Customer Acquisition, Engagement, and Retention in Online Advertising

Abstract

Online advertising continues to evolve at a rapid as the internet and the digital marketing landscape mature. Firms face new challenges in acquiring, engaging and retaining customers. Entire new markets and technologies have grown out of the race digital marketing dominance. This dissertation aims to examine some of these advances and offer practical insights for today’s firms that need to navigate this new world. In the first essay, I explore the effects of user attention to online display advertising. Using two observational studies, I show that attention is highly heterogeneous and predictable during the user browsing session. The implications are that publishers should be more selective in ad placement and that advertisers should be more selective in ad purchases. The second essay examines how programmatic advertising firms should efficiently allocate ads in real-time bidding environments on behalf of their client advertisers. I introduce the demand side platform problem which is related to both the adwords and publisher problems, but distinct in that the supply of ad space assumed unlearnable. I provide a real-time mechanism for efficient ad allocation in this setting and demonstrate efficacy using real-time bidding data. In the final essay, I examine cross-merchant spillovers in coalition loyalty programs. I examine a natural experiment where a large grocery store joined a large loyalty program coalition. Using a quasi-difference-in-difference approach and Bayesian Structural Time Series for causal inference, I find that adding a large complementary merchant into a coalition loyalty program increases sales and purchase frequency of existing customers at existing merchants.