Investigating the android apps' success: An empirical study

Measuring the success of software systems was not a trivial task in the past. Nowadays, mobile apps provide a uniform schema, i.e., the average ratings provided by the apps' users to gauge their success. While recent research has focused on examining the relationship between change-and fault-proneness and apps' lack of success, as well as qualitatively analyzing the reasons behind the apps' users dissatisfaction, there is little empirical evidence on the factors related to the success of mobile apps. In this paper, we explore the relationships between the mobile apps' success and a set of metrics that not only characterize the apps themselves but also the quality of the APIs used by the apps, as well as user attributes when they interact with the apps. In particular, we measure API quality in terms of bugs fixed in APIs used by apps and changes that occurred in the API methods. We examine different kinds of changes including changes in the interfaces, implementation, and exception handling. For user-related factors, we leverage the number of app's downloads and installations, and users' reviews. Through an empirical study of 474 free Android apps, we find that factors such as the number of users' reviews provided for an app, app's category and size appear to have an impact on the app's success