In September 2018, the European Commission (“EC”) sent out formal requests for information (“RFIs”) to investigate allegations of an anticompetitive conduct by Amazon. The investigation relates to the interdependencies between Amazon’s third-party sales platform for retailers (“Amazon Marketplace”) and Amazon’s own online retail operations. Operating both on an upstream intermediation market for businesses (“merchants”) and downstream retail markets vis-à-vis its end customers (“shoppers”) has created a strong conflict of interest for Amazon. This article outlines the background to the EC’s investigation and the current focus and likely theories of harm regarding Amazon’s conduct involved in the investigation
Amazon Marketplace offers merchants a wide range of functionalities. They can use the Amazon Marketplace as a new or additional sales channel, building on Amazon’s brand. They can also purchase additional Amazon services such as warehousing their products in Amazon’s fulfillment centers, where Amazon handles the packing and shipment of the goods and provides customer service for the merchants. In addition, Amazon collects and transfers shopper payments to the merchants.
At the same time, Amazon is one of the largest online retailers itself. Depending on the product, Amazon’s own retail offerings may directly compete with those of the merchants using the Amazon Marketplace. Due to the comparable platform and similar shopping experience, many shoppers may not recognize any difference between Amazon’s own retail services and its marketplace activities for other merchants.
Over the years, Amazon’s dual role as both marketplace sales representative and online retailer (“hybrid platform”) has raised concerns both in the US and in Europe. Most allegations concern the manner in which Amazon collects and analyzes retailer data to learn which products sell well. Merchants have claimed that Amazon is using the data it collects to identify successful new product offerings on its platform to then market an Amazon-own version. In other words, Amazon is alleged to use the data it collects from the retailers using its Marketplace to compete against them with an own offering. To market its own version, Amazon is said to either approach the actual manufacturers of the original (successful) products with a view to reselling the products at a lower price or to even sell those products as so called “Amazon Basics”– its own brand. Amazon reportedly offers its own products at a price below that of the original merchant so that Amazon’s offering appears as the lowest priced offer on its Marketplace platform – and thus as a first choice in Amazon’s Buy Box.
Following the EC’s findings in its e-commerce sector enquiry in 2017, the EC recently commenced investigating allegations that Amazon uses data from third-party transactions generated via its Marketplace to enhance its own online retail offerings.
On 19th September 2018, EU Commissioner for Competition, Margrethe Vestager, explained that the investigation was at an early stage and primarily dealt with the issue of Amazon’s use of data. In particular, she raised the question whether:
“[I]f you as Amazon get the data from the smaller merchants that you host – which can be, of course, completely legitimate because you can improve your service to these smaller merchants – do you then also use this data to do your own calculations: as what is the new big thing, what is it that people want, what kind of offers do people like to receive, what makes them buy things?”.
To shed more light on the matter, the EC sent formal RFIs to many merchants active on the Amazon Marketplace. With a deadline to answer until the 9th October 2018, the questions aimed at discovering any potential Amazon strategy to use merchants’ data sets generated through their sales on the marketplace and to determine the respective value of that data for Amazon.
To this end, the EC asked whether Amazon ever started selling products that merchants had also separately sold within the previous five years. Merchants answering in the affirmative were requested to provide further details on Amazon’s approach (e.g. whether it temporarily suspended the merchants’ sales operations) and on the impact of that conduct on the merchant’s business. The EC further inquired if Amazon ever approached the manufacturers of products sold by merchants on the marketplace.
A second string of questions related to the relevant types of data merchants would make available to Amazon, including (average) prices, quantities sold, specific conversion rates, rebate campaigns or quantities stocked. The RFI also seeks information on the respective producers and factors that are relevant for the positioning and displaying in “Amazon Buy Boxes”. Finally, merchants are asked whether they use Amazon software such as “Amazon Seller Central” or Amazon (or third-party) price-monitoring software.
Following its own sector enquiry into online price comparison services in October 2017, in June 2018 the German Federal Cartel Office (“Bundeskartellamt”) revealed to have “received a lot of complaints” and to be “looking at the role and market power of Amazon” with regards to Amazon’s hybrid function. Germany is Amazon’s second largest market after the US.
In a policy paper published in October 2018, the Bundeskartellamt summarized its concerns as follows:
“Not only did this dual business model allow the platforms to develop a strong market position based on increased network effects resulting from the large variety they offer. For instance, they provide a rapidly available wide and deep range of products, which can result in dealers’ businesses being dependent on the platforms. There is also the risk that, when cooperating with the manufacturers, independent dealers can be disadvantaged or even squeezed out of the market due to unfavourable conditions. As a competition authority, the Bundeskartellamt wants to keep markets open and prevent e-commerce from being concentrated in the hands of only a few players, i.e. the manufacturers themselves, some large dealers and even fewer leading platforms, which would dramatically reduce customers’ choice options.”
With the EC’s initiation of proceedings for the adoption of a prohibition decision against Amazon, the Bundeskartellamt will lose its competence to further investigate the same conduct (Article 11(6) 1/2003). It is unknown whether it has yet referred the case to the EC. In any case, the Bundeskartellamt’s preliminary findings and concerns are likely to also play a role in the EC’s investigation.
Amazon’s conduct as described in the EC’s RFI raises several competition concerns. If the EC finds its suspicions confirmed, this could amount to a violation of Articles 101 and/or 102 of the Treaty on the Functioning of the European Union (“TFEU”). In banning anticompetitive agreements, Article 101 TFEU covers, inter alia, the exchange of competitively sensitive information between rivals. Article 102 TFEU prohibits the abuse of a dominant market position. In case of an infringement, the EC could impose a fine of up to 10% of Amazon’s total turnover in the preceding business year.
Unlike other RFIs, the EC has left open whether it is pursuing the case under Article 101 and/or 102 TFEU, and what the relevant theory of harm would be. The EC has indicated that Amazon’s dual position as merchant platform and online retailer is at the core of the concerns. The dual position allows Amazon to adjust its own offerings on the retail level to the success or failure of other companies selling identical or comparable goods via the Amazon Marketplace. The risk of a commercial failure regarding the recoupment of an investment is close to zero once Amazon has learned from its Marketplace data that a certain product sells well. By better positioning and/or pricing its own follow-on product, Amazon may possibly push all consumer demand (originally created by the pioneer merchant on the Amazon Marketplace) to its own “copy-cat” product.
It is also possible that the EC assesses Amazon’s collection and use of the market data gathered through its Marketplace as an “exchange of information” amongst competing merchants, triggering concerns under Article 101 TFEU. For that, Amazon’s practice would have to amount to an “agreement between undertakings” or a “concerted practice” with the object or effect of restricting competition between Amazon and the merchants in retail sales.
In cases of an exchange of information, the EC views the making available of up-to-date, non-aggregated data as particularly problematic if such data is competitively sensitive. Relevant to competition are company data a competitor can take into account for its own strategic market behavior, in particular sales figures, market shares, prices/discounts/bonuses, customer names and costs.
With regards to Amazon, two different competitive relationships must be distinguished. In its capacity as a merchant platform, Amazon engages in a vertical relationship with merchants. In this vertical relationship, the market data can be used to improve the product search and matching functionalities of the Amazon Marketplace for merchants and shoppers alike. Any product-specific data allows Amazon to constantly enhance its search and ranking algorithms. The need to improve and innovate the service could thus justify the gathering of such merchants’ data that is required for the matching of merchants with shoppers. Both sides of the market profit from more relevant product search results. In contrast, there is no apparent legitimate need to gather merchant data that is unrelated to the matching functionality of the marketplace but rather linked to Amazon’s own activities as an online retailer.
The EC will likely also investigate whether Amazon’s conduct amounts to an abuse of dominance on the (likely national) markets for online merchant platforms. Whether Amazon can be found to be dominant on these markets will depend on the extent to which other market players such as eBay Marketplaces, AliExpress, Fnac or smaller European and international companies actually compete with and restrain the Amazon Marketplace. It may well be that Amazon is dominant individually in some member states of the European Union while it forms a duopoly with eBay in other states – or lacks any dominance in yet other countries due to strong domestic competition.
In countries in which Amazon is found to be dominant, the crucial question will be whether its conduct can be considered an abuse of dominance. To that end, the EC is likely to first assess well-established theories of harm such as predatory pricing or exploitation. However, in lack of an exhaustive list of abusive practices, the EC may also consider the conduct to be a novel type of an abuse that is linked to the particularities of the markets in question.
One well-established form of abuse that has been suggested in connection with Amazon’s strategy to undercut the prices of merchants on its Marketplace is predatory pricing. While it appears that Amazon has offered its own “copy-cat” products at prices below those of other merchants, this does not mean that Amazon is selling below cost. The competitive concern is not that Amazon offered its products at a lower price than the original seller (which may actually benefit shoppers). The concern originates from the fact that Amazon is only able to offer such prices due to its use of third-party data. The data excludes any risks regarding the launch of a new product and also makes it easier to adjust the scope of production to the actual demand. This is not about predatory pricing but more about its ability to benefit from economies of scale and scope because of unfairly obtained information advantage.
It could be argued that Amazon is abusing its dominance by exploiting merchants whose market data it collects without approval or against their will. In this respect the investigation has similarities with the German investigation into Facebook. In a preliminary assessment, the German Bundeskartellamt had found that Facebook abuses its dominance on the market for social networks by making the use of its social network conditional on the collection of unlimited amounts of user data from other sources and then linking it to the user’s account for marketing purposes.
Amazon’s impermissible use of merchant data appears even more problematic from a competition law standpoint than Facebook’s practice because Amazon uses the extracted data directly against the commercial interests of its Marketplace merchants (which are also competitors), namely to outcompete them at the online retail level. In a competitive environment, one would expect that customers leave a platform as soon as they realize that the platform is using their data to place competing products.
One question may be whether small and medium-sized merchants have become so dependent on using the Amazon Marketplace to reach shoppers that they now cannot easily withdraw from the platform. Some recent studies would suggest that they have. In Germany, for example, almost one third of all merchants have claimed that they depend on online merchant platforms such as the Amazon Marketplace, and more than half of them were concerned about becoming too dependent.
There is also a link between Amazon’s conduct and the EC’s abuse of dominance investigations into Google’s unauthorized use of third-party content such as news, images or reviews for its specialized search services (Google News, Images, Local).
In a preliminary assessment, the EC had identified Google’s use of content that it had “scraped” from third-party websites without consent in its specialized search services as a potential abuse of dominance. Websites did not have the technical option to block Google’s use of their content in Google’s specialized search services (with which they competed) without this affecting their visibility also in Google’s general search service (with which they did not compete). The EC concluded that Google had created an anticompetitive “link between getting the right to use material from other sites on its specialised search services and the appearance that these sites have on Google’s general search results – a practice that allows Google to benefit from investments made by other firms.”  Accordingly, the EC had “asked Google to sever this link to restore competitive incentives.”
Amazon has created a similar link. Severing this link (e.g. by means of a functional separation of the different Amazon units) would allow merchants to freely use the Amazon Marketplace without concern that it empowers Amazon to launch a competing product on the online retail market.
Amazon’s conduct also bears strong resemblance to Google’s favoring of its own services as prohibited by the EC in Google Search (Shopping). Both cases deal with a severe conflict of interest resulting from a company’s dominance on an upstream intermediation market for downstream services which the company offers itself. Google was found to be active both on the upstream intermediation market for general search services (which it dominates) and the downstream markets for specialized search services such as comparison shopping services (which it sought to dominate). The EC found that Google abused its dominance in the upstream market by favoring its downstream comparison shopping service on general search results as compared to competing services.
Similarly, Amazon is active both on the upstream intermediation market for merchant platforms (which it may dominate) and on the downstream online retail market (which it may be seeking to dominate).
Ultimately, the overarching theory of harm relates to leveraging. Since 1974, the European Courts have consistently held that an undertaking with a dominant position in a given market may not extend that position to a separate but related market by means that do not reflect competition on the merits. If confirmed by the responses to the EC’s RFIs, Amazon may be found to have used its dominance in the market for merchant platforms to extend that position into online retail markets by collecting and using its merchant customers’ market data. Amazon was only able to acquire that data due to the dependency of those merchants on its marketplace. No competing retailer is in that position.
Raising various issues of market definition, dominance and abuse in several multi-sided online markets, the EC’s Amazon investigation is both complex and highly relevant for the e-commerce sector. While having to address several novel issues, the EC will likely rely to a large extent on its findings in the Google investigation and the e-commerce sector inquiry. It may draw further conclusions from the German Amazon and Facebook investigations. In any case, the EC appears to be fully aware of the relevance of data being amassed in the hands of a few. The Amazon case may determine whether the established principles of Article 101 and 102 TFEU suffice to address this issue.
 See e.g. Greg Bensinger, The Wall Street Journal, 27 June 2012, Competing With Amazon on Amazon.
 See e.g. Jennifer Rankin, The Guardian, 23 June 2015, Third-party sellers and Amazon - a double-edged sword in e-commerce.
 The “Amazon Buy Box” is a box on the product’s details page, where customers begin the purchasing process. Often, the same product is offered by multiple sellers (i.e. several Merchants, including in some instances Amazon itself). Thus, these sellers compete to be in the Buy Box. Amazon may select one seller among the eligible sellers based on algorithms and then grant that seller the Buy Box.
 Case AT.40462 – Amazon Marketplace. Bundeskartellamt, 27 October 2017, Bundeskartellamt launches sector inquiry into comparison websites https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2017/24_10_2017_Vergleichsportale.html?nn=3591568
 Nicholas Hirst, MLEX, 27 June 2018, Amazon’s ‘hybrid function’ catches eye of German antitrust enforcers.
 Bundeskartellamt, Competition restraints in online sales after Coty and Asics - what’s next?, Series of papers on Competition and Consumer Protection in the Digital Economy, October 2018, page 6.
 Article 23 of 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 (“Regulation 1/2003”).
 EC, Guidelines on the applicability of Article 101 of the Treaty on the Functioning of the European Union to horizontal co-operation agreements of 14 January 2011, at para 86.
 Lina M. Khan, Amazon’s Antitrust Paradox, The Yale Law Journal 2017, 710 et seqq.
 See the Bundeskartellamt, 19 December 2017, Preliminary assessment in Facebook proceeding: Facebook's collection and use of data from third-party sources is abusive, available at https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2017/19_12_2017_Facebook.html
 See Bundeskartellamt, id.
 DHL, 25 October 2018, Onlinehändler im Spannungsfeld von Wachstum und Marktkonzentration.
 See EC, 25 April 2013, MEMO/13/383, Commission seeks feedback on commitments offered by Google to address competition concerns – questions and answers, available at http://europa.eu/rapid/press-release_MEMO-13-383_en.htm.
 See in greater detail Hoppner/Davies The EU Competition Investigation of Internet Search. Why Google's Market Tested Proposals for Re-Labelling and Opt-Out Solutions Do Not Suffice, Comp. Law Review Int. (CRi) 2013, p. 107, 110 et sub, available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2597543
 Vice-President Almunia, 9 December 2013, SPEECH/13/1042, Intellectual property and competition policy, available at http://europa.eu/rapid/press-release_SPEECH-13-1042_en.htm.
 See id..
 Case AT.39740 Google Search (Shopping), on appeal: Case T-612/17 – Google and Alphabet v Commission.
 Joined cases 6 and 7-73, Commercial Solvents, ECR 1974, 223.
 See Case 311/84 Télémarketing (CBEM), EU:C:1985:394, para. 27; Case T-201/04, Microsoft, EU:T:2007:289, para. 1344; see in greater detail Hoppner, Monopoly Leveraging & Equal Treatment: the EU Commission’s Google Shopping Decision, available at https://www.hausfeld.com/news-press/monopoly-leveraging-equal-treatment-the-eu-commissions-google-shopping.
*Prof. Dr. Thomas Höppner is a partner and Philipp Westerhoff is an associate in the Berlin office.