Are Personalized Recommendations the Savior for Online Content Providers?
Advertising seems to be the major stream for generating revenue with online content. Therefore, it is crucial for online content providers to create heavy traffic on their pages. Recommender systems have already been evaluated on their positive impact in e-commerce settings. In our evaluation we show that personalized recommendation as well change the user behavior and improve the relevant economic KPIs for content providers. To show these effects we put up a model that inter-relates all components of an advertising-based revenue stream. We formulate a set of hypotheses on those components, which can be influenced by recommender systems: exposure and stickiness, indicating customer engagement, as well as the content portfolio that impairs the appealed user base. The hypotheses are tested in a study that is being conducted based on a real-world data set from a German newspaper.