Adaptive Ad Network Selection for Publisher-Return Optimization in Mobile-App Advertising

摘要

In the era of online advertising, the new norm of mobile-app advertising defines a novel revenue generation channel for publishers through user- and context-targeted advertisements. Unlike online advertising, mobile-app advertising has unique behaviors due to certain constraints and attributes. As a result, the existing solutions of return optimization for publishers do not always provide the expected outcome. This has created a new research gap. Finding a solution at the app instance level of the mobile advertising ecosystem has a high potential to bridge this gap. This study provides a full-fledged mechanism to determine the ad network, which gains the highest return to the publisher at the app instance level based on the attributes of both the advertisement and the ad network. Using such attributes and the mobile-app user's click behavior, we estimate the ad network effectiveness, the advertisement effectiveness, and the click-through rate to determine the optimal ad network, which provides the highest return for the publisher. Through a simulation experiment based on data generated in real-life scenarios, we demonstrate that the publisher-return is higher in our proposed approach than that obtained from advertisements from a single ad network for all the mobile-app users.