Effectiveness of Targeted Digital Marketing

Abstract

Digital technologies have contributed to significant changes in the targeting of marketing. For example, the music industry---where consumers used to be targeted directly, e.g., through ads in print media, on the radio, on TV and online---now focuses its promotional efforts on the curators of playlists from major music streaming services. Playlist curators are important tastemakers that can influence the demand of consumers for tracks and hence the revenue of record labels and artists. A key challenge for managers in the music industry is to maximize a track's expected revenue by targeting the most effective playlists. Though the scale of the challenges posed by digitization may be largest in sectors with predominantly analog business models, new technologies have also transformed inherently digital industries such as online display advertising. While many advertisers consider display advertising an important---if not the most important---marketing tool, there are concerns over its effectiveness. Managers are interested in new targeting frameworks that increase the effectiveness of online display advertising. The goal of this dissertation is to identify important challenges in digital marketing and to provide targeting-based solutions. It illustrates these challenges and solutions by using examples from online display advertising and music streaming. The paper looking at online display advertising describes how advertisers can target users at different stages of the purchase process with different ad content, thus better satisfying their informational needs. The other papers focus on the targeting of curated playlists in digital music streaming. More specifically, the second paper describes an estimation framework for gauging the incremental effects of a track's playlist inclusions on streams. It uses the estimation framework as the basis for an attribute-based targeting approach that can predict the most effective playlist for a given track---even if the track is not yet released. The third paper looks at the factors affecting how playlist curators create playlists---i.e., whether and where (at which positions)---tracks are included on playlists. Thus, the second and third paper jointly allow record labels, artists and managers to improve the targeting of curated playlists. A common thread of the papers of this dissertation is their usage of empirical data to answer important research questions. For example, the first paper uses data from an online scenario experiment and data from a digital field experiment. Data from the online scenario experiment are used to describe the perceived effectiveness and likely popularity of different ad content among marketing managers. The digital field experiment allows for gauging the effectiveness of ads with different content with respect to their influence on consumer behavior in early and late stages of the purchase process. Similarly, the second and third paper improve the targeting of curated playlists by using field data. The field data come from a major international record label and are supplemented with data from Spotify, which is the world's largest music streaming platform, and data from a music analytics provider. The papers of this dissertation contribute to the marketing literature in different ways. First, all papers identify important challenges in the targeting of marketing and describe how these challenges relate to previous marketing literature. Second, the papers develop new conceptual frameworks and estimation approaches that can be used to predict the effects of marketing tools on consumer behavior and hence revenue. Third, the papers propose solutions that allow for increasing the effectiveness of marketing through targeting and usage of empirical data, e.g., field data from industry partners. The papers of this dissertation contribute to literature on online display advertising, (purchase) goal framing, music streaming and playlists.