant undertaking must demonstrate that its conduct does not eliminate effective competition, by removing all or most existing sources of actual or potential competition in the relevant market (490). Rivalry between undertakings is an essential driver of economic efficiency and consumer welfare, including longer-term efficiencies in the form increased investments in R&D, innovation or infrastructure. When all or most sources of remaining effective competition are eliminated, the competitive process is brought to an end and short-term efficiency gains are outweighed by longer-term efficiency and consumer welfare losses stemming from, for example, expenditures incurred by the incumbent undertaking to maintain its position (rent seeking), misallocation of resources, reduced choice, innovation and higher prices (491).

244.

The assessment of this condition must take due account of the fact that market competition is already weakened because of the very presence of the dominant undertaking. Accordingly, the degree of dominance, and, in particular whether the dominant undertaking has an entrenched dominant or even monopolistic position, are relevant for the assessment of whether the conduct eliminates all or most sources of competition (492).

245.

Assessing whether the conduct eliminates all or most sources of remaining effective competition requires a realistic analysis of the various (actual or potential) sources of competition in the market, the level of competitive constraint that they impose on the dominant undertaking and the impact of the conduct at issue on that competitive constraint (493).

6.   FINAL REMARKS

246.

The Guidance on the Commission’s enforcement priorities in applying Article 82 of the EC Treaty to abusive exclusionary conduct by dominant undertakings (494) is hereby withdrawn and does not apply as of 30 days after the publication of these Guidelines in the Official Journal.

---

(1)  All references to ‘product(s)’ in these Guidelines include goods and services.

(2)  Judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 37; and judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 73.

(3)  Judgment of 22 October 2002, Roquette Frères, C-94/00, EU:C:2002:603, paragraph 42; judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 22; judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 124; judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 41 and 46 (the latter stating that ‘the well-being of both intermediary and final consumers must be regarded as the ultimate objective warranting the intervention of competition law’). See also judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 1028: in that case, the General Court stated that the abusive practices were detrimental to the interest of consumers in having more than one source for obtaining information on the internet, ensuring the protection of their privacy or using products having specific linguistic features within the EEA. Such interests were not only consistent with competition on the merits, in that they encouraged innovation for the benefit of consumers, but were also necessary in order to ensure plurality in a democratic society. See also Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions of 18 November 2021, A competition policy fit for new challenges, COM(2021) 713 final.

(4)  The term ‘quality’ in these Guidelines covers all aspects of the quality of a product, such as its sustainability, resource efficiency, durability, the value and variety of uses offered by the product, the possibility to integrate the product with other products, the image conveyed by or the security and privacy protection afforded by the product, as well as its availability, including in terms of lead time, resilience of supply chains, reliability of supply and transport costs. See Commission Notice on the definition of the relevant market for the purposes of Union competition law (‘Market Definition Notice’) (OJ C, C/2024/1645, 22.2.2024, ELI: http://data.europa.eu/eli/C/2024/1645/oj), paragraph 15.

(5)  See e.g. judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraphs 862 and 892.

(6)  Judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 281. See also judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 131 and 133, where the Court refers, amongst other things to ‘limiting production, product or alternative service development’.

(7)  Judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 135. In these Guidelines, the Commission uses the expression ‘distort effective competition’ to mean conduct which, absent an objective justification, constitutes an abuse of a dominant position.

(8)  Network effects arise when the value of a product for a user depends on the number of users for that product (direct network effects). For multi-sided platforms, network effects can also arise when the value of a product on one side of the platform depends on the number of users on another side of the platform (indirect network effects). See Market Definition Notice, paragraph 94 and footnote 83.

(9)  ‘Winner-takes-all’ dynamics refer to a combination of economic features such as significant scale economies, network effects, switching costs, etc., that in the medium to long run make markets prone to the emergence of a dominant undertaking or in some cases even a monopolist.

(10)  Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions of 18 November 2021, A competition policy fit for new challenges, COM(2021) 713 final.

(11)  Council Regulation (EC) No 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty [now Articles 101 and 102 TFEU] (OJ L 1, 4.1.2003, p. 1, ELI: http://data.europa.eu/eli/reg/2003/1/2009-07-01), Articles 1, 3, 5 and 6.

(12)  In these Guidelines, the term ‘European Union courts’ refers to the Court of Justice and the General Court.

(13)  Judgment of 12 May 2022, Servizio Elettrico Nazionale, C-377/20, EU:C:2022:379, paragraphs 44 and 46. Consequently, unless specified otherwise, in these Guidelines, the concept of ‘consumers’ encompasses all direct and indirect users of the products affected by the conduct of a dominant undertaking, including intermediate producers that use the products as an input, as well as distributors, wholesalers, retailers and final consumers.

(14)  Judgment of 21 February 1973, Europemballage Corporation and Continental Can Company v Commission, C-6/72, EU:C:1973:22, paragraph 26; judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 24; and judgment of 12 May 2022, Servizio Elettrico Nazionale, C-377/20, EU:C:2022:379, paragraph 44.

(15)  Judgment of 12 May 2022, Servizio Elettrico Nazionale, C-377/20, EU:C:2022:379, paragraphs 44 and 68; judgment of 4 July 2023, Meta Platforms and Others (General terms of use of a social network), C-252/21, EU:C:2023:537, paragraph 47; judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 131 and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 187.

(16)  For the notion of ‘exclusionary effects’, see Section 3.3.1.

(17)  Pursuant to the case law of the European Union Courts, the Commission may depart from guidelines in individual cases by giving reasons that are compatible with the principle of equal treatment, see judgment of 13 December 2012, Expedia, C-226/11, EU:C:2012:795, paragraphs 23 to 31; judgment of 13 July 2023, Commission v CK Telecoms UK Investments, C-376/20 P, EU:C:2023:561, paragraph 123; judgment of 14 July 2006, Endesa v Commission, T-417/05, EU:T:2006:219, paragraph 203; and 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, 211 and 213. In this regard, the Commission considers that one possible reason to deviate from these Guidelines is the need to align its practice dynamically to the interpretation given by the European Union Courts to Article 102 TFEU.

(18)  Judgments of 21 February 1973, Europemballage and Continental Can v Commission, C-6/72, EU:C:1973:22, paragraph 26; judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 26; and judgment of 16 March 2023, Towercast, C-449/21, EU:C:2023:207, paragraph 46.

(19)  Judgment of 13 February 1979, Hoffmann-La-Roche v Commission, C-85/76, EU:C:1979:36, paragraph 116; and judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 119. Articles 101 and 102 TFEU must be interpreted consistently, in light, however, of their specific characteristics; see judgment of 21 December 2023, International Skating Union v Commission, C-124/21, EU:C:2023:1012, paragraph 128; and judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 186. Rules on the application of Article 101 TFEU include, for example, Commission Regulation (EU) 2022/720 of 10 May 2022 on the application of Article 101(3) of the Treaty on the Functioning of the European Union to categories of vertical agreements and concerted practices (OJ L 134, 11.5.2022, p. 4, ELI: http://data.europa.eu/eli/reg/2022/720/2024-02-11); Commission Regulation (EU) 2023/1066 of 1 June 2023 on the application of Article 101(3) of the Treaty on the Functioning of the European Union to certain categories of research and development agreements (OJ L 143, 2.6.2023, p. 9, ELI: http://data.europa.eu/eli/reg/2023/1066/oj); and Commission Regulation (EU) 2023/1067 of 1 June 2023 on the application of Article 101(3) of the Treaty on the Functioning of the European Union to certain categories of specialisation agreements (OJ L 143, 2.6.2023, p. 20, ELI: http://data.europa.eu/eli/reg/2023/1067/oj). Moreover, as regards the application of merger control pursuant to Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations between undertakings (OJ L 24, 29.1.2004, p. 1, ELI: http://data.europa.eu/eli/reg/2004/139/oj) (‘the Merger Regulation’), the Court of Justice has held that ‘Article 21(1) of Regulation No 139/2004 must be interpreted as not precluding the competition authority of a Member State from regarding a concentration of undertakings which has no Community dimension within the meaning of Article 1 thereof, is below the thresholds for mandatory ex ante control laid down in national law, and has not been referred to the Commission under Article 22 of that regulation, as constituting an abuse of a dominant position prohibited under Article 102 TFEU, in the light of the structure of competition on a market which is national in scope’ (judgment of 16 March 2023, Towercast, C-449/21, EU:C:2023:207, paragraph 53).

(20)  Judgment of 4 July 2023, Meta Platforms and Others (General terms of use of a social network), C-252/21, paragraphs 47-54; and judgment of 12 January 2023, Lietuvos geležinkeliai v Commission, C-42/21 P, EU:C:2023:12, paragraph 88. See also judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraph 366, where the Court held that the existence of remedies specific to another regulatory system (in that case, the patent system) does not alter the conditions of application of the competition law regime.

(21)  The application of the principle non bis in idem is subject to a twofold condition, namely, first, that there must be a prior final decision (the ‘bis’ condition) and, second, that the prior decision and the subsequent proceedings or decisions must concern the same facts (the ‘idem’ condition) (judgment of 22 March 2022, bpost, C-117/20, EU:C:2022:202, paragraph 28). The European Union courts have held that an undertaking can be subject to a duplication of proceedings and penalties if: (i) there are clear and precise rules making it possible to predict which acts or omissions are liable to be subject to a duplication of proceedings and penalties, and also to predict that there will be coordination between the two competent authorities; (ii) the two sets of proceedings have been conducted in a sufficiently coordinated manner within a proximate timeframe; and (iii) the overall penalties imposed correspond to the seriousness of the offences committed. See judgment of 22 March 2022, bpost, C-117/20, EU:C:2022:202, paragraph 58.

(22)  Judgment of 10 July 2014, Telefónica and Telefónica de España v Commission, C-295/12 P, EU:C:2014:2062, paragraph 133. If a conduct is imposed on undertakings by legislation, or if legislation creates a legal framework which eliminates any possibility of competitive activity, Article 102 TFEU does not apply. In such situations, the abusive conduct is not attributable to the dominant undertaking, as Article 102 TFEU implicitly requires the autonomous conduct of that undertaking. Article 102 TFEU may apply, however, if that legislation leaves open the possibility of competition which may be prevented, restricted or distorted by the autonomous conduct of undertakings. Judgment of 14 October 2010, Deutsche Telekom v Commission, C-280/08 P, EU:C:2010:603, paragraph 80.

(23)  Commission Notice – Guidelines on the effect on trade concept contained in Articles 81 and 82 of the Treaty (OJ C 101, 27.4.2004, p. 81).

(24)  Article 102 TFEU also stipulates that the dominant position must be held in the whole or a substantial part of the internal market. This requirement has been met even in cases where the dominant position of the undertaking concerned was limited to a region within a Member State or even to a port. See judgment of 10 December 1991, Merci convenzionali porto di Genova, C-179/90, EU:C:1991:464, paragraph 15; and judgment of 20 June 2018, České dráhy a.s., T-325/16, EU:T:2018:368, paragraph 146.

(25)  Judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraph 65.

(26)  See judgment of 17 December 2003, British Airways v Commission, T-219/99, EU:T:2003:343, paragraphs 101 and 191; and judgment of 15 December 1994, Gøttrup-Klim u.a. Grovvareforeninger v Dansk Landbrugs Grovvareselska, C-250/92, EU:C:1994:413, paragraphs 46-52. To establish dominance on the buyer side, it is necessary to define a purchasing market; for more guidance in this respect, see Market Definition Notice, paragraph 7. Unless otherwise stated, the remainder of these Guidelines does not deal with dominance and abusive conduct on the buyer side. However, the principles set out in these Guidelines may apply mutatis mutandis to dominance on the buyer side, taking into account the specificities of this type of situation.

(27)  Judgment of 13 February 1979, Hoffmann-La-Roche v Commission, C-85/76, EU:C:1979:36, paragraph 70; and judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraph 113.

(28)  Judgment of 30 January 2007, France Télécom v Commission, T-340/03, EU:T:2007:22, paragraph 101.

(29)  Judgment of 21 February 1973, Europemballage Corporation and Continental Can Company v Commission, C-6/72, EU:C:1973:22, paragraph 32; judgment of 30 January 2020, Generics UK & Others, C-307/18, EU:C:2020:52, paragraphs 127 and 128; and judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraph 83. See also the Market Definition Notice, paragraph 9 a). Where an undertaking holds a dominant position in an upstream market, it may not be necessary to define the downstream market in which the allegedly abusive conduct produces exclusionary effects. See judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraphs 84-86.

(30)  Market definition is not discussed in detail in these Guidelines.

(31)  One issue that can arise in the context of market definition in Article 102 TFEU cases is that the price charged by the undertaking concerned may have been raised above the competitive level. In such a scenario, the use of the so-called SSNIP test (Small but Significant and Non-transitory Increase in Price test) as a tool for defining the boundaries of the relevant market may not be appropriate, as market definition must be based on substitution at the competitive prices and not at prices that have already been raised above the competitive level. The risk that a SSNIP test analysis starting at an already inflated price leads to a false conclusion of wide relevant markets is called the ‘cellophane fallacy’. If used at all, the SSNIP test must therefore be carefully applied in Article 102 TFEU cases. See, for example, Commission decision of 15 October 2014 in case AT.39253 – Slovak Telekom, paragraphs 158-171. See also the Market Definition Notice, paragraph 18 c) and footnote 55.

(32)  See the Market Definition Notice, paragraph 18 c).

(33)  Market power relates to the ability of an undertaking to profitably maintain prices above competitive levels for a period of time or to profitably maintain output in terms of product quantities, product quality, variety or innovation below competitive levels for a period of time.

(34)  Judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraphs 80 and 81; judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraph 180; and judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 265. See also paragraph 106(a) of these Guidelines.

(35)  See, to that effect, judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraph 65.

(36)  Judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraph 66; and judgment of 15 December 1994, Gøttrup-Klim u.a. Grovvareforeninger v Dansk Landbrugs Grovvareselska, C-250/92, EU:C:1994:413, paragraph 47.

(37)  A finding of dominance is not excluded by the fact that the undertaking concerned may have been temporarily not profitable (or even loss making); see judgment of 9 November 1983, Nederlandsche Banden-Industrie-Michelin v Commission (Michelin I), C-322/81, EU:C:1983:313, paragraph 59.

(38)  These may – depending on the case – include an analysis of competitive constraints exerted from outside the product market on which the undertaking concerned may hold a dominant position. On out-of-market constraints see the Market Definition Notice, paragraph 17. See also the judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraphs 146-154; Commission decision of 4 March 2024 in case AT.40437 – Apple – App Store (music streaming), paragraphs 335 and 357-364. See also Section 2.2.4 for guidance on the assessment of dominance in the presence of after-markets, where effective competition on the primary market may constrain an undertaking’s market power in the secondary market.

(39)  See also the Market Definition Notice, paragraph 18 b).

(40)  What a sustained period of time is will depend on the specifics of the case and the product and the market in question.

(41)  Judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraph 41.

(42)  Judgment of 3 July 1991, Akzo v Commission, C-62/86, EU:C:1991:286, paragraph 60. For example, in the judgment of 12 December 1991, Hilti v Commission, T-30/89, EU:T:1991:70, paragraph 92, a market share of between 70 % and 80 % was considered a clear indication of the existence of a dominant position.

(43)  Judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraphs 108 and 109, where dominance was found with a market share of between 40 % and 45 %.

(44)  See judgment of 17 December 2003, British Airways v Commission, T-219/99, EU:T:2003:343, paragraphs 210-211 and 224; and judgment of 15 December 1994, Gøttrup-Klim u.a. Grovvareforeninger v Dansk Landbrugs Grovvareselska, C-250/92, EU:C:1994:413, paragraph 48.

(45)  See, to that effect, judgment of 15 December 1994, Gøttrup-Klim and Others Grovvareforeninger v Dansk Landbrugs Grovvareselskab, C-250/92, EU:C:1994:413, paragraph 48; and judgment of 17 December 2003, British Airways v Commission, T-219/99, EU:T:2003:343, paragraphs 216 and 217.

(46)  As a general rule, the Commission relies on market shares computed over one year reference periods, but, depending on the characteristic of the relevant market, a different reference period may be appropriate; see Market Definition Notice, paragraph 113.

(47)  For more information, see the Market Definition Notice, paragraph 108. In zero-price markets, other metrics such as the numbers of users, or transactions, or indicators of the intensity of usage may provide a better basis for assessing dominance; see, for example, Commission decision of 27 June 2017 in case AT.39740 – Google Shopping, paragraphs 275-284.

(48)  Judgment of 17 December 2003, British Airways v Commission, T-219/99, EU:T:2003:343, paragraph 210. See, also Commission decision of 15 October 2014 in case AT.39523 – Slovak Telekom, paragraph 312.

(49)  Judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraphs 111 and 112.

(50)  Judgment of 14 December 2005, General Electric v Commission, T-210/01, EU:T:2005:456, paragraphs 150 and 151; and Commission decision of 13 May 2019 in case AT.40134 – AB InBev beer trade restrictions, paragraphs 69 and 70.

(51)  Judgment of 11 December 2013, Cisco Systems and Messagenet v Commission, T-79/12, EU:T:2013:635, paragraph 69.

(52)  Judgment of 30 January 2007, France Télécom v Commission, T-340/03, EU:T:2007:22, paragraphs 103-108.

(53)  See, for example, the Market Definition Notice, paragraph 108.

(54)  Judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraphs 122 and 124; and judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraphs 48 and 49.

(55)  Regulations intended to control or regulate the construction, demolition, alteration or use of land or buildings.

(56)  For example, in pharmaceutical markets, potential market entrants typically have to obtain marketing authorisations and negotiate pricing and reimbursement conditions with national authorities; see Commission decision of 15 June 2005 in case COMP/A. 37.507/F3 – AstraZeneca, paragraph 527; and Commission decision of 10 February 2021 in case AT.40394 – Aspen, paragraph 67. Another form of legal and regulatory barriers are regulations that permit certain activities to be carried out only after the completion of a professional qualification. For more examples, see Commission decision of 4 March 2024 in case AT.40437 – Apple – App Store (music streaming), paragraphs 341 and 342.

(57)  See for example judgment of 23 April 1991, Höfner v Macrotron, C-41/90, EU:C:1991:161, paragraph 28; and Commission decision of 2 October 2017 in case AT.39813 – Baltic rail, paragraph 162.

(58)  The mere possession of IP rights does not necessarily confer a dominant position, but their possession may in certain circumstances create a dominant position by enabling an undertaking to prevent effective competition on the market; see judgment of 6 December 2012, AstraZeneca v Commission, C-457/10, EU:C:2012:770, paragraphs 186-187. While ownership of a standard essential patent (‘SEP’) does not on its own equate to dominance, it may nevertheless be established that a SEP confers a dominant position based on an assessment of all the relevant factors. See in this respect Commission decision of 29 April 2014 in case AT.39985 – Motorola, paragraphs 223 and 241; and Commission decision of 29 April 2014 in case AT.39939 – Samsung, paragraph 46. For another example of dominance in the field of industry standards, see Commission decision of 9 December 2009 in case AT.38636 – Rambus, paragraphs 18-26.

(59)  Such other barriers to entry or expansion may be raised by the undertaking concerned itself; see e.g. judgment of 14 September 2022, Google Android, T-604/18, EU:T:2022:541, paragraphs 228-233.

(60)  Judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraph 48; and Commission decision of 13 May 2019 in case AT.40134 – AB InBev beer trade restrictions, paragraph 74.

(61)  Judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraph 122; and Commission decision of 22 June 2011 in case COMP/39.525 – Telekomunikacja Polska, paragraph 656.

(62)  Commission decision of 15 October 2014 in case AT.39523 – Slovak Telekom, paragraph 332.

(63)  Judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraphs 70-81 and 122; and Commission decision of 15 October 2014 in case AT.39523 – Slovak Telekom, paragraph 291. Where most companies in an industry are vertically integrated, potential market entrants may need to enter the relevant market at all levels in order to compete. This in turn increases the financial and managerial resources required to enter and compete.

(64)  Commission decision of 24 May 2018 in case AT.39816 – Upstream gas supplies in Central and Eastern Europe, paragraph 34.

(65)  Judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraph 105 (on the inertia of doctors in their prescribing habits); and judgment of 6 December 2012, AstraZeneca v Commission, C-457/10 P, EU:C:2012:770, paragraph 50. See also Commission decision of 31 October 2024 in case AT.40588 – Teva Copaxone, paragraphs 971-980.

(66)  Judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraphs 91-94. See also Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraphs 709-712; Commission decision of 13 May 2019 in case AT.40134 – AB InBev beer trade restrictions, paragraph 74; and Commission decision of 20 December 2022 in cases AT.40462 – Amazon Marketplace and AT.40703 – Amazon Buy Box, paragraph 89.

(67)  Judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraph 278; and Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraph 637.

(68)  Commission decision of 7 October 2020 in case AT.40608 – Broadcom, paragraph 49; and Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraph 634.

(69)  Judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraphs 285-286; Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraph 635.

(70)  Judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 278.

(71)  Commission decision of 22 June 2011 in case COMP/39.525 – Telekomunikacja Polska, paragraph 648; Commission decision of 20 March 2019 in case AT.40411 – Google Search (AdSense), paragraph 242.

(72)  Commission decision of 4 March 2024 in case AT.40437 – Apple – App Store (music streaming), paragraph 399.

(73)  Judgment of 11 December 2013, Cisco Systems and Messagenet v Commission, T-79/12, EU:T:2013:635, paragraph 73; judgment of 14 September 2022, Google Android, T-604/18, EU:T:2022:541, paragraphs 115 and 202-205. See also Commission decision of 4 May 2017 in case AT.40153 – E-book MFNs and related matters (Amazon), paragraph 65.

(74)  Judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraphs 115 and 184.

(75)  For a definition of direct and indirect network effects, see footnote 8.

(76)  Commission decision of 4 March 2024 in case AT.40437 – Apple – App Store (music streaming), paragraphs 344 and 345; Commission decision of 20 March 2019 in case AT.40411 – Google Search (AdSense), paragraphs 249-251; Commission decision of 20 December 2022 in cases AT.40462 – Amazon Marketplace and AT.40703 – Amazon Buy Box, paragraph 90; and Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraphs 611-616. See, to that effect, also judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 171.

(77)  ‘Single-homing’ refers to the use by consumers of one platform for a given product, as opposed to the use of multiple platforms in parallel for the same product (multi-homing). See the Market Definition Notice, footnote 131.

(78)  A digital ecosystem exists, for example, where several categories of suppliers, customers and consumers are brought together and interact on a platform, and where the products comprising that ecosystem (i.e. the interlinked products) may overlap with, or be connected to, each other in terms of their horizontal or vertical complementarity; see judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 116.

(79)  See Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraph 636; and Commission decision of 4 March 2024 in case AT.40437 – Apple – App Store (music streaming), paragraphs 365 and 389-397.

(80)  Commission decision of 27 June 2017 in case AT.39740 – Google Shopping, paragraph 300.

(81)  In the case of possible dominance on the buyer side, constraints can be exerted by suppliers of the undertaking concerned.

(82)  Judgment of 7 October 1999, Irish Sugar v Commission, T-228/97, EU:T:1999:246, paragraphs 97-104, in which it was considered whether the alleged lack of independence of the undertaking in relation to its two main customers should be seen as an exceptional circumstance preventing the finding of a dominant position even though the undertaking accounted for a very large share of sales on the industrial sugar market in Ireland. See also Commission decision of 17 October 2013 in case COMP/39.866 – Ryanair/DAA-Aer Lingus, paragraph 78.

(83)  Commission decision of 29 April 2014 in case AT.39985 – Motorola, paragraphs 242, 243 and 257.

(84)  Commission decision of 29 April 2014 in case AT.39985 – Motorola, paragraph 244; and Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraph 322.

(85)  Commission decision of 4 March 2024 in case AT.40437 – Apple – App Store (music streaming), paragraph 353; and Commission decision of 20 December 2012 in case AT.39230 – Rio Tinto Alcan, paragraph 48.

(86)  Commission decision of 20 December 2012 in case AT.39654 – Reuters Instrument Codes, paragraph 36; Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraph 644; and Commission decision of 10 February 2021 in case AT.40394 – Aspen, paragraph 70.

(87)  This can be the case if the products of the undertaking concerned are considered ‘must-have’ products by buyers; see Commission decision of 23 May 2024 in case AT.40632 – Mondelez Trade Restrictions, paragraphs 211-215.

(88)  For more guidance on market definition in the presence of after-markets see the Market Definition Notice, paragraphs 99-104. As regards possible similarities between after-market settings and digital ecosystems, see Commission decision of 4 March 2024 in case AT.40437 – Apple – App Store (music streaming), paragraphs 336, 337, 362, 363 and 414.

(89)  If the markets for the primary and secondary products exhibit the characteristics of a system market (see the Market Definition Notice, paragraphs 100(a) and 101), the assessment of dominance will focus on that market. Furthermore, the considerations outlined in the following paragraphs – regarding the link between competition in the primary market and competition in the after-market – are less relevant in the context of an after-market where secondary products from different suppliers are compatible with all or most of the primary products (see the Market Definition Notice, paragraphs 100(c) and 102).

(90)  Commission decision of 20 May 2009 in case COMP/C-3/39.391 – EFIM confirmed in the judgments of 24 November 2011, EFIM v Commission, T-296/09, EU:T:2011:693, paragraphs 60, 90 and 91, and of 19 September 2013, EFIM v Commission, C-56/12, EU:C:2013:575, paragraphs 12 and 36-37. See also Commission decision of 13 December 2011 in case COMP/C-3/39692 – IBM Maintenance Services, paragraphs 21-24; and Commission decision of 7 January 1999 in case IV/E 2/36.431 – Info-Lab/Ricoh, paragraphs 35 and 37-50.

(91)  Judgments of 24 November 2011, EFIM v Commission, T-296/09, EU:T:2011:693, paragraphs 60, 90 and 91 and of 19 September 2013, EFIM v Commission, C-56/12, EU:C:2013:575, paragraphs 12 and 36-37.

(92)  See, for example, the judgment of 19 September 2013, EFIM v Commission, C-56/12, EU:C:2013:575, paragraphs 12 and 36-37.

(93)  Judgment of 16 March 2000, Compagnie Maritime Belge Transports and Others v Commission, joined cases C-395/96 P and C-396/96 P, EU:C:2000:132, paragraph 36. The concept of collective dominance has developed in parallel under the case-law on Article 102 TFEU and the case-law on merger control, including under the Merger Regulation, with a similar concept of collective dominance applied under both instruments. See, for example, judgment of 26 January 2005, Laurent Piau v Commission, T-193/02, EU:T:2005:22, paragraphs 109-111, on the applicability of the concept of collective dominance based on tacit coordination to Article 102 TFEU.

(94)  Judgment of 30 September 2003, Atlantic Container Line and Others v Commission, joined cases T-191/98 and T-212/98 to T-214/98, EU:T:2003:245, paragraphs 631-633 and 653-655.

(95)  Judgment of 7 October 1999, Irish Sugar v Commission, T-228/97, EU:T:1999:246, paragraph 66. See also Commission decision of 26 November 2008 in cases COMP/39.388 – German Electricity Wholesale Market and COMP/39.389 – German Electricity Balancing Market, paragraph 27.

(96)  Judgment of 31 March 1998, French Republic and others v Commission, joined cases C-68/94 and C-30/95, EU:C:1998:148, paragraph 221; judgment of 7 October 1999, Irish Sugar v Commission, T-228/97, EU:T:1999:246, paragraph 46; and judgment of 16 March 2000, Compagnie Maritime Belge Transports and Others v Commission, joined cases C-395/96 P and C-396/96 P, EU:C:2000:132, paragraph 42.

(97)  Judgment of 16 March 2000, Compagnie Maritime Belge Transports and Others v Commission, joined cases C-395/96 P and C-396/96 P, EU:C:2000:132, paragraph 45; and judgment of 25 March 1999, Gencor v Commission, T-102/96, EU:T:1999:65, paragraphs 273-276.

(98)  Judgment of 16 March 2000, Compagnie Maritime Belge Transports and Others v Commission, joined cases C-395/96 P and C-396/96 P, EU:C:2000:132, paragraphs 36 and 42-44.

(99)  Judgment of 16 March 2000, Compagnie Maritime Belge Transports and Others v Commission, joined cases C-395/96 P and C-396/96 P, EU:C:2000:132, paragraphs 44 and 45.

(100)  Two or more undertakings have interlocking directorships when they have one or more members of their management boards in common.

(101)  A collective market share above 50 % is, in the absence of evidence to the contrary, an indication of the ability of the collective entity to behave to an appreciable extent independently of its competitors, its customers and ultimately of consumers; see judgment of 30 September 2003, Atlantic Container Line and Others v Commission, joined cases T-191/98 and T-212/98 to T-214/98, EU:T:2003:245, paragraphs 932-936; and judgment of 25 March 1999, Gencor v Commission, T-102/96, EU:T:1999:65, paragraph 206.

(102)  Judgment of 7 October 1999, Irish Sugar v Commission, T-228/97, EU:T:1999:246, paragraph 63; see also Commission decision of 22 June 2005 in case COMP/A.39.116/B2 – Coca Cola, paragraphs 23-25.

(103)  Judgment of 16 March 2000, Compagnie Maritime Belge Transports and Others v Commission, joined cases C-395/96 P and C-396/96 P, EU:C:2000:132, paragraph 45.

(104)  Judgment of 25 March 1999, Gencor v Commission, T-102/96, EU:T:1999:65, paragraphs 273 and 276.

(105)  For instance, the increasing use of algorithms may lead to more findings of collective dominance based on tacit coordination, as algorithms can facilitate such coordination also in more complex environments (see paragraphs 50 and 51 of these Guidelines).

(106)  See also judgment of 6 June 2002, Airtours v Commission, T-342/99, EU:T:2002:146, paragraphs 61 and 62; and judgment of 26 January 2005, Laurent Piau v Commission, T-193/02, EU:T:2005:22, paragraph 111.

(107)  Judgment of 10 July 2008, Bertelsmann and Sony Corporation of America v Impala, C-413/06, EU:C:2008:392, paragraphs 125 and 126.

(108)  Judgment of 10 July 2008, Bertelsmann and Sony Corporation of America v Impala, C-413/06, EU:C:2008:392, paragraph 123.

(109)  Judgment of 6 June 2002, Airtours v Commission, T-342/99, EU:T:2002:146, paragraph 61. Coordination is typically easier when a small number of undertakings are involved, and it becomes more difficult when many undertakings are involved.

(110)  Judgment of 25 March 1999, Gencor v Commission, T-102/96, EU:T:1999:65, paragraph 134.

(111)  Corporate links such as cross-shareholding or participation in joint ventures, may also help align incentives and encourage parallel behaviour; see Commission decision of 13 June 2000 in case COMP/M.1673 – VEBA/VIAG, paragraph 226.

(112)  See e.g. Commission decision of 7 May 2018 in case M.8444 – ArcelorMittal/Ilva, paragraph 1128.

(113)  See judgment of 6 June 2002, Airtours v Commission, T-342/99, EU:T:2002:146, paragraph 62, first indent.

(114)  Judgment of 6 June 2002, Airtours v Commission, T-342/99, EU:T:2002:146, paragraph 62, second indent.

(115)  Judgment of 25 March 1999, Gencor v Commission, T-102/96, EU:T:1999:65, paragraph 276.

(116)  Judgment of 6 June 2002, Airtours v Commission, T-342/99, EU:T:2002:146, paragraph 195; and judgment of 13 July 2006, Impala v Commission, T-464/04, EU:T:2006:216, paragraph 466.

(117)  See Commission decision of 4 July 2018 in Case M.8451 – Tronox/Cristal, paragraphs 365-370.

(118)  See judgment of 10 July 2008, Bertelsmann AG and Sony Corporation of America v Impala, C-413/06 P, EU:C:2008:392, paragraph 123.

(119)  Sections 3, 4 and 5 of these Guidelines generally refer to conduct by one undertaking that has a single dominant position but, all considerations equally, they apply mutatis mutandis to conduct by collectively dominant undertakings.

(120)  Actions by third parties (for instance, a dominant undertaking’s distributors) may be attributed to a dominant undertaking if it is established that those actions were not adopted independently by those third parties, but form part of a policy that is decided unilaterally by the dominant undertaking (judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 33).

(121)  Judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 135. The actual scope of that special responsibility must be considered in the light of the specific circumstances of each case. See judgment of 14 November 1996, Tetra Pak v Commission, C-333/94 P, EU:C:1996:436, paragraph 24.

(122)  See judgment of 19 April 2012, Tomra and Others v Commission, C-549/10 P, EU:C:2012:221, paragraphs 19-21; judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 60-62; and judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 45. However, proof of intent may constitute a relevant factor to be taken into consideration in the assessment of the abuse, see paragraph 106(f) of these Guidelines, and is a condition to establish predatory pricing within the meaning of paragraph 113(b) of these Guidelines.

(123)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 130; judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraph 179; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 190-199 and 222. This contextual examination is different from the assessment of the actual actions or reactions of third parties once the conduct is implemented (see paragraph 103 of these Guidelines).

(124)  For instance, as regards self-preferencing it is necessary to establish according to the relevant legal standard that the conduct distorts effective competition (see Section 4.9), whereas for exclusive dealing it is sufficient to show that the conduct amounts to exclusive dealing in order to trigger the application of a presumption that the conduct distorts effective competition (see Section 4.5). Moreover, when it is established that a conduct is by its very nature harmful to competition, this conduct is deemed as such to distort effective competition (see Section 4.10).

(125)  See to that effect, judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraphs 137-139; and judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraphs 46-48, and 51-52.

(126)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 68 and 103; judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 129-131; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 264.

(127)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 47.

(128)  See judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 170 and 171.

(129)  See paragraph 107 below for a description of the different types of conduct that are subject to specific analytical frameworks. These Guidelines are without prejudice to the European Union courts developing additional specific analytical frameworks for other types of conduct by dominant undertakings.

(130)  See judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 166. See also judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraph 179; judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 130; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 269.

(131)  See paragraphs 85, 86, 89 and 92 below. According to the case-law of the European Union courts, a hypothetical equally efficient competitor is a hypothetical competitor that is as efficient as the dominant undertaking but does not hold a dominant position on the market in question. Accordingly, such a hypothetical equally efficient competitor does not have access to resources or means that are inherent to the holding of the dominant position; see judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 78.

(132)  See Section 4.10 below.

(133)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 86 and 103.

(134)  Judgment of 17 July 1998, ITT Promedia v Commission, T-111/96, EU:T:1998:183, paragraph 138; and judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraph 24. See also judgment of 3 July 1991, Akzo v Commission, C-62/86, EU:C:1991:286, paragraph 70; judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraph 106, which refer to the concept of ‘competition on the basis of quality’; and judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 171, which refers to the ‘merits of the service’.

(135)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 85; judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 265.

(136)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 85. However, the mere fact of putting a new good on the market or increasing the quantity or quality of the goods produced is not as such sufficient to demonstrate that the undertaking is competing on the merits, given that all the elements mentioned in Section 3.2.2 have to be taken into account.

(137)  Judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 37; judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 45 and 73; judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 134; and judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraphs 21 and 22; judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 267.

(138)  Judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraph 189; judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraph 46; and judgment 30 January 2020, Generics UK & Others, C-307/18, EU:C:2020:52, paragraphs 149-151.

(139)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 74 and 75.

(140)  For the notion of competition on the merits, see paragraph 67 above.

(141)  Judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 168.

(142)  Judgment of 6 December 2012, AstraZeneca v Commission, C-457/10 P, EU:C:2012:770, paragraph 98. See also, on disparagement of competing products, Commission decision of 31 October 2024 in case AT.40588 – Teva Copaxone, paragraph 1316. See also, in relation to the disparagement of competing products vis-à-vis relevant stakeholders, Commission decision of 22 July 2024 in case AT.40577 – Vifor (IV iron products), paragraph 82.

(143)  Judgment of 6 December 2012, AstraZeneca v Commission, C-457/10 P, EU:C:2012:770, paragraph 134; and judgment of 17 July 1998, ITT Promedia v Commission, T-111/96, EU:T:1998:183, paragraph 72.

(144)  Judgment of 4 July 2023, Meta Platforms and Others (General terms of use of a social network), C-252/21, EU:C:2023:537, paragraphs 47 and 51.

(145)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 78; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 268.

(146)  This should not be understood as an exhaustive list of all the factors that may be relevant to establish that a given conduct departs from competition on the merits. In addition, one factor may be sufficient to conclude that a given conduct departs from competition on the merits in light of the specific circumstances at hand.

(147)  See judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, in particular, paragraphs 1046-1047, 1057-1058 and 1069-1070; judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 269. See also judgment of 23 October 2003, Van den Bergh Foods v Commission, T-65/98, EU:T:2003:281, paragraphs 148, 152 and 157.

(148)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 96-99; and judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 131 and 135. See also, to that effect, judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 226. As regards the more favourable treatment of the dominant undertaking’s own products, see Section 4.9 and judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 186.

(149)  See judgment of 19 April 2012, Tomra and Others v Commission, C-549/10 P, EU:C:2012:221, paragraph 22.

(150)  Judgment of 30 January 2020, Generics (UK) and Others, C-307/18, EU:C:2020:52, paragraph 162; judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 63; and judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 45.

(151)  Judgment of 16 March 2000, Compagnie Maritime Belge Transports and Others v Commission, joined cases C-395/96 P and C-396/96 P, EU:C:2000:132, paragraph 131; judgment of 30 September 2003, Atlantic Container Line and Others v Commission, joined cases T-191/98 and T-212/98 to T-214/98, EU:T:2003:245, paragraphs 1124 and 1460; and judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 613.

(152)  See judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 85 and 86; judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 188 and 266; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 229-230.

(153)  For the avoidance of doubt, in case of conflict, the factors set out in Section 4 regarding the assessment of exclusionary effects specific to certain types of conduct are to prevail over the general elements and principles outlined in the present section.

(154)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 129-131; and judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 50. See also judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 77; and judgment of 19 April 2012, Tomra and Others v Commission, C-549/10 P, EU:C:2012:221, paragraph 68. See however paragraph 58 above regarding the existence of presumptions that certain types of conduct distort effective competition.

(155)  See, to that effect, judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 124; judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 87 and 165; judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraphs 37, 51 and 54; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 264.

(156)  Judgment of 6 December 2012, AstraZeneca v Commission, C-457/10 P, EU:C:2012:770, paragraphs 134 and 154.

(157)  Judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 281.

(158)  Judgment of 8 October 1996, Compagnie Maritime Belge Transports and Others v Commission, joined cases T-24/93, T-25/93, T-26/93 et T-28/93, EU:T:1996:139, paragraph 149; judgment of 4 July 2023, Meta Platforms and Others (General terms of use of a social network), C-252/21, EU:C:2023:537, paragraph 47; and judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 131.

(159)  See judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 61; judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 31; judgment of 15 March 2007, British Airways v Commission, T-219/99 P, EU:T:2003:343, paragraphs 297 and 298; judgment of 8 October 1996, Compagnie Maritime Belge Transports and Others v Commission, joined cases T-24/93, T-25/93, T-26/93 and T-28/93, EU:T:1996:139, paragraph 149; judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 55 and 56; and judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraphs 58-60.

(160)  Judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 264.

(161)  See e.g. judgment of 6 April 1995, C-241/91 P and C-242/91 P, RTE and ITP v Commission, EU:C:1995:98, paragraph 54; and judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 792.

(162)  Judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 65; judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 70 and 98; and judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 42. A practice cannot be deemed to be abusive if it has remained at the planning stage (see judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 43).

(163)  Judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 41. See also the judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraphs 55-57.

(164)  See judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraphs 110-113.

(165)  Article 102 TFEU only applies to conduct that is capable of affecting trade between Member States, see paragraph 15 and footnote 23 above.

(166)  See, to that effect, judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraphs 72 and 73; and judgment of 19 April 2012, Tomra Systems ASA and Others v Commission, EU:C:2012:221, paragraphs 46 and 48.

(167)  Judgment of 19 April 2012, Tomra and Others v Commission, C-549/10 P, EU:C:2012:221, paragraph 42; judgment of 30 January 2020, Generics (UK) and Others, C-307/18, EU:C:2020:52, paragraph 161; and judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, paragraph 596.

(168)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 129; and judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 165.

(169)  See Section 3.3.2.1 below.

(170)  See judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 131; judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 167 and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 271. The facts in the judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726 concern the leveraging by Google (Google LLC) of its dominant position on the market for online general searches into the market for comparison shopping services, where Google was a new entrant and other competing undertakings were already present. The Commission therefore considers, as a general matter, that the reference to ‘potentially competing undertakings’ in paragraph 167 of that judgment cannot be seen as referring only to undertakings not operating in the market but must extend to all undertakings whose entry into or expansion in the relevant market may be hampered.

(171)  See judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 131; judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 263-264 and 267-269 (where the Court of Justice agreed with the General Court that a price-cost test was neither ‘mandatory in the context of the application of Article 102 TFEU’ nor relevant ‘in the circumstances of the present case’). See also judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 272-273, 278 and 284, where the Court of Justice held that in that case it was not possible, nor would it make sense to assess exclusionary effects by reference to a hypothetical equally efficient competitor.

(172)  This is without prejudice of the possibility, under some specific circumstances, of applying elements of the assessment that are relevant for the analysis of pricing abuses to specific non-pricing abuses. See for example paragraph 158 as regards the applicability of the price-cost test to exclusive dealing.

(173)  For conditional rebates, there are situations where a price-cost test is not appropriate. See paragraphs 94-96 and 145 below.

(174)  The price-cost test (in some instances also referred to as an ‘as efficient competitor’ test) refers ‘to various tests which have in common the aim of assessing the ability of a practice to produce anti-competitive exclusionary effects by reference to the ability of a hypothetical competitor of the undertaking in a dominant position, which is as efficient as the dominant undertaking in terms of cost structure, to offer customers a rate which is sufficiently advantageous to encourage them to switch supplier, despite the disadvantages caused, without that causing that competitor to incur losses’, judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 56. This analysis is carried out having regard to the specificities of the type of pricing conduct at hand. See in this respect paragraph 112 as regards predatory pricing, paragraph 125 as regards margin squeeze and paragraph 144 as regards conditional rebates.

(175)  The assessment does not concern the actual or potential competitors of the dominant undertaking that may be affected by its conduct, see judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraph 343. See also paragraphs 119, 134 and 146 of these Guidelines.

(176)  For instance, the choice of long-run average incremental costs (LRAIC) (which include product-specific sunk costs) as the relevant cost measure for a price-cost test allows accounting for the costs incurred by a hypothetical equally efficient competitor to enter and compete in R&D-driven markets characterised by very low variable and high fixed costs. See Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraph 787; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 442. Dynamic considerations may also inform how the total amount of product-specific development costs is allocated across the lifecycle of a given product for the purpose of carrying out a price-cost test. In the past, the Commission has, for instance, used cost allocation methodologies that account for the fact that the recoupment of product-specific development is not constant but varies over the life cycle of the product; see Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraphs 893-913; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 509-514.

(177)  Judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 264; judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 272-273, 278 and 284.

(178)  See, for example, judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 272, 278 and 284; judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110; judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726; judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011; judgment of 6 December 2012, AstraZeneca v Commission, C-457/10 P, EU:C:2012:770; and judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289.

(179)  Judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 134; judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 127; and judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 164.

(180)  Judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 57; judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 539; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 273 and 278, where the Court of Justice stated that the conduct at stake did not lend itself to a quantitative, price-based analysis. While there is no legal obligation to carry out in all cases a price-cost test, its relevance in non-pricing cases cannot be automatically ruled out, when it is possible to reliably quantify the non-price elements of the conduct, see judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraphs 58 and 59. In light of the enforcement experience of the Commission, it is in general very challenging to reliably quantify the non-pricing elements of a conduct. As such, the Commission considers that the application of a price-cost test to non-pricing behaviour could be considered only in exceptional circumstances. However, if a dominant undertaking submits an analysis based on a price-cost test, the Commission is required to examine its probative value. See judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 60.

(181)  Judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, paragraph 666. On the relevance of actual market developments, see paragraph 106(g) below.

(182)  See paragraph 86 above, and in particular footnote 170.

(183)  Judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 273 and 278.

(184)  Judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 57; judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 59.

(185)  See, by analogy, judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 60.

(186)  Judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 224.

(187)  The presence of other contextual facts or conduct, including conduct of the dominant undertaking itself, which may also increase the likelihood of the exclusionary effects in question does not prevent a finding of abuse in relation to a given conduct. See judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 537, where the General Court ruled that it was irrelevant who established other barriers to entry or expansion that may contribute to the exclusionary effect of a dominant undertaking’s conduct; judgment of 18 November 2020, Lietuvos geležinkeliai v Commission, T-814/17, paragraph 227, which, in view of the competitive situation absent the conduct, considered that the conduct at issue ‘aggravated the situation existing at the time of the suspension of traffic’ and made the repair of the track ‘more difficult’; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 203, 211-213 and 222, where the Court of Justice noted that conduct must be analysed in its context, in particular where ‘different behaviours are interconnected in such a way that they take effect in conjunction with each other, and distinguishing the effects of one from those of the other would be an artificial exercise’.

(188)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 412.

(189)  Judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraph 91; and judgment of 21 February 1973, Europemballage Corporation and Continental Can Company v Commission, C-6/72, EU:C:1973:22, paragraph 27. See also judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraph 354. However, see paragraph 73 of these Guidelines on the relevance of the fact that the conduct relies on the use of resources or means inherent to the holding of a dominant position.

(190)  See, to that effect, judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 225 and 228. See also judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 225, 226, 228 and 343.

(191)  However, not every instance of the existence of factors external to the dominant undertaking will be considered for this purpose, see judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 670.

(192)  Judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 199.

(193)  See, to that effect, judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 231. See also judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 377 and 378.

(194)  Judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 225, 228 and 343, where the Court found that it is permissible for the Commission to rely ‘on a range of evidence, without being required systematically to use any single methodology, in particular a counterfactual analysis, to prove the existence of such a causal link’.

(195)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 98-99. See also judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 203 and 213, where the Court of Justice ruled that when users’ behaviour has been distorted by the conduct, it may be impossible to distinguish such a bias from the intrinsic characteristics of the product or service concerned, and this precludes a finding that users’ choice was, in principle, attributable to the superior performance of the dominant undertaking.

(196)  Judgment of 6 December 2012, AstraZeneca v Commission, C-457/10 P, EU:C:2012:770, paragraph 110; see, to that effect, judgment of 27 June 2024, Commission v Servier, C-176/19 P, EU:C:2024:549, paragraphs 337-358 (in the context of the assessment of the causal link between a practice and its effects under Article 101 TFEU).

(197)  See, in the context of Article 101 TFEU, judgment of 8 September 2016, Lundbeck v Commission, T-472/13, EU:T:2016:449, paragraphs 363, 539 and 572.

(198)  The Court of Justice also refers to the ‘intrinsic capacity’ of the conduct to produce effects: see judgment of 6 September 2017, C-413/14 P, Intel v Commission, EU:C:2017:632, paragraph 140; and judgment of 14 September 2022, T-604/18, Google and Alphabet v Commission (Google Android), EU:T:2022:541, paragraphs 640-641.

(199)  Judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraph 360.

(200)  Judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraph 58. See also, to the same effect, judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 98-99 and 102, where the Court of Justice considered that its conclusion could not be called into question by ‘the actions that competitor undertakings potentially took or could have taken, such as purchasing files from third parties containing information on customers in the protected market, in order to limit the harmful consequences of that practice’.

(201)  Judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraphs 84-87; and judgment of 16 July 2026, RRC Sports, C-209/23, EU:C:2026:597, paragraphs 256, 264 and 265, for a specific scenario where the dominant undertaking was not in competition with those which would be affected by the measure.

(202)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 437.

(203)  See paragraph 106(a) of these Guidelines.

(204)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 129-130; and judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraph 179.

(205)  Judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 190-191, 199 and 206-207, also stressing that theoretical possibilities for users may not be enough to negate the capability of restricting competition.

(206)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 183; judgment of 19 April 2012, Tomra and Others v Commission, C-549/10 P, EU:C:2012:221, paragraph 39; and judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 81.

(207)  Judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 68. For instance, with regard to network effects, see judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 226; judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 273 and 278; and judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 562, 1061 and 1062.

(208)  Commission decision of 16 October 2019 in case AT.40608 – Broadcom, paragraphs 475 and 478.

(209)  Judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 68; judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 139; and judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraphs 44 and 48.

(210)  Judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 696.

(211)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 904, 1038, 1049-1058; Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraphs 399-406 and 1218; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 641-643.

(212)  See paragraphs 113(b) and 142(h) of these Guidelines in relation to predatory pricing and conditional rebates.

(213)  Judgment of 30 January 2020, Generics (UK) and Others, C-307/18, EU:C:2020:52, paragraph 162; judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 63; and judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 45.

(214)  Judgment of 9 September 2010, Tomra Systems and Others v Commission, T-155/06, EU:T:2010:370, paragraph 35; and judgment of 22 March 2012, Slovak Telekom v Commission, T-458/09 and T-171/10, EU:T:2012:145, paragraph 61.

(215)  See, to that effect, judgment of 14 October 2010, Deutsche Telekom v Commission, C-280/08 P, EU:C:2010:603, paragraphs 258-259; judgment of 29 March 2012, Telefónica and Telefónica de España v Commission, T-336/07, EU:T:2012:172, paragraph 402; and judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 1078-1090.

(216)  Judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 345-346, on exit or failed entry of rivals.

(217)  Judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 65.

(218)  See judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 55 and 56.

(219)  Judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraph 58.

(220)  Judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraph 602.

(221)  The applicability of a given analytical framework to specific conduct must be determined in the light of all relevant legal and economic circumstances. See to that effect, judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 166 and 92-140.

(222)  See judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraphs 70-72; judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraph 107; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 618.

(223)  See judgment of 14 November 1996, Tetra Pak v Commission, C-333/94 P, EU:C:1996:436, paragraph 44; and judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraphs 37, 110, 112 and 113. That interpretation does not preclude the Commission from taking the possibility to recoup losses into account as a relevant factor in the assessment of the conduct in question; see judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraph 111.

(224)  Judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 641-643.

(225)  See Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraphs 399-401 and 1060-1066; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 641-643.

(226)  See Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraphs 402-406; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 523.

(227)  See judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraphs 43 and 115; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 523.

(228)  In specific circumstances, predatory pricing may take place and distort competition in an associated non-dominated market, for instance: (a) where there is a close associative link between the markets (demonstrated, for example, by the fact that a significant number of the customers that purchase goods from the undertaking on the market on which it is dominant also purchase goods from that undertaking on the market on which it is not dominant, as well as the fact that the competitors of the undertaking are also active on both markets) and (b) the undertaking has a position on the dominated market that enables it to count on a favoured status on the non-dominated market (demonstrated, for example, by reference to the undertaking’s quasi-monopolistic position on the dominated market as well as its position as a leading supplier on the non-dominated market). See, to that effect, judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraphs 35-45 and 58-62; Commission decision of 14 December 1985 in case IV/30.698 – ECS/AKZO, paragraphs 62-87; and judgment of 14 November 1996, Tetra Pak v Commission, C-333/94 P, EU:C:1996:436, paragraphs 27-31.

(229)  Judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraphs 70-72; and judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraphs 27 and 28.

(230)  Judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraph 71; judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraph 109; and judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraph 27.

(231)  Judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraph 72; judgment of 14 November 1996, Tetra Pak v Commission, C-333/94 P, EU:C:1996:436, paragraph 41; judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraph 109; and judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraph 27.

(232)  Judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraphs 75, 82 and 109; and Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraph 1118.

(233)  Judgment of 30 January 2007, France Télécom v Commission, T-340/03, EU:T:2007:22, paragraphs 198, 199 and 206-209.

(234)  Direct evidence includes, in particular, contemporaneous statements made within the dominant undertaking, such as in emails, letters, presentations, minutes and meeting notes, as well as external statements, such as threats to competitors. See judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraphs 76-82; judgment of 8 October 1996, Compagnie Maritime Belge Transports and Others v Commission, joined cases T-24/93, T-25/93, T-26/93 and T-28/93, EU:T:1996:139, paragraph 147; judgment of 30 January 2007, France Télécom v Commission, T-340/03, EU:T:2007:22, paragraphs 199-209; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 546-552 and 560-579.

(235)  Indirect evidence can be described as a series of important and convergent factors, which indicate the existence of a plan to eliminate or reduce competition as such, and which may relate, in particular, to the duration, the continuity and the scale of the below-cost sales, as well as the targeted nature of the below-cost pricing and the importance of the market (segment) in which it takes place. See judgment of 30 January 2007, France Télécom v Commission, T-340/03, EU:T:2007:22, paragraphs 210-215; judgment of 6 October 1994, T-83/91, Tetra Pak v Commission, EU:T:1994:246, paragraphs 151 and 190; and Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraphs 1138-1146.

(236)  Judgment of 30 January 2007, France Télécom v Commission, T-340/03, EU:T:2007:22, paragraph 197, with reference to judgment of 6 October 1994, T-83/91, Tetra Pak v Commission, EU:T:1994:246, paragraph 151.

(237)  See judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraph 36.

(238)  See judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraph 109, with references to judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraphs 70 and 71 and judgment of 14 November 1996, Tetra Pak v Commission, C-333/94 P, EU:C:1996:436, paragraph 41. See also judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraph 27; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 521, 654, 662, 663 and 694. To date, the European Union courts have never upheld claims by dominant undertakings that a given predatory pricing conduct did not distort competition in the specific circumstances of the case.

(239)  Judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 522 and 627.

(240)  Judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraphs 71-72; and judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraph 108. Whether, in a particular case, a cost element is to be considered as variable, as opposed to being fixed, can depend on the output variation that is subject to the conduct and the time period over which costs are to be assessed. In that respect, the general rule is that the longer the time period and the greater the output variation subject to the conduct, the more likely it is that costs will be classified as variable and the larger will be the resulting variable cost estimate.

(241)  See as regards the use of AAC (in the context of a conditional rebate scheme), judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraph 308, with reference to judgment of 26 January 2022, Intel v Commission, T-286/09 RENV, EU:T:2022:19, paragraphs 157 and 158, and as regards the use of LRAIC (in the context of predatory pricing) judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 435-443; and Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraphs 780-796.

(242)  For example, if the dominant undertaking had to expand capacity in order to be able to predate, then the costs of that extra capacity should be taken into account in looking at the dominant undertaking’s losses.

(243)  For example, the dominant undertaking may be able to sell assets that would not be needed in the absence of the conduct.

(244)  See, to that effect, judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraphs 33-35, notably the method of attributing common costs using percentages.

(245)  Take the example of a store that offers two categories of products (e.g. books and electronic devices). If the store only offered books, some common costs would be incurred in any event (e.g. paying for the managing director) while other common costs would not have been incurred in full (e.g. the store’s surface area would be proportionately smaller, the store would have fewer check-out counters, etc.). Such a change in costs would be included in LRAIC even though the costs are classified (accounting-wise) as common costs. For the case of single product undertakings, LRAIC and ATC are the same, since, by definition, there are no common costs in this instance.

(246)  Commission decision of 4 July 2007 in case COMP/38784 – Wanadoo España v Telefónica, paragraphs 319-320.

(247)  Judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 41, with references to judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraph 74, and judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraph 108. See also judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 422.

(248)  While a dominant undertaking knows what its own costs and charges are, it does not, as a general rule, know what its competitors’ costs and charges are, judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 44; and judgment of 14 October 2010, Deutsche Telekom v Commission, C-280/08 P, EU:C:2010:603, paragraph 202.

(249)  Judgment of 30 January 2007, France Télécom v Commission, T-340/03, EU:T:2007:22, paragraphs 131 and 154.

(250)  Judgment of 30 January 2007, France Télécom v Commission, T-340/03, EU:T:2007:22, paragraphs 131 and 154.

(251)  Judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraphs 99 and 100; and judgment of 6 October 1994, T-83/91, Tetra Pak v Commission, EU:T:1994:246, paragraphs 201 and 202.

(252)  Judgment of 30 January 2007, France Télécom v Commission, T-340/03, EU:T:2007:22, paragraphs 131, 132 and 137.

(253)  Opportunity costs represent profits foregone by not using an asset or product for the best alternative profitable opportunity. If there is clear evidence that the product subject to the investigation would have been sold at a profit in another market absent the conduct, it may be appropriate to account for this profit as an opportunity cost in the price-cost test.

(254)  For example, where a dominant social media platform that operates in a market which is characterised by the fact that users are not charged and the platform is monetised by charging advertisers, it may be appropriate to look at revenues and costs on all sides of the platform.

(255)  For example, in Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), the Commission separately assessed the prices applied by the dominant undertaking to each of its chipsets in question and in relation to each of the two strategically important customers (see paragraphs 399-401 and 939). It might have also been possible to conduct a more aggregate assessment grouping products and/or customers into a single price-cost test. The General Court upheld the decision, noting that the Commission was entitled to limit the analysis to assess the price reductions applied to two crucial customers and in relation to the chipsets sold in a market segment, see judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 641-643. See also Commission decision of 14 December 1985 in case IV/30.698 – ECS/AKZO, paragraphs 82-87, in which the Commission analysed the pricing policy adopted in relation to flour additives in the larger organic peroxides market.

(256)  See judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 334-339; and Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraphs 630-632. See also judgment of 30 January 2007, France Télécom v Commission, T-340/03, EU:T:2007:22, paragraphs 131 and 137, as regards the period chosen over which the costs were spread. Where appropriate, robustness checks can be used to verify the accuracy of the price-cost test, see Commission decision of 18 July 2019 in case AT.39711 – Qualcomm (Predation), paragraphs 1007-1033.

(257)  This includes a situation in which the input is at the same level as or downstream from the market for which it is needed. This may, for instance, happen where one undertaking controls a downstream distribution network needed to access customers.

(258)  Judgment of 14 October 2010, Deutsche Telekom v Commission, C-280/08 P, EU:C:2010:603, paragraph 178; and judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 70.

(259)  Judgment of 18 December 2025, ‘Lukoil Bulgaria’ EOOD v Komisia za zashtita na konkurentsiata, C-260/24, EU:C:2025:988, paragraphs 33 and 35.

(260)  The actual form of integration of the vertically integrated undertaking (e.g. sole vertically integrated undertaking, different divisions, separate companies controlled by the same group) is not relevant.

(261)  Judgment of 18 December 2025, ‘Lukoil Bulgaria’ EOOD v Komisia za zashtita na konkurentsiata, C-260/24, EU:C:2025:988, paragraph 41; judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 89; and judgment of 29 March 2012, Telefónica and Telefónica de España v Commission, T-336/07, EU:T:2012:172, paragraph 146.

(262)  Judgment of 14 October 2010, Deutsche Telekom v Commission, C-280/08 P, EU:C:2010:603, paragraphs 169 and 183; judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 32; and judgment of 25 March 2021, Slovak Telekom v Commission, C-165/19 P, EU:C:2021:239, paragraph 73.

(263)  Judgment of 18 December 2025, ‘Lukoil Bulgaria’ EOOD v Komisia za zashtita na konkurentsiata, C-260/24, EU:C:2025:988, paragraph 31; judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 34; and judgment of 29 March 2012, Telefónica and Telefónica de España v Commission, T-336/07, EU:T:2012:172, paragraph 187.

(264)  Judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 100.

(265)  See judgment of 18 December 2025, ‘Lukoil Bulgaria’ EOOD v Komisia za zashtita na konkurentsiata, C-260/24, EU:C:2025:988, paragraphs 41 and 43-46; judgment of 25 March 2021, Slovak Telekom v Commission, C-165/19 P, EU:C:2021:239, paragraph 73; and judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraphs 31-32, 61, 63 and 73.

(266)  Based on the Commission’s experience, the elements to consider may concern the coverage and duration of the margin squeeze or the importance of the input.

(267)  Judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraphs 55, 56 and 72; judgment of 10 July 2014, Telefónica and Telefónica de España v Commission, C-295/12 P, EU:C:2014:2062, paragraphs 75 and 96; and judgment of 25 March 2021, Slovak Telekom v Commission, C-165/19 P, EU:C:2021:239, paragraph 52.

(268)  See, to that effect, judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraphs 70 and 71.

(269)  Judgment of 14 October 2010, Deutsche Telekom v Commission, C-280/08 P, EU:C:2010:603, paragraphs 200-204; and judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraphs 31-34.

(270)  See e.g. Commission decision of 4 July 2007 in AT.38784 – Wanadoo España v Telefónica, paragraph 318. See paragraph 118 above for an explanation of LRAIC.

(271)  Judgment of 14 October 2010, Deutsche Telekom v Commission, C-280/08 P, EU:C:2010:603, paragraphs 198 and 200-202; and judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraphs 41, 42 and 44.

(272)  See judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraphs 44 and 45.

(273)  Judgment of 29 March 2012, Telefónica and Telefónica de España v Commission, T-336/07, EU:T:2012:172, paragraph 209.

(274)  In past Commission cases, the application of the price-cost test involved the aggregation of all the products that a hypothetical equally efficient competitor could supply on the downstream market by using the upstream input of the dominant undertaking. See, for example, Commission decision of 15 October 2014 in case AT.39523 – Slovak Telekom, paragraph 831-832; and Commission decision of 4 July 2007 in case AT.38784 – Wanadoo España v Telefónica, paragraphs 386-388.

(275)  See footnote 297.

(276)  See for instance paragraph 198(a).

(277)  Judgment of 9 November 1983, Nederlandsche Banden-Industrie-Michelin v Commission (Michelin I), C-322/81, EU:C:1983:313 paragraph 73; judgment of 15 March 2007, British Airways v Commission, C-95/04 P, EU:C:2007:166, paragraph 67; and judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 29.

(278)  Judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraphs 30 and 39.

(279)  See the case-law referred to in paragraph 106(d) of these Guidelines.

(280)  Judgment of 15 March 2007, British Airways v Commission, C-95/04 P, EU:C:2007:166, paragraph 75; and judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 40.

(281)  Judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraphs 97-100; and judgment of 7 October 1999, Irish Sugar v Commission, EU:T:1999:246, T-228/97, paragraphs 207-214.

(282)  For example, a lack of transparency can put pressure on customers and make it less attractive to switch to a competitor where the effects of complying or failing with the conditions that govern the rebate scheme are uncertain. Judgment of 9 November 1983, Nederlandsche Banden-Industrie-Michelin v Commission (Michelin I), C-322/81, EU:C:1983:313, paragraphs 83 and 84.

(283)  Judgment of 9 November 1983, Nederlandsche Banden-Industrie-Michelin v Commission (Michelin I), 322/81, EU:C:1983:313, paragraph 81; and judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraphs 32 and 33.

(284)  Judgment of 9 September 2010, Tomra Systems and Others v Commission, T-155/06, EU:T:2010:370, paragraphs 261, 262 and 269.

(285)  See by analogy judgment of 3 July 1991, AKZO v Commission, C-62/86, EU:C:1991:286, paragraph 72; judgment of 14 November 1996, Tetra Pak v Commission, C-333/94 P, EU:C:1996:436, paragraph 41; judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraph 109; and judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraph 27.

(286)  Judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraphs 57 and 61; judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraphs 58 and 62; judgment of 12 May 2022, Servizio Elettrico Nazionale, C-377/20, EU:C:2022:379, paragraph 81; and judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 643. See also, by analogy, judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraph 181.

(287)  The price-cost test is typically based on elements referred to at letters (c), (d), (e), (f) and (g) in paragraph 142 above.

(288)  See, by analogy, judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraph 278.

(289)  Judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 61; see also, by analogy, judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraphs 57 and 58; and judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraph 181.

(290)  See, by analogy, judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 57. In the Commission’s experience, converting non-monetary inducements to monetary amounts can be difficult or impossible; see Commission decision of 16 October 2019 in case AT.40608 – Broadcom, paragraph 352. In cases where it is possible to reliably determine the monetary amount of the inducement, the assessment must be done, in principle, by taking into account a hypothetical competitor with the cost structure of the dominant undertaking.

(291)  See also paragraph 119.

(292)  This effective price (per unit) is not the average price (per unit) applied by the dominant undertaking, but the list price (per unit) less the total value of the rebate that the customer loses by switching the contestable units to the competitor, distributed over the contestable units in the relevant period of time.

(293)  The lack of sufficient data, for example for a given period affected by the conduct, does not necessarily prevent the calculation of the contestable demand if the unknown data are inferred from representative data that are known and such an inference is based on a concrete logical pattern, which is clearly set out. See judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraphs 270 and 271.

(294)  The complexity of data, the imperfect data quality, the existence of different data sources, lack of perfect information, or subjectivity in assumptions may lead to different estimates of the contestable demand, such that the use of one or another estimate can alter the result of a price-cost test (i.e., from negative to positive, or vice versa).

(295)  Judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraph 89.

(296)  All references to ‘exclusive’, ‘exclusivity’ or ‘exclusively’ in this section equally apply to situations where the purchase or supply obligation or the incentive scheme relates to most rather than all of a customer’s demand or supplier’s supply (see judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraph 89; and judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 137). For example, a rebate conditional on customers purchasing 75 % of their requirements from a dominant undertaking has been held to be an exclusivity rebate, see judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraph 83.

(297)  Such advantages may consist of price advantages such as discounts, rebates, payments, or bonuses, or non-price advantages, such as technical support, free-of-charge upgrades or installations, or early access to a technology. See Commission decision of 29 March 2006 in case COMP/E-1/38.113 – Prokent-Tomra, paragraph 317; and Commission decision of 16 October 2019 in case AT.40608 – Broadcom, paragraph 364(b)(2) and (3), and footnote 269.

(298)  The exclusivity conditions may result from contractual agreements or arrangements, as well as emails or other correspondence between the customer and the dominant undertaking, including oral arrangements. For the avoidance of doubt, these also include situations where a certain rebate or advantage is linked to exclusivity.

(299)  Judgment of 23 October 2003, Van den Bergh Foods v Commission, T-65/98, EU:T:2003:281, paragraph 160; and judgment of 9 September 2010, Tomra Systems and Others v Commission, T-155/06, EU:T:2010:370, paragraphs 297 and 298.

(300)  Stocking requirements refer to obligations to reserve a given space only for the products of the dominant undertaking, in a way that in practice excludes the possibility for competitors’ products to be offered to customers. See, for example, judgment of 23 October 2003, Van den Bergh Foods v Commission, T-65/98, EU:T:2003:281, paragraphs 159 and 160, where the obligation to use freezers exclusively for the products of the dominant undertaking was considered to lead de facto to outlet exclusivity.

(301)  Judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraph 90; judgment of 9 September 2010, Tomra Systems and Others v Commission, T-155/06, EU:T:2010:370, paragraphs 209 and 210; judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraphs 46 and 49; and judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, paragraph 581.

(302)  Judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 51.

(303)  Judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraph 89; judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 137; and judgment of 26 January 2022, Intel v Commission, T-286/09 RENV, EU:T:2022:19, paragraph 124.

(304)  Judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraph 89; judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 137.

(305)  Judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 138; judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraphs 47 and 52; and judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, paragraphs 380 and 381.

(306)  See paragraph 161(b); judgment of 26 January 2022, Intel v Commission, T-286/09 RENV, EU:T:2022:19, paragraph 521 and 524-527 and judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraphs 90 and 337-341.

(307)  See, to that effect, judgment of 15 June 2022, Qualcomm v Commission (Exclusivity payments), T-235/18, EU:T:2022:358, paragraphs 412-417.

(308)  For a definition of ‘contestable’ and ‘non-contestable’ demand, see paragraph 148.

(309)  Judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraphs 59 and 60.

(310)  See also judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 56.

(311)  See also judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 59.

(312)  For instance, this is the case when the rebate is conditioned on the purchase of a minimum share of the buyer’s total purchases over the reference period from the dominant undertaking (as opposed to a minimum volume). When the rebate is conditioned on a minimum share of purchases, an increase in purchases from the rival, even without any decrease in the volume purchased from the dominant firm, would still put at risk that the buyer meets the requirements for the rebate. A minimum share requirement in a rebate scheme hence leads to a stronger disincentive for the buyer to purchase from rivals than a corresponding minimum volume requirement, which can lead to exclusion even without below-cost pricing by the dominant undertaking. Another scenario concerns industries presenting economies of scale and a customer base that is sufficiently fragmented. In these cases, each customer may not have a sufficient size to individually allow a smaller rival or new entrant to gain sufficient scale. This can lead to a coordination problem among customers to switch sufficient volume to the new entrant, which the dominant undertaking can exploit by signing up customers via exclusivity agreements that do not involve a pricing below cost, but which nevertheless deny the rival scale.

(313)  See paragraphs 94-96.

(314)  Judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraph 54.

(315)  See, to that effect, judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 428; judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 206-207.

(316)  See, for example, the scenarios outlined in paragraphs 157 and 158.

(317)  Judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 139. Other factors referred to in paragraph 106 of these Guidelines may also be relevant for the purposes of this assessment.

(318)  Judgment of 19 April 2012, Tomra and Others v Commission, EU:C:2012:221, C-549/10 P, paragraph 39; and Commission decision of 20 March 2019 in case AT.40411 – Google Search (AdSense), paragraph 364 (annulled by the General Court in judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, but not on this point).

(319)  Judgment of 13 February 1979, Hoffmann-La Roche v Commission, C-85/76, EU:C:1979:36, paragraphs 41 and 133; judgment of 23 October 2003, Van den Bergh Foods v Commission, T-65/98, EU:T:2003:281, paragraphs 154 and 156; judgment of 9 September 2010, Tomra Systems and Others v Commission, T-155/06, EU:T:2010:370, paragraph 269, and judgment of 19 April 2012, Tomra and Others v Commission, EU:C:2012:221, C-549/10 P, paragraph 79. See as an example Commission decision of 16 October 2019 in case AT.40608 – Broadcom, paragraph 365. See also paragraph 157(c) above.

(320)  Similarly, for exclusive supply obligations, the relevant element is the share of suppliers and the overall share of supply covered by the obligation; see judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, paragraphs 602 and 603.

(321)  Judgment of 19 April 2012, Tomra and Others v Commission, EU:C:2012:221, C-549/10 P, paragraphs 43-46; and judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 139. See also Commission decision of 16 October 2019 in case AT.40608 – Broadcom, paragraphs 366, 385-389; and Commission decision of 20 March 2019 in case AT.40411 – Google Search (AdSense), paragraphs 381-385 (annulled by the General Court in judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, but not on this point).

(322)  Judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 696.

(323)  The scope and intensity of this assessment may vary depending on the circumstances of the case. See Commission decision of 16 October 2019 in case AT.40608 – Broadcom, paragraph 367 (see also 382, 383, 384, 390 and 391), where the duration of one to three years was considered long, in view of the characteristics of the industry and the presence of automatic renewal clauses. See also judgment of 23 October 2003, Van den Bergh Foods v Commission, T-65/98, EU:T:2003:281, paragraph 105; and order of 28 September 2006, Unilever Bestfoods v Commission, C-552/03 P, EU:C:2006:607, paragraph 55.

(324)  Commission decision of 16 October 2019 in case AT.40608 – Broadcom, paragraphs 364(c) and 381-383; and judgment of 23 October 2003, Van den Bergh Foods v Commission, T-65/98, EU:T:2003:281, paragraphs 92 and 93.

(325)  Judgment of 9 September 2010, Tomra Systems and Others v Commission, T-155/06, EU:T:2010:370, paragraph 260; Commission decision of 16 October 2019 in case AT.40608 – Broadcom, paragraph 364(b).

(326)  See also footnote 312.

(327)  The European Union courts have referred to this factor as a ‘strategy aiming to exclude competitors that are at least as efficient as the dominant undertaking from the market’. The fact that such element may or may not be relevant, depending on the circumstances of the case, has been acknowledged by the European Union courts’ references to its ‘possible existence’, see judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 139; judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraphs 144, 145 and 181; judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraphs 48 and 50; and judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, paragraphs 678, 679, 883 and 981.

(328)  See paragraph 158 above.

(329)  Judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651,paragraphs 57 and 61; judgment of 19 January 2023, Unilever Italia Mkt Operations, C-680/20, EU:C:2023:33, paragraphs 58 and 62; judgment of 12 May 2022, Servizio Elettrico Nazionale, C-377/20, EU:C:2022:379, paragraph 81; judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 643; see also, by analogy, judgment of 24 October 2024, Commission v Intel, C-240/22 P, EU:C:2024:915, paragraph 181.

(330)  Judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 269.

(331)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraph 963.

(332)  Judgment of 6 October 1994, Tetra Pak International v Commission, T-83/91, EU:T:1994:246, paragraph 34.

(333)  Conversely, in the context of tying, the tied product may also be offered stand-alone.

(334)  Judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 283. For example, tying can lead to efficient product integration, lower production costs and lower transaction costs, as well as increased convenience and quality improvements for consumers. It remains possible for the dominant undertaking to raise these benefits as objective justifications, see Section 5.

(335)  See judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 1046, 1047, 1069 and 1070.

(336)  For instance, this may be the case when entry into the market of the tied product is a precursor for entry into the market of the tying product. The dominant undertaking may even face the threat of entry on both the tied and tying product markets. In that case, tying and bundling can restrict competition for the tied product but also make successful entry in either market dependent on simultaneous successful entry in the other market, thereby increasing the risk or costs for potential entrants.

(337)  Judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 284; and judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 842, 859, 862, 864, 867 and 869.

(338)  See Section 2. In bundling cases, the undertaking needs to be dominant in one of the markets concerned. In the case of tying in after-markets, the undertaking must be dominant in the tying market.

(339)  See also Section 3.3 on the notion of capability to have exclusionary effects. In addition, see judgment of 14 November 1996, Tetra Pak v Commission, C-333/94 P, EU:C:1996:436, paragraph 27, for the specific case of closely associated markets.

(340)  Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraph 744.

(341)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraph 925.

(342)  For instance, evidence that customers purchase the tying and the tied products separately from different sources of supply, or that there are companies specialising in the manufacture and sale of the tied product on an autonomous basis, see judgment of 12 December 1991, Hilti v Commission, T-30/89, EU:T:1991:70, paragraph 67; judgment of 14 November 1996, Tetra Pak v Commission, C-333/94 P, EU:C:1996:436, paragraph 36; and judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraph 927.

(343)  See judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 925 and 928-931.

(344)  See the Market Definition Notice.

(345)  Judgment of 12 December 1991, Hilti v Commission, T-30/89, EU:T:1991:70, paragraph 66. However, the definition of the relevant market has a different purpose than the assessment of the ‘separate product’ criterion in the context of tying. The definition of the relevant product market serves to determine whether two products are sufficiently close substitutes for them to be placed in the same product market. The decisive criterion to establish whether two products are ‘separate products’ for the purpose of tying is not whether they are close substitutes (in fact, often they are complementary products and therefore not substitutes), but whether there is independent demand for the tied product without the tying product.

(346)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 921, 922 and 932.

(347)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraph 935.

(348)  Judgment of 14 November 1996, Tetra Pak v Commission, C-333/94 P, EU:C:1996:436, paragraph 37; and judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraph 942.

(349)  Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraphs 858 and 865.

(350)  Judgment of 12 December 1991, Hilti v Commission, T-30/89, EU:T:1991:70, paragraphs 16 and 100.

(351)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraph 962.

(352)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 967-969.

(353)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraph 970. Accordingly, tying does not require proof that customers are obliged to use the tied product or to purchase it on an exclusive basis. Coercion can also be reinforced by the fact that it is not possible to uninstall the tied product; see judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraph 963. Granting customers the possibility to uninstall or to opt out from the tied product after they have obtained it together with the tying product is generally insufficient to alter the conclusion that coercion exists. The impact of such possibilities is typically considered when assessing whether the tie is capable of producing exclusionary effects; see Commission decision of 12 September 2025 in case AT.40721 – Microsoft Teams and AT.40873 – Microsoft Teams II, paragraph 124.

(354)  Judgment of 12 December 1991, Hilti v Commission, T-30/89, EU:T:1991:70, paragraphs 16 and 101.

(355)  However, in general, the fewer the customers that are subject to coercion, the less the tie is capable of having exclusionary effects; see also paragraph 106(d) of these Guidelines.

(356)  Judgment of 14 November 1996, Tetra Pak v Commission, C-333/94 P, EU:C:1996:436, paragraph 25.

(357)  Judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 283.

(358)  Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraph 753.

(359)  See, for instance, judgment of 12 December 1991, Hilti v Commission, T-30/89, EU:T:1991:70, in relation to the requirement that users of Hilti‘s patented nail cartridges should also buy nails; and judgment of 6 October 1994, Tetra Pak International v Commission, T-83/91, EU:T:1994:246, in relation to the requirement by Tetra Pak that buyers of liquid packaging machines should also buy cartons and maintenance services from it.

(360)  Judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraphs 292-295; and Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraphs 1013 and 1014.

(361)  See judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraphs 296 and 297; and Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraphs 1017-1048.

(362)  This is the case, for instance, when, due to its distribution together with the tying product, the tied product achieves such a level of market penetration that competitors are unable to counterbalance it or match it with alternative means to reach customers or end users. See judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 1036-1039; Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraphs 805 and 811; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 413. See also Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraph 829.

(363)  Conversely, if the tying concerns two entirely unrelated products, the link between the tying and the tied product may be weaker and exclusionary effects less likely.

(364)  Commission decision of 20 December 2012 in case AT.39230 – Rio Tinto Alcan, paragraph 73; Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraphs 833, 837 and 839.

(365)  Commission decision of 20 December 2012 in case AT.39230 – Rio Tinto Alcan, paragraphs 68 and 69. Where the dominant undertaking is also dominant on the market for the tied product, tying can help to maintain and strengthen its dominant position in that market; see Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraph 858.

(366)  Commission decision of 20 December 2012 in case AT.39230 – Rio Tinto Alcan, paragraphs 70-72; Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraphs 859 and 860; and Commission decision of 14 November 2024 in case AT.40684 – Facebook Marketplace, paragraphs 981, 982 and 994.

(367)  Commission decision of 21 April 2004 in case COMP/C-3/37.792 – Microsoft, paragraphs 878-896 and 980; and Commission decision of 16 December 2009 in case COMP/39.530 – Microsoft (Tying), paragraphs 55 and 56.

(368)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 1041 and 1042; judgment of 14 September 2022, Google Android, T-604/18, EU:T:2022:541, paragraphs 583 and 593; judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 203-210; Commission decision of 16 December 2009 in case COMP/39.530 – Microsoft (Tying), paragraphs 47-54. This factor may be less present if the tying targets business customers.

(369)  In the context of access restrictions and refusal to supply, the term ‘input’ refers to various types of assets, for example, goods, services, infrastructure, networks, intellectual property rights or data. In some cases, the input may not exist ‘as such’ in a directly accessible form in the market. Steps may be needed to create it, which may include, for example, drawing up interoperability information which was not drawn up in an accessible way before (judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 249 and 807); or making changes to computer systems, or detailed preparations and numerous series of tests (judgment of 9 September 2009, Clearstream v Commission, T-301/04, EU:T:2009:317, paragraph 106).

(370)  Judgment of 25 March 2021, Slovak Telekom v Commission, C-165/19 P, EU:C:2021:239, paragraphs 50 and 51; and judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraphs 44-51.

(371)  Judgment of 25 March 2021, Slovak Telekom v Commission, C-165/19 P, EU:C:2021:239, paragraph 50.

(372)  Judgment of 6 March 1974, Istituto Chemioterapico Italiano and Commercial Solvents Corporation v Commission, joined cases C-6/73 and C-7/73, EU:C:1974:18, paragraph 25; and judgment of 16 September 2008, Sot. Lélos kai Sia EE and Others v GlaxoSmithKline AEVE Farmakeftikon Proïonton, joined cases C-468/06 to C-478/06, EU:C:2008:504, paragraph 77.

(373)  Judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraphs 47 and 48.

(374)  Judgment of 18 December 2025, Lukoil Bulgaria and Lukoil Neftohim Burgas, C-245/24, EU:C:2025:987, paragraph 50; judgment of 13 December 2018, Slovak Telekom v Commission, T-851/14, EU:T:2018:929, paragraph 121; and judgment of 25 March 2021, Slovak Telekom, C-165/19 P, EU:C:2021:239, paragraphs 54-60.

(375)  Judgment of 25 March 2021, Slovak Telekom, C-165/19 P, EU:C:2021:239, paragraph 50; and judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 111 and 112.

(376)  Alternatively, the dominant undertaking may have a right over the input that ensures ‘an equivalent degree of control’. However, where the dominant undertaking’s decision-making autonomy regarding access to the input is limited by prerogatives or obligations imposed by legislation, regulations or contracts, that undertaking cannot be regarded as ‘owning’ the input. In this latter case, the general framework outlined in Section 3 applies (see also paragraph 182(c) above). Judgment of 18 December 2025, Lukoil Bulgaria and Lukoil Neftohim Burgas, C-245/24, EU:C:2025:987, paragraph 50.

(377)  If the input was acquired by the dominant undertaking at a price and under the conditions resulting from a competitive procedure, the input is akin to input established or developed by that dominant undertaking. Judgment of 18 December 2025, Lukoil Bulgaria and Lukoil Neftohim Burgas, C-245/24, EU:C:2025:987, paragraph 55.

(378)  Judgment of 25 March 2021, Slovak Telekom v Commission, C-165/19 P, EU:C:2021:239, paragraph 45; judgment of 12 January 2023, Lietuvos geležinkeliai v Commission, C-42/21 P, EU:C:2023:12, paragraph 79; and judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraph 47. A refusal to supply can be found even if the input has not yet been traded. It is sufficient for there to be a potential or hypothetical market for the input, as indicated by demand from potential purchasers; see the judgment of 29 April 2004, IMS v NDC Health, C-418/01, EU:C:2004:257, paragraphs 43 and 44.

(379)  Judgment of 25 March 2021, Slovak Telekom v Commission, C-165/19 P, EU:C:2021:239, paragraphs 46 and 47; and judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraphs 41 and 42.

(380)  Judgment of 26 November 1998, Bronner, C-7/97, EU:C:1998:569, paragraph 41; judgment of 9 September 2009, Clearstream v Commission, T-301/04, EU:T:2009:317, paragraph 147; judgment of 25 March 2021, Slovak Telekom, C-165/19 P, EU:C:2021:239, paragraph 44; and judgment of 12 January 2023, Lietuvos geležinkeliai v Commission, C-42/21 P, EU:C:2023:12, paragraph 79. The European Union courts have also specified that the refusal must not be objectively justified. For the assessment of objective justifications, see Section 5.

(381)  Judgment of 25 March 2021, Deutsche Telekom v Commission, C-152/19 P, EU:C:2021:238, paragraphs 48 and 49; and judgment of 25 March 2021, Slovak Telekom v Commission, C-165/19 P, EU:C:2021:239, paragraphs 48 and 49.

(382)  Judgment of 26 November 1998, Bronner, C-7/97, EU:C:1998:569, paragraphs 41 and 44-46; judgment of 29 April 2004, IMS v NDC Health, C-418/01, EU:C:2004:257, paragraph 28; judgment of 9 September 2009, Clearstream v Commission, T-301/04, EU:T:2009:317, paragraph 147; and judgment of 25 March 2021, Slovak Telekom, C-165/19 P, paragraph 49. See also Commission decision of 17 January 2024 in case AT.40735 – Online rail ticket distribution in Spain, paragraphs 107-109.

(383)  The time required for the duplication of the input is also relevant in determining this. See, for example, Commission decision of 24 March 2004 in case COMP/C-3/37.792 – Microsoft, paragraph 685.

(384)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 229 and 457.

(385)  Judgment of 26 November 1998, Bronner, C-7/97, EU:C:1998:569, paragraph 43; and judgment of 29 April 2004, IMS v NDC Health, C-418/01, EU:C:2004:257, paragraph 28.

(386)  Judgment of 26 November 1998, Bronner, C-7/97, EU:C:1998:569, paragraph 41; judgment of 12 January 2023, Lietuvos geležinkeliai v Commission, C-42/21 P, EU:C:2023:12, paragraph 79; and judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 229, 332 and 563. See Commission decision of 17 January 2024 in case AT.40735 – Online rail ticket distribution in Spain, paragraphs 115, 116, 118 and 119.

(387)  Judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 563 and 593. See also Commission decision of 11 July 2024 in case AT.40452 – Apple – Mobile Payments, paragraph 74.

(388)  Judgment of 5 October 1988, AB Volvo v Erik Veng, C-238/87, EU:C:1988:477; judgment of 6 April 1995, RTE v Commission, joined cases C-241/91 P and C-242/91 P, EU:C:1995:98; and judgment of 29 April 2004, IMS v NDC Health, C-418/01, EU:C:2004:257.

(389)  Judgment of 6 April 1995, RTE v Commission, joined cases C-241/91 P and C-242/91 P, EU:C:1995:98, paragraph 54; and judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 647-665.

(390)  Judgment of 29 April 2004, IMS v NDC Health, C-418/01, EU:C:2004:257, paragraph 49.

(391)  Conducts described in other sections of these Guidelines may also imply some degree of favouring by the dominant undertaking, such as margin squeeze and tying. See, to that effect, judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 163. Where the behaviour of a dominant undertaking clearly has the constituent elements of one of the types of conduct outlined in another section of these Guidelines, the specific guidance provided in that section will in principle apply.

(392)  Judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 186 and 187.

(393)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 163, 237, 240 and 279; and judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 113, 180 and 267.

(394)  For the notion of digital ecosystems, see footnote 78. See also judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 149.

(395)  For instance, where customers of the product in one market use it to transact with customers of the product in the other market, and vice versa. See Commission decision of 5 September 2025 in case AT.40670 – Google – Adtech and Data-related practices, concerning practices in the display advertising intermediation value chain, where publishers and advertisers used intermediation services in the respective markets to match their offers and bids.

(396)  Commission decision of 27 June 2017 in case AT.39740 – Google Shopping, where comparison shopping services were using and relying on the Google search engine and its search results for their traffic.

(397)  See, to that effect, Commission decision of 27 June 2017 in case AT.39740 – Google Shopping, section 7.2.1 and judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 283. See also Commission decision of 20 December 2022 in case AT.40462 – Amazon Marketplace and AT.40703 – Amazon Buy Box, paragraph 203.

(398)  See, by analogy, judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 96-99.

(399)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 187 and 329. While in some instances the preferential treatment of the products of the dominant undertaking may be combined with the less favourable treatment of the products of competitors, based on the Commission’s experience the latter is not necessary to establish the existence of self-preferencing.

(400)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 162, 164 and 175, confirmed on appeal by judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 171, 172 and 186.

(401)  These elements may be relevant both for the assessment of whether the conduct departs from competition on the merits and is capable of having exclusionary effects, see paragraph 61 above and judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 170 and 171.

(402)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 170, 171 and 174. The conditions set out in the judgment of 26 November 1998, Bronner, 7/97, EU:C:1998:569 do not apply because self-preferencing is a different type of abuse from a refusal to supply; see judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 230 and 240, confirmed on appeal by judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 110-114. However, a finding of indispensability may provide a strong indication that the conduct amounts to an abuse; see, by analogy, judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraphs 70 and 71. The fact that the leveraging market is an important source of business for competitors that cannot be effectively replaced may also indicate the capability of self-preferencing to produce exclusionary effects; see judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 454; judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 170-171; Commission decision of 27 June 2017 in case AT.39740 – Google Shopping, paragraphs 591-592; and Commission decision of 20 December 2022 in case AT.40462 – Amazon Marketplace and case AT.40703 – Amazon Buy Box, paragraphs 173–176.

(403)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 155, 197, and 240, confirmed on appeal by judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraphs 146, 158 and 187.

(404)  For instance, the conduct of the dominant undertaking in the leveraging market is contrary to its interests in that market or to those of its customers in that same market or does not correspond to the market standard. Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 176-185, noted by the General Court for the sake of completeness; and Commission decision of 5 September 2025 in case AT.40670 – Google – Adtech and Data-related practices, sections 8.3.4.1 and 8.4.4.1.

(405)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 176-184, discussing the open nature of a search engine, noted by the General Court for the sake of completeness; see, by analogy, judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraphs 49 and 52, concerning the development of an open platform.

(406)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 172. The possible influence of self-preferencing on user behaviour is relevant in particular in those instances where the conduct relies on the behavioural bias of users. See also, to that effect, judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 213 and 226.

(407)  The European Union courts have referred to such conduct as ‘naked restrictions’; see judgment of 26 January 2022, Intel Corp. v Commission, T-286/09 RENV, EU:T:2022:19, paragraph 96, readopting the finding made in the judgment of 12 June 2014, Intel v Commission, T-286/09, EU:T:2014:547, paragraphs 209-210; or as abuses having the ‘object’ of restricting competition; see judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 131 and 185; judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping), C-48/22 P, EU:C:2024:726, paragraph 167; and judgment of 15 October 2025, Red Bull v Commission, T-306/23, EU:T:2025:959, paragraphs 96-98.

(408)  Judgment of 12 June 2014, Intel v Commission, T-286/09, EU:T:2014:547, paragraph 32; judgment of 26 January 2022, Intel Corp. v Commission, T-286/09 RENV, EU:T:2022:19, paragraphs 91, 96 and 531; and judgment of 10 December 2025, Intel v Commission, T-1129/23, EU:T:2025:1091, paragraphs 23-24.

(409)  See, to that effect, judgment of 7 October 1999, Irish Sugar v Commission, T-228/97, EU:T:1999:246, paragraphs 228-234; and judgment of 12 June 2014, Intel v Commission, T-286/09, EU:T:2014:547, paragraphs 199 and 209-219.

(410)  See, to that effect, judgment of 12 January 2023, Lietuvos geležinkeliai v Commission, C-42/21 P, EU:C:2023:12, paragraphs 83-84 and 89-91. See also, to that effect, Commission decision of 2 October 2017 in case AT.39813 – Baltic rail, paragraph 33.

(411)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 133, 147,148 and 185.

(412)  See judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 119, 186 and 201; and judgment of 5 December 2024, Tallinna Kaubamaja Grupp and KIA Auto, C-606/23, EU:C:2024:1004, paragraphs 35 and 36.

(413)  This would mean, among other things, that the actual or potential effects of the conduct on competition are not relevant for its categorisation as an abuse ‘by object’, see judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 119, 159, 166 and 186; judgment of 27 June 2024, Commission v Servier, C-176/19 P, EU:C:2024:549, paragraphs 288 and 289; judgment of 5 December 2024, Tallinna Kaubamaja Grupp and KIA Auto, C-606/23, EU:C:2024:1004, paragraphs 35 and 36.

(414)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 205 and 209. Moreover, the possibility to invoke an objective justification is without prejudice to the application of the case-law of the European Union courts according to which conduct that is not anti-competitive in nature may fall outside the scope of Article 102 TFEU because it is justified by legitimate objectives in the public interest, judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 183 and 185.

(415)  Judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraphs 40 and 41; judgment of 30 January 2020, Generics (UK) and Others, C-307/18, EU:C:2020:52, paragraph 165; judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 84; and judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 201 and 202. The examples provided in this section are not exhaustive of the possible objective justifications that can be raised in Article 102 TFEU cases.

(416)  Judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 49; judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 688 and 1144; judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 554; judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraph 71; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 361. It is for the undertaking invoking the benefit of that defence against the finding of an infringement to demonstrate that the conditions for applying the defence are satisfied; see to that effect, Council Regulation (EC) No 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the EC [now Articles 101 and 102 TFEU] (OJ L 1, 4.1.2003, p. 1, ELI: http://data.europa.eu/eli/reg/2003/1/2009-07-01), recital 5 and Article 2.

(417)  See, to that effect, judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraph 686 (see also paragraphs 688-689 and 693, indicating that the claim of an objective justification was not credible, in view of the lack of any internal document to support the alleged justification and the inconsistency of AstraZeneca’s conduct in different Member States); and judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 577.

(418)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 205; and judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraphs 577 and 591.

(419)  Judgment of 30 January 2020, Generics (UK) and Others, C-307/18, EU:C:2020:52, paragraph 166.

(420)  Judgment of 30 January 2020, Generics (UK) and Others, C-307/18, EU:C:2020:52, paragraph 168.

(421)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 206; and judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraph 79.

(422)  See, to that effect, judgment of 3 October 1985, CBEM v CLT and IPB, C-311/84, EU:C:1985:394, paragraphs 26 and 27.

(423)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 552.

(424)  Judgment of 16 September 2008, Sot. Lélos kai Sia EE and Others v GlaxoSmithKline AEVE Farmakeftikon Proïonton, C-468/06 to C-478/06, EU:C:2008:504, paragraph 70.

(425)  Judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraph 73.

(426)  Judgment of 14 February 1978, United Brands v Commission, C-27/76, EU:C:1978:22, paragraphs 182-187.

(427)  Judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 594-603.

(428)  Judgment of 7 October 1999, Irish Sugar v Commission, T-228/97, EU:T:1999:246, paragraphs 111-112 and 190; and judgment of 2 April 2009, France Télécom v Commission, C-202/07 P, EU:C:2009:214, paragraphs 47 and 56.

(429)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 552; judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraph 73; and judgment of 17 September 2007, Microsoft v Commission, T-201/04, EU:T:2007:289, paragraphs 1146 and 1159, where the argument that the integration of Windows Media Player in Windows created technical efficiencies or led to ‘superior technical product performance’ was rejected.

(430)  To assess whether the period of time required by the dominant undertaking for the implementation of the requested interoperability is objectively necessary, the relevant factors are: (i) the degree of technical difficulty in developing the solution that allows the interoperability, (ii) constraints related to the fact that it is impossible for the dominant undertaking to equip itself, within a short time, with the resources, in particular human resources, necessary to develop the technical solution, and (iii) constraints external to the dominant undertaking that have an impact on its ability to develop the technical solution, such as constraints resulting from the applicable regulatory framework. See judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraphs 73-75. The fact that the dominant undertaking demands an appropriate financial contribution from the undertaking that has requested interoperability does not preclude a defence based on objective necessity. Such a contribution must be fair and proportionate, allowing the dominant undertaking to derive an appropriate benefit, considering the actual cost of developing the solution. See judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraph 76.

(431)  Judgment of 25 February 2025, Alphabet and Others, C-233/23, EU:C:2025:110, paragraph 77.

(432)  Public interest objectives can be considered genuine and objectively defined if they are set by legislation, soft-law or impartial standardisation processes.

(433)   Commission decision of 22 July 2024 in case AT.40577 – Vifor, paragraphs 200-202; and Commission decision of 31 October 2024 in case AT.40588 – Teva Copaxone, paragraphs 1480 and 2142.

(434)  More specifically, an objective necessity may relate to compliance with a public ban on selling dangerous substances to certain customers for reasons of safety or health. See, by analogy, Commission Guidelines on vertical restraints (‘Vertical Guidelines’) (OJ C 248, 30.6.2022, p. 1), paragraph 180. In AstraZeneca, the undertaking put forward the compliance with pharmacovigilance obligations as an objective justification. The General Court dismissed this argument, relying on several grounds, and held that the mere burden of having to comply with certain obligations may not be enough to justify an abuse of dominant position. See judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraphs 685-694. The General Court ruled in particular that ‘the obligation to submit, at five-yearly intervals, reports on other suspected adverse reactions does not constitute such a significant pharmacovigilance burden that it could constitute a serious objective ground of justification’. Judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraph 692.

(435)   Commission decision of 7 December 2018 in case AT.40461 – DE/DK Interconnector, paragraph 67.

(436)  The European Union’s resilience may be enhanced where the conduct is necessary to either (a) reduce dependencies and mitigate shortages and disruptions in supply chains, or (b) ensure sufficient production capacity for the EU’s security and defence industry. In so far as such enhanced resilience increases the quality of the dominant undertaking’s product (for example, in terms of lead times or reliability of supply) to the benefit of consumers or enables the dominant undertaking more generally to compete more effectively on parameters such as price, choice, quality and innovation, it will be assessed as an efficiency defence pursuant to the principles set out in Section 5.3.

(437)  Judgment of 12 December 1991, Hilti v Commission, T-30/89, EU:T:1991:70, paragraph 118; and judgment of 6 October 1994, Tetra Pak v Commission, T-83/91, EU:T:1994:246, paragraph 138.

(438)  See, to that effect, judgment of 12 December 1991, Hilti v Commission, T-30/89, EU:T:1991:70, paragraphs 116-118.

(439)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 103. See also, to that effect, judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 883; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 363-371.

(440)  Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraph 103. See also, to that effect, judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 883; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 363-371.

(441)  See, to that effect, judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 203.

(442)  Judgment of 15 March 2007, British Airways v Commission, C-95/04 P, EU:C:2007:166, paragraph 86; judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651, paragraph 42; and judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632, paragraph 140.

(443)  See, by analogy, Commission Guidelines on the application of Article 101(3) (formerly 81(3)) of the Treaty (‘Article 101(3) Guidelines’) (OJ C 101, 27.4.2004, p. 9), paragraphs 66 and 67.

(444)  Such efficiencies may include savings in logistical costs due to economies of scale in transport and distribution that may be generated by, for instance, exclusive dealing. See, by analogy, Vertical Guidelines, paragraph 16(g).

(445)  Article 101(3) Guidelines, paragraph 59.

(446)  Assessed in Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraphs 993-999; and Commission decision of 20 March 2019 in case AT.40411 – Google Search (AdSense), paragraphs 447-449 (annulled by the General Court by judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, but not on this point).

(447)  Relationship-specific investments are investments that suppliers or buyers make in their specific vertical relationship, for example by investing in specific equipment or training to satisfy the requirements of the other party. These investments are typically sunk and have little or no value outside the specific vertical relationship. See Vertical Guidelines, paragraph 16(e). Relationship-specific investments were assessed in the Commission decision of 24 January 2018 in case AT.40220 – Qualcomm (Exclusivity payments), paragraphs 507-523 (annulled by the General Court by judgment of 15 June 2022, Qualcomm (Exclusivity payments), T-235/18, EU:T:2022:35, but not on this point); and Commission decision of 20 March 2019 in case AT.40411 – Google Search (AdSense), paragraphs 443-445 (annulled by the General Court in judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, but not on this point).

(448)  See, by analogy, Vertical Guidelines, paragraph 16(b), (c), (e), and (i).

(449)  Under these Guidelines, the term sustainability includes, but is not limited to, addressing climate change (for instance, through the reduction of greenhouse gas emissions), reducing pollution, limiting the use of natural resources, upholding human rights, ensuring a living income, fostering innovation or resilient infrastructure, reducing food waste, facilitating a shift to healthy and nutritious food, ensuring animal welfare. See, to that effect, Commission Guidelines on the applicability of Article 101 of the Treaty on the Functioning of the European Union to horizontal co-operation agreements (‘Horizontal Co-operation Agreements Guidelines’) (OJ C 259, 21.7.2023, p. 97), paragraph 517.

(450)  See Horizontal Co-operation Agreements Guidelines, paragraph 558.

(451)  See Vertical Guidelines, paragraph 316.

(452)  Judgment of 10 November 2021, Google and Alphabet v Commission (Google Shopping), T-612/17, EU:T:2021:763, paragraph 188. See also, to that effect, judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379, paragraphs 84-86.

(453)  Judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 592. These conditions are to be interpreted consistently with those laid down in Article 101(3) TFEU. See, to that effect, judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 201, 204, 205, 207 and 208.

(454)  For the purpose of this condition, the term ‘consumers’ refers to all groups of consumers affected by the conduct of the dominant undertaking and not individual consumers. See, to that effect, judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 193-195, 234-237; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 592 and 593. Accordingly, the effects to be assessed are those on all consumers in the relevant markets, not on each consumer individually. See, to that effect, judgment of 23 November 2006, Asnef-Equifax, C-238/05, EU:C:2006:734, paragraph 70; and judgment of 21 March 2002, Shaw and Falla v Commission, T-131/99, EU:T:2002:83, paragraph 163.

(455)  An efficiency defence cannot be established if the practice bears no relation to the alleged benefits for consumers, or if it goes beyond what is necessary to achieve these benefits. See judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 76.

(456)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 204; judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 592; See also, to that effect, judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraph 42; judgment of 30 January 2020, Generics (UK) and Others, C-307/18, EU:C:2020:52, paragraph 166; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraph 362.

(457)  See, by analogy, Article 101(3) Guidelines, paragraph 46.

(458)  See, to that effect judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 204; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 592.

(459)  See, by analogy, judgment of 21 November 2009, GlaxoSmithKline Services and Others v Commission and Others, C-501/06 P, C-513/06 P, C-515/06 P and C-519/06 P, EU:C:2009:610, paragraphs 93-95; judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 190.

(460)  See judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 237.

(461)  See Commission decision of 24 January 2018 in case AT.40220 – Qualcomm (Exclusivity payments), paragraph 519 (annulled by the General Court in judgment of 15 June 2022, Qualcomm (Exclusivity payments), T-235/18, EU:T:2022:35, but not on this point).

(462)  See, by analogy, judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 196.

(463)  See, by analogy, Article 101(3) Guidelines, paragraph 56. See also, by analogy, judgment of 21 November 2009, GlaxoSmithKline Services and Others v Commission and Others, C-501/06 P, C-513/06 P, C-515/06 P and C-519/06 P, EU:C:2009:610, paragraphs 93-95; judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 190.

(464)  See, by analogy, Article 101(3) Guidelines, paragraph 57.

(465)  Judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172, paragraph 42.

(466)  Judgment of 15 March 2007, British Airways v Commission, C-95/04 P, EU:C:2007:166, paragraph 86; judgment of 17 February 2011, TeliaSonera Sverige, C-52/09, EU:C:2011:83, paragraph 76. See, by analogy, Article 101(3) Guidelines, paragraph 54.

(467)  See, by analogy, Article 101(3) Guidelines, paragraphs 53 and 54.

(468)  See, to that effect, by analogy, judgment of 13 July 2023, CK Telecoms, v Commission, C-376-20 P, EU:C:2023:561, paragraph 82; and judgment of 17 May 2023, EVH v Commission, T-312/20, EU:T:2023:252, paragraphs 233 and 234.

(469)  See Article 101(3) Guidelines, paragraph 88.

(470)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 193-195; judgment of 21 December 2023, Royal Antwerp Football Club, C-680/21, EU:C:2023:1010, paragraphs 123 and 130; and judgment of 23 November 2006, Asnef-Equifax, C-238/05, EU:C:2006:734, paragraph 70.

(471)  See judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 193-196 and 233-237; judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraphs 592 and 593; and judgment of 23 November 2006, Asnef-Equifax, C-238/05, EU:C:2006:734, paragraph 70.

(472)  See, by analogy, Article 101(3) Guidelines, paragraph 98.

(473)  The Commission understands the benefits consumers can derive from qualitative efficiencies as encompassing individual use-value benefits (i.e. benefits consumers derive from the consumption of a specific product), non-use value benefits (i.e. benefits consumers derive from their concern for the impact of the consumption of a specific product on others) and collective benefits (i.e. benefits accruing to a wider section of society from addressing negative or positive externalities). See, by analogy, Horizontal Co-operation Agreements Guidelines, paragraphs 571-584.

(474)  See, by analogy, Horizontal Co-operation Agreements Guidelines, paragraphs 577-579.

(475)  See, by analogy, Article 101(3) Guidelines, paragraph 104.

(476)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 204; and judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 592.

(477)  Judgment of 27 September 2006, GlaxoSmithKline Services and Others v Commission, T-168/01, EU:T:2006:265, paragraphs 248 and 251; and judgment of 11 September 2014, MasterCard Inc, C-382/12 P, EU:C:2014:2201, paragraph 242.

(478)  See, by analogy, Horizontal Co-operation Agreements Guidelines, paragraphs 582-584.

(479)  See, by analogy, Horizontal Co-operation Agreements Guidelines, paragraph 587.

(480)  Judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 204; see also Commission decision of 20 March 2019 in case AT.40411 – Google Search (AdSense), paragraphs 452, 455, 621 (annulled by the General Court in judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search), T-334/19, EU:T:2024:634, but not on this point); Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraphs 1003 and 1329; and Commission decision of 24 January 2018 in case AT.40220 – Qualcomm (Exclusivity payments), paragraphs 516 and 523 (annulled by the General Court in judgment of 15 June 2022, Qualcomm (Exclusivity payments), T-235/18, EU:T:2022:35, but not on this point).

(481)  See, to that effect, judgment of 15 March 2007, British Airways v Commission, C-95/04 P, EU:C:2007:166, paragraph 86; judgment of 18 September 2024, Qualcomm (Predation), T-671/19, EU:T:2024:626, paragraph 593; and judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraphs 602, 607-614, 891; and judgment of 2 July 2026, Google and Alphabet v Commission (Google Android), C-738/22 P, EU:C:2026:533, paragraphs 363-371.

(482)  See, by analogy, Article 101(3) Guidelines, paragraphs 74 and 75.

(483)  See, by analogy, Article 101(3) Guidelines, paragraph 80.

(484)  Judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 612.

(485)  See, to that effect, Commission decision of 4 March 2024 in case AT.40437 – Apple – App Store Practices (music streaming), paragraphs 806 and 807.

(486)  See, for instance, Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraphs 995-997; judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraphs 607-611. See also, to that effect, Commission decision of 4 March 2024 in case AT.40437 – Apple – App Store Practices (music streaming), paragraphs 803, 806-808, 812 and 823.

(487)  Commission decision of 24 January 2018 in case AT.40220 – Qualcomm (Exclusivity), paragraphs 521 and 525 (annulled by the General Court in judgment of 15 June 2022, Qualcomm v Commission, T-235/18, EU:T:2022:35, but not on this point).

(488)  See, for instance, Commission decision of 18 July 2018 in case AT.40099 – Google Android, paragraph 998; judgment of 14 September 2022, Google and Alphabet v Commission (Google Android), T-604/18, EU:T:2022:541, paragraph 613.

(489)  See Commission decision of 24 January 2018 in case AT.40220 – Qualcomm (Exclusivity), paragraph 523 (annulled by the General Court in judgment of 15 June 2022, Qualcomm v Commission, T-235/18, EU:T:2022:35, but not on this point).

(490)  See judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraphs 204 and 207.

(491)  See, by analogy, Article 101(3) Guidelines, paragraph 105.

(492)  See, to that effect, judgment of 21 December 2023, European Superleague Company, C-333/21, EU:C:2023:1011, paragraph 207.

(493)  For a more detailed discussion of the relevant factors to be assessed see Article 101(3) Guidelines, paragraphs 108-116.

(494)   OJ C 45, 24.2.2009, p. 7.