Market Segmentation Through Information

Abstract

Prodigious amounts of data are being collected by internet companies about their users' preferences. We consider the information design problem of how to share this information with traditional companies which, in turn, compete on price by offering personalised discounts to customers. We provide a necessary and sufficient condition under which the internet company is able to perfectly segment and monopolize all such markets. This condition is surprisingly mild, and suggests room for regulatory oversight.

Voluntary Disclosure and Personalized Pricing

“…Several papers also study the role of information in competitive markets with differentiated products. Elliott et al (2020) show that an information designer can segment the market so that consumers are allocated efficiently while nevertheless guaranteeing that consumers obtain no surplus. Armstrong and Zhou (2021) study optimal information structures for a consumer that does not know her tastes, and show that the consumer-optimal signal may involve learning a little so as to amplify price competition.…”

Consumer Information and the Limits to Competition

“…If all ways to partition consumers are possible, the paper shows that any combination of pro…t (above the no-discrimination benchmark) and consumer surplus which sum to no more than maximum total welfare can be implemented. Elliot, Galeotti, and Koh (2020) extend Bergemann et al to the competition case with product di¤erentiation, and derive conditions under which market segmentation through information can earn …rms 8 When we study the "top product"signal structure in section 4.3, the resulting posterior valuation distribution is binary. There we construct the mixed strategy pricing equilibrium similarly as in Moscarini and Ottaviani (2001) but with an arbitrary number of …rms.…”

Data Sharing and Incentives

“…Our paper contributes to an emerging literature regarding the welfare consequences of data markets and algorithmic scoring. This literature has tackled several important social questions, such as whether predictive algorithms discriminate (Chouldechova, 2017;Kleinberg et al, 2017); how to protect consumers from loss of privacy (Acquisti et al, 2015;Dwork and Roth, 2014;Fainmesser et al, 2019;Eilat et al, 2019); how to price data (Bergemann et al, 2018;Agarwal et al, 2019); whether seller or advertiser access to big data harms consumers (Jullien et al, 2018;Gomes and Pavan, 2019); and how to aggregate big data into market segments or consumer scores (Ichihashi, 2019;Bonatti and Cisternas, 2019;Yang, 2019;Hidir and Vellodi, 2019;Elliott and Galeotti, 2019). There is additionally a growing literature about strategic interactions with machine learning algorithms: see Eliaz and Spiegler (2018) on the incentives to truthfully report characteristics to a machine learning algorithm, and Olea et al (2018) on how economic markets select certain models for making predictions over others.…”