Essays on Marketing Strategies in Online Advertising

Abstract

Digital advertising expenditures in the United States reached 139.8 billion in 2020, a 12.2% increase from 2019. Along with the rapid growth of ad spending, ad platforms adopt several marketing strategies, aiming to make online advertising more effective. In this dissertation, I model and provide insights to some of the main challenges that online advertising currently faces. In the first chapter, I study advertisers’ learning of consumer information enabled by tracking technology and examine how the learning incentives affect advertisers’ bidding strategies across time. It is shown that when the differentiation between advertisers is low, the pursuit of individualized information motivates advertisers to overbid (bid more than expected consumer valuation) and the learning of consumer information increases advertisers’ auction competition in the first stage. On the other hand, when advertisers are highly differentiated, our findings suggest that despite the value of information, advertisers may prefer being ignorant and thus are motivated to start with underbidding (bid less than expected consumer valuation). Thus, the learning of consumer information could decrease advertisers’ auction competition in the first stage. Furthermore, our analysis also shows that it may not always be in the ad platform’s best interest to adopt tracking technology that allows advertisers to learn consumer information over time. In the second chapter, I study the quality score, an index introduced by ad platforms to rank advertisers in auction, and examine optimal quality investments for heterogeneous advertisers to increase their quality score. The results suggest that a quality score function that rewards bidders in excess of their click-through rates does increase the quality investments but in a non-monotonic way. Higher rewards of landing page quality do not necessarily mean higher investments. Furthermore, such an improvement in quality levels often comes with a sacrifice of the auction revenue despite the higher bids. While the potential revenue impact for the auctioneer depends on the bidding strategies and auction formats, there exists a quality score function that ensures increased revenues by rewarding quality improvement.