The Effects of Microtargeting on Conversion Rates and the Role of Information Asymmetry

Description

Behavioral advertising targets consumers based on who they are, while contextual advertising targets consumers based on what they see. In many websites today a combination of the two (referred to asmicrotargeting) is used to offer more targeted ads to consumers. There is still an ongoing debate whether more precise targeting (using behavioral data of consumers) can benefit or hurt publishers’ revenues, with evidence supporting both sides of the debate, both from the industry (Ravichandran and Nitish, 2019; Lomas, 2019) and the academic literature (Iyer et al., 2005; Levin and Milgrom, 2010). However, until recently it was a common wisdom that ads which are more tailored to consumers should have higher conversion rates, even in cases where revenues are lower. The idea is simple; with more available data about consumers, the matching between advertisers and consumers should be improved. Recent empirical evidence though suggests that less targeted ads can achieve the same or sometimes even more conversions, compared to microtargeted ones (Snelders et al., 2020).In this project, we propose a game-theoretic model to study this phenomenon and provide a rationale as to why a publisher who turns off microtargeted behavioral advertising, can end up with a higher conversion rate. We link this phenomenon to the information asymmetry that exists in the online advertising market. If a publisher allows the use of third-party cookies on her website, then some advertisers can have access to behavioral data provided by big ad exchanges, which might not be available to everyone. Also, some advertisers might not have the advanced tools necessary to make a good use of all the extra data. We predict that this information asymmetry can result in low efficiency in the market. Surprisingly, turning off microtargeting can result in a more efficient market (better matching) by giving an equal opportunity to everyone, counterbalancing the fact that all advertisers are now less informed.This project contributes to the growing literature on targeted advertising and online advertising auctions. In this proposal, we present some interesting findings regarding the role of information asymmetry in targeted advertising and highlight their implications for the online advertising industry