Associate Professor Linli Xu

The rapid growth of digital content platforms has led to information overload for consumers; hence, different interfaces have been widely used to facilitate the process of digital content consumption. In general, most recommendation systems could be classified into two types. The first type, self-navigation recommender, generates a list of alternatives for the consumer to browse and choose from, while the other type only recommends a single item embedded in a chosen content. With coexistence of both recommenders, users commonly switch back and forth between exploration and consumption.

This project examines this dynamic consumer engagement across various activities using data from a short-video mobile platform. The researchers' model-free evidence suggests a non-sequential response path under the self-navigation mode and users’ dynamic engagement between content exploration and consumption mode. Following these findings, the researchers constructed an empirical model encompassing a sequence of micro decisions to describe consumers’ browsing and viewing decisions and how their prior engagement affects those decisions. They first find that the satiation effect of engagement with the app will lead to less subsequent engagement in both exploration and consumption. However, a higher satisfaction level from prior consumption could drive more subsequent browsing and viewings. We also find that users’ variety-seeking tendency significantly affects the exploration and consumption processes.