Pricing Externalities in Real-Time Bidding Markets

Online publishers rely on real-time bidding (RTB) markets to sell their remnant inventory and increasingly to command higher prices for “premium” content. However, content providers have lagged advertisers historically in the sophistication of their pricing models, as evidenced by the large increase in demand-side platforms without corresponding investment on the sell-side. Informed advertisers collect user-intent and demographic information in addition to the publisher context in order to score the relevance of impressions. The resulting differential impression pricing is only visible to publishers as a positive externality to revenue unless they collect the same audience targeting information. In this paper, we introduce a Bayesian hierarchical model of auction clearing price that can naturally account for the presence of publisher-advertiser information asymmetry and quantify its impact on price. Although the current model is simply exploratory, it suggests that richer models of dynamic floor pricing may improve publisher revenue.