Search, targeting and media

University Of Virginia Main Campus, Charlottesville VA

Investigators

Abstract

Search, targeting, and media. The project innovates 3 topical themes. The intellectual contribution of the proposal is to push forward the economic analysis of the Internet and new communication technologies. The broader impact is in engaging new techniques and models to analyze these modern communication phenomena that encompass Marketing and Media, and in highlighting the inherent differences among firms. First, consumers searching on the web frequently follow the order in which sponsored links are presented. Moreover, firms pay more for higher slots via a position auction conducted by the search engine. A tractable model of pricing under directed search is proposed and integrated with a position auction for better slots (which rationalizes the way consumers search). Although the position auction induces the right order of searching, firms tend to price so as to discourage further exploration, and thus consumers don?t search enough. ?Better? firms (with higher quality products) tend to bid more for earlier positions, even though product prices tend to increase in the order of search. Bids for slots factor in the position interdependence across firms because product prices and profits depend on the qualities of products both before and after in the sequence of positions. Second, targeted advertising will imminently transform marketing by inducing competition for individual consumers. Analyzing the impact of such targeting is rendered tractable because it is shown that expected profits per consumer (technically, in a mixed strategy equilibrium of both prices and advertising) can be characterized in a simple familiar form. Using this ingredient enables the study of price competition among firms that rationally anticipate subsequent poaching and retention through targeted advertising. In a benchmark first case, market base prices are neutral to targeting costs, so firms lose and consumers gain through the subsequent extra competition. However, in broader contexts firms can be better off despite extra marketing expenses when targeting facilitates higher base store prices because the margin of competition is shifted. Consumers are then worse off in aggregate, and further market inefficiency is induced through poaching. The third theme engages competition policy for media markets. The common business model of financing through ad revenue is known as a ?two-sided? market structure because two groups (consumers and advertisers) are brought together on a media platform (web-site, magazine, TV program, etc.) The project exploits a novel ?aggregative game? approach to such two-sided media markets that ties together the impacts of platforms? actions on both sides of the market and shows they are diametrically opposed. This is the ?see-saw? effect that structural changes adversely affecting one side of the market can benefit the other side. In this context, a merger worsens viewer welfare in the classic way of modeling media economics, but can improve their welfare in an alternative specification when ads clutter viewer attention. In both cases, the effects on advertiser profits go in the opposite direction. The analysis is noteworthy for delivering results on mergers of heterogeneous firms, and in a two-sided setting on top. All three themes make foundational contributions in proposing how to progress tractably, and highlight the fundamental role of firm heterogeneity that characterizes markets and their performance. The search framework delivers a full-fledged integration of position auctions and pricing with sequential directed search. This set-up describes the search environment on the Internet, and should open the way for further research development. The targeting work sets the stage for structural empirical analysis of big data and its use by firms in targeting. The aggregative game framework applied to two-sided markets stresses impacts of market structure changes on three groups of economic actors (viewers, advertisers, and platforms) and shows how these impacts are tied together.