Large platforms acting as “digital gatekeepers” are increasingly drawing competition agencies’ attention. While no legal definition of a “gatekeeper” has been laid down yet, this concept is meant to cover platforms that rely on significant network effects. A “gatekeeper” may also be a go-between, controlling access from one point to another. However, gatekeepers may refuse or control access to services, as shown by the Commission’s latest probe into Amazon’s Marketplace.
On 10 November, the Commission addressed a statement of objections to Amazon, informing it of its preliminary position that it allegedly breached EU competition rules in online retail markets. In the Commission’s view, Amazon has systematically breached antitrust rules as it has been relying on non-public data collected from independent sellers using Amazon’s Marketplace as an intermediary. The Commission press release suggests that Amazon abused its access to non-public data of retailers to gain a competitive advantage for its retail sales business.
Misuse of Marketplace retailers’ data
Just like many other digital platforms, Amazon plays the role of an intermediary by providing a marketplace for independent retailers. It is also active as a retailer in the same market.
This position gives Amazon access to rather sensitive business data, including:
- the number of products ordered and shipped;
- the turnover generated by retailers on Amazon’s Marketplace;
- shipping data;
- data related to retailers’ past performance; and
- various claims of consumers on products, including guarantees.
The Commission concluded that employees of Amazon’s retail business units had access to a very large amount of non-publicly available retailer data that was then transmitted to the automated systems of Amazon’s retail business. The data was subsequently used by Amazon to adjust retail offers, serving as a basis for adopting Amazon’s own retail business strategy. The Commission found that, as a result, Amazon was able to pick up the best-selling products in various categories.
Theory of harm
The Commission considered this practice to be solely the result of Amazon’s access to sensitive data rather than the result of normal market conditions. The access and the subsequent collection of non-public retailers’ data allowed Amazon to avoid usual competitive risks and to leverage its dominance in the market for the provision of marketplace services in France and Germany to strengthen its position on the neighbouring direct retail market.
A more careful read of the Commission’s press release could lead to the conclusion that such practices are anticompetitive “per se” as this may be the way to circumvent or “avoid the normal risks of retail competition”.
This statement implies that the use of data of retailers is automatically anticompetitive without providing any detailed reasoning for this finding. It seems that Amazon makes use of this strategy to put competitive pressure on already established players in the retail sales market by replicating what others do well and by getting an insight into the products that are the most relevant for end costumers. In addition, this business conduct does not involve any infringements of intellectual property rights, e.g. by counterfeiting established trademarks. Paradoxically, this rather “improper” method is considered acceptable in oligopoly-like markets characterized by higher transparency so that competitors are able to monitor each other’s conduct and adjust their respective commercial strategies. While in Airtours, the Court of First Instance found that in oligopolistic markets, each member of the oligopoly knows how the other members are behaving and is thus able to adopt the same policy. However, while in oligopolistic markets this is the result of a transparent market structure, in the case of Amazon, adjusting one’s own commercial strategy based on competitors’ data is a hardcore anticompetitive practice. Thus, it seems that adopting competitors’ commercial strategies in a market where very few operators are present is not contrary to competition rules, while replicating customers’ strategy is contrary to Art. 102 TFEU. In addition, undertakings are free to make use of any competitive assets they dispose of to be more successful. In the absence of any IP rights or data protection infringements, it is unclear why the Commission considers the use of non-public data as a “per object” restriction instead of opting for an effects-based test.
That said, the Commission’s approach opens a number of questions rather than providing for answers: Is the scale of Amazon’s conduct problematic? Are big platforms punished for gathering and using third-party data in a more efficient way? Is there a double standard as to the conduct of smaller players as opposed to gatekeepers’ strategies? When did exploiting own competitive advantages become anti-competitive?
Just like in the Google Shopping case, it seems that the mere exploitation by a company of its own competitive advantages is considered to be of an anti-competitive nature.
This reasoning is difficult to follow, as the very purpose of competition rules is not to set undertakings at an equal level and to impose on them to compete with the same assets and means. Competition rules are rather set to ensure that all players – big as well as small ones – have equal chances and incentives to succeed. In that sense, the point of competition rules is to allow market players to make or be able to make the best of the assets they have (IP rights, business know-how, customer networks, commercial strategies etc.).
That said, it remains unclear when an undertaking may exploit its own competitive advantages. The Commission should clarify what type of competitive advantages can be exploited and whether there are some limits as to the size of undertakings exploiting their own advantages. While the exploitation of certain competitive advantages by smaller players could be considered as logical or even pro-competitive, comparable conduct by large players with a “gatekeeping” role may appear to constitute an exploitative abuse.
Anticompetitive object or anticompetitive effects?
The Commission’s press release suggests that the use of non-public data would facilitate Amazon in leveraging “its dominance in the market for the provision of marketplace services”. The press release does not specify the markets affected by Amazon’s alleged conduct. There are also so far no references to potential pro-competitive effects resulting from Amazon’s adjusted commercial strategies.
Further questions arise as to eventual remedies. While splitting Amazon’s retail business and its marketplace business appear unlikely, behavioural remedies, especially the cease of data collection and data misuse, could be expected. However, there may be a number of issues regarding the actual implementation of such remedies or the appointment of monitoring trustees. Indeed, it seems probable for Amazon to try to circumvent such remedies. In order for potential distortions of competition to be avoided, such remedies should be implemented for at least five to ten years. It remains unclear if such remedies would have any long-term effects.
The investigation into Amazon’s practices regarding its “Buy Box” and Prime label
On the same date that it announced having sent a statement of objections to Amazon, the Commission opened a formal probe into Amazon’s Buy Box and Prime label units. Once again, the Commission’s attention has been drawn to alleged self-preferencing practices. In the Commission’s preliminary view, Amazon artificially favors its own retail offers as well as offers of retailers using its marketplace and delivery services.
The selection for the “Buy Box” display makes the respective product and retailer more visible for end customers and grants them a competitive advantage compared to other competing products and operators. In that sense, the “Buy Box” option functions like a “must-have” infrastructure for retailers. Visibility among customers seems to be crucial for a breakthrough in the online retail market. Indeed, in the Google Shopping case, the Commission deemed abusive the systematic granting by Google of prominent placement to its own comparison shopping service. The Commission also found that Google had demoted rival comparison shopping services based on its own general search algorithms. The Commission probe demonstrated that the most highly ranked rival services appeared on average on page four of Google’s search results while others were even further down.
That said, the recent Commission investigation shows a clear trend towards incriminating self-preferencing by leading platforms as abusive conduct under Art. 102 TFEU. While waiting for the adoption of the Digital Services Act, it remains to be clarified whether such conduct should only apply to digital gatekeepers and whether any form of safe harbor may be envisaged.
 See on this the Commission’s recent proposal for a Digital Markets Act, which defines gatekeepers as “core platform providers” that “feature an ability to connect many business users with many end users through their services which, in turn, allows them to leverage their advantages, such as their access to large amounts of data, from one area of their activity to new ones. Some of these providers exercise control over whole platform ecosystems in the digital economy and are structurally extremely difficult to challenge or contest by existing or new market operators, irrespective of how innovative and efficient these may be. Contestability is particularly reduced due to the existence of very high barriers to entry or exit, including high investment costs, which cannot, or not easily, be recuperated in case of exit, and absence of (or reduced access to) some key inputs in the digital economy, such as data .”, see European Commission Communication of 15 December 2020, COM(2020) 842 final, “Proposal for a Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector” (“Digital Markets Act”), Rec. 3.
 Commission Press Release of 10 November 2010, “Commission sends Statement of Objections to Amazon for the use of non-public independent seller data and opens second investigation into its e-commerce business practices”, at: https://ec.europa.eu/commission/presscorner/detail/en/ip_20_2077 [last retrieved on 29 November 2020].
 Judgment of the Court of First Instance of 6 June 2002 in Case T-342/99, Airtours plc v Commission, ECLI:EU:T:2002:146, para. 62.
 Commission Decision of 27 June 2017 in Case AT.39740, Google Search (Shopping).
 European Commission, “The Digital Services Act package”, at: https://ec.europa.eu/digital-single-market/en/digital-services-act-package [last retrieved on 4 December 2020].