Advertising on Online Marketplaces: Information Asymmetry and the Relevance of Sponsored Listings

Abstract

Advertising on e-commerce marketplaces, wherein sponsored product listings are interleaved with organic product listings, is a large and growing phenomenon. In this paper, we both theoretically and empirically study whether including sponsored listings improves or hurts the overall quality and relevance of the set of products displayed to consumers. Our theoretical analysis reveals that the relevance of the ads to the consumers' search queries depends on the level of information asymmetry between the marketplace owner and the sellers who sell products with different degrees of relevance. Specifically, when information asymmetry is low (high), i.e., the platform can (cannot) easily distinguish between high- and low-relevance sellers, then low-relevance (high-relevance) sellers have a greater incentive to advertise. However, even when low-relevance products are displayed as ads, consumers end up finding well-matching products as long as search and evaluation costs are reasonable; therefore, the overall impact on sales is relatively small while the marketplace benefits from the additional revenue from selling ads. We obtain data from a large-scale field experiment run at Flipkart, a leading online marketplace in India, and find that various empirical patterns implied by our theoretical results hold in the data. Our study provides several practical implications for managers of online marketplaces.

Keywords: E-commerce, Sponsored Search, Advertising, Product Heterogeneity, Platform, Asymmetric Information

JEL Classification: M31, M37, L81, D82, D83, C93