CYIL vol. 11 (2020)

MICHAL PETR CYIL 11 (2020) prices, in 10% of cases using automated pricing algorithms, i.e. without a “human” decision directly involved. 16 In order to appreciate the magnitude of such changes, please consider that Amazon performed already in 2012, 2.5 million price changes in a single day; by contrast, the “brick and mortar” Wall-Mart performed only 50 000 price changes in a month, 17 i.e. on average 1 700 changes a day, which is approximately 1 500 times less than Amazon. 2. Personalised Pricing Algorithms The crucial distinction of personalised pricing is that the algorithm sets a specific price for an individual customer or a group of customers, based on their willingness to pay. As the OECD summarizes it, personalised pricing stands for any practice of price discriminating final consumers based on their personal characteristics and conduct, resulting in prices being set as an increasing function of consumers’ willingness to pay. 18 This amounts to price discrimination. Economic theory distinguishes between three forms of price discrimination: 19 first-degree (or perfect) discrimination, where each consumer is charged their full willingness to pay, 20 second-degree discrimination (versioning), where the seller offers a number of versions of the same product at different prices, leaving the consumers the decision of choosing a version according to their preferences, and third-degree discrimination, a practice of setting different prices to different groups of consumers, which are partitioned according to their observed characteristics. We will not discuss the second- degree discrimination (versioning) further, as the use of pricing algorithms does not influence it significantly. 21 Pricing algorithms determine the “individual” price using three sets of data: volunteered (e. g. phone number, date of birth etc.), observed (search history, users’ location etc.), and inferred (e. g. health status, hobbies etc.) ones. 22 Digitalization has enormously increased the amount of observable data and employment of advanced algorithms and will increase the possibility to work with these data and increase the quality of the inferred ones. It may 16 Commission Staff Working Document accompanying the document Report from the Commission to the Council and the European Parliament. Final Report on the E-commerce Sector Inquiry. 10 May 2017, SWD(2017) 154 final ( E-commerce Report ), available at: https://eur-lex.europa.eu/resource.html?uri=cellar:9d1137d3-3570-11e7- a08e-01aa75ed71a1.0001.02/DOC_1&format=PDF (1 July 2020), p. 175. 17 MEHRA, S. K. Antitrust and the Robo-Seller: Competition in the Time of Algorithms. Minnesota Law Review, 2016 (100), p. 1335. Available at: http://www.minnesotalawreview.org/wp-content/uploads/2016/04/Mehra_ ONLINEPDF1.pdf (1 June 2020). 18 OECD Report on Personalised Pricing , p. 9. 19 PIGOU, A. C. The Economics of Welfare . Palgrave Macmillan (2013, originally in 1920). 20 This is believed to be – so far – only a hypothetical scenario. See e. g. European Commission. Consumer market study on online market segmentation through personalised pricing/offers in the European Union (2018), available at: https://ec.europa.eu/info/sites/info/files/aid_development_cooperation_fundamental_rights/aid_and_ development_by_topic/documents/synthesis_report_online_personalisation_study_final_0.pdf (1 July 2020). 21 Accordingly, the UK Study defines personalised pricing as: „ the practice where businesses may use information that is observed, volunteered, inferred, or collected about individuals’ conduct or characteristics, to set different prices to different consumers (whether on an individual or group basis), based on what the business thinks they are willing to pay “. UK Study , p. 36. 22 OECD Report in Personalised Pricing , p. 11.

242