Optimal Platforms for Search and Matching

Abstract

Search and matching increasingly take place on online platforms with elements of centralized and decentralized matching. Platforms can alter the search process for users, but cannot eliminate search frictions entirely. I study a model where users pay to change the distribution of potential partners they search over and characterize the revenue-maximizing platform when match characteristics are private information. With complementary match characteristics, I show that the only possible source of inefficiency on the optimal platform is exclusion, unlike standard screening problems. Matching on the optimal platform is perfectly assortative --- there is no equilibrium mismatch. This conclusion is robust to various manners in which fees are collected (e.g. subscription fees or match fees), and to environments where the platform designer can also change search speeds.