Real-Time Bidding in Online Display Advertising

Abstract

Display advertising is a major source of revenue for many of the online publishers and content providers. Historically, display advertising impressions have been sold through pre-negotiated contracts, known as reservation contracts, between publishers and advertisers. In recent years, a growing number of impressions are being sold in real-time bidding (RTB), where advertisers bid for impressions in real-time, as consumers visit publishers’ websites. RTB allows advertisers to target consumers at an individual level using browser cookie information, and enables them to customize their ads for each individual. The rapid growth of RTB has created new challenges for advertisers and publishers on how much budget and ad inventory to allocate to RTB. In this paper, we use a game theory model with two advertisers and a publisher to study the effects of RTB on advertisers’ and publishers’ strategies and their profits. We show that symmetric advertisers use asymmetric strategies where one advertiser buys all of his impressions in RTB whereas the other advertiser focuses on reservation contracts. Interestingly, we find that while both advertisers benefit from existence of RTB, the advertiser that focuses on reservation contracts benefits more than the advertiser that focuses on RTB. We show that while RTB lowers the equilibrium price of impressions in reservation contracts, it increases the publisher’s total revenue. Despite many analysts’ belief that, because of being more efficient, RTB will replace reservation contracts in the future, we show that publishers have to sell a sufficiently large fraction of their impressions in reservation contracts in order to maximize their revenue. We extend our model to consider premium consumers, publisher’s uncertainty about the number of future visitors, and effectiveness of ad customization.