How Is Mobile User Behavior Different? — A Hidden Markov Model of Mobile Application Usage Dynamics

Abstract

Driven by the ubiquity and strong context dependence of mobile app use, Internet companies are in a race of cross-industry expansion to build a seamless ecosystem incorporating various contexts. This paper offers several insights on improving app use in the era of mobile Internet. In contrast to PC Internet, in addition to hedonic and utilitarian states, we uncover a novel social state that is prevalent but transient, indicating mobile users have a fundamental need for frequent light-social activities. Thus, one strategy to increase use is to enrich an app’s social components, specifically on light-social functionalities. In addition, our results show that app use interdependence is the strongest under the hedonic state. This indicates the strategic value of boosting current app use is the highest in the hedonic state, providing guidance to companies on better spending of their limited marketing resources. Furthermore, we show that these internal states are interdependent of each other and their dynamic is affected by contextual factors that are distinct in the mobile context. Thus, companies should put more weight on tailoring their engagement strategies under different contexts in the ear of mobile Internet than the traditional PC context.