Price customization and targeting in many-to-many matching markets

Recent technologies permit matching intermediaries to engage in unprecedented levels of targeting. Yet, regulators fear that the welfare gains of such targeting be hindered by the high degree of price customization practiced by matching intermediaries, whereby prices finely depend on the characteristics of the matching partners. To shed light on this debate, we develop a matching model in which agents' preferences are both vertically and horizontally differentiated. Mirroring current practices, we show how platforms maximize profits by offering menus of matching plans defined by (a) a baseline configuration, (b) a baseline price, and (c) a collection of nonlinear tariffs for customization. We illustrate how, under such plans, prices are linked to structural elasticities, and derive primitive conditions under which market power distortions increase with the targeting level of a match. We then study the effects on targeting and consumer welfare of uniform-pricing regulation mandating that the price charged to the side-i agents be invariant in that side's observable characteristics (e.g., the requirement that the price charged to advertisers be invariant in the ads' content). Finally, we examine the transition of matching markets from a centralized structure to a decentralized one where sellers post prices and matching is unmediated. The analysis has implications for ad-exchanges, media platforms, cable TV, business-to-business platforms, and large online retailers.