15 November 2019

Exclusive Data, Price Manipulation, and Market Leadership

Whether the collection of data on customer behavior by companies like Amazon, Tesco, or Starbucks can entail pro- or anti-competitive effects is largely debated. The authors find that both directions are possible, depending on the size of a company’s exclusive customer list. Mandatory data sharing can be an effective tool to restore market competition, but it should be considered on a case-by-case basis.

Key issue

Market leadership and exclusive access to data are observed in several industries. Several firms, such as tech giants, as well as bricks-and-mortar companies (like the UK’s Tesco supermarkets or Starbucks cafés), have collected, stored, and analyzed individual data for years. In the same vein, Amazon’s competitors are known to be adjusting their prices in response to online price changes, and so do supermarkets and other chains. We ask the following questions: (i) does exclusive access to customer-level data create incentives for market leadership? (ii) are there anti- or pro-competitive effects from exclusive access to data?

Approach and methodology

We present a game theoretic model of price competition with horizontal product differentiation in which one firm has access to exclusive customer data. This data, which covers only a share of all consumers, enables the firm holding the data to price discriminate the profiled consumers. The rest of the consumers remain anonymous to both firms. In our setting, the two firms independently and simultaneously decide on whether to announce their price “early” or “late.” If their choices coincide, then simultaneous competition follows. Otherwise, the firm that selects “early” announces its price first and acts as a price leader. Then, the other firm sets its price after observing the rival’s move.

Key findings and conclusions

Exclusive data creates incentives for market leadership if a sufficient number of consumers are profiled. In particular, when an intermediate number of consumers is profiled, market leadership also entails a semi-collusive outcome in which market segmentation arises as an equilibrium feature. By acting as a price leader, the data holder can set a very high posted price, which will then be undercut by the other firm. In equilibrium, the data holder covers only the market with profiled consumers who are price discriminated, whereas the rival firm serves only non-profiled consumers. This way, the data holder can distort market competition and lead to remarkably high prices. However, when too many consumers are profiled, competition intensifies, and this leads to lower prices and substantial benefits to consumers. This paper contributes to the current literature on data-driven practices and the implications reveal that competition policy concerns arise on top of privacy ones.