There is no doubt that we are living in the mobile age. Massive commercial data from mobile users provide huge potential opportunities for both merchants and researchers. However, compared with Web analytics, how to analyze the mobile data is still under debate. Some companies have adopted similar measurements or metrics to those that have been used in Web analytics tools to depict mobile activities. However, such measurements might bias the interpretation because the behaviors on mobile devices are fundamentally different from behaviors in an online context. Hence, we propose a prototype of an Android-based mobile analytics tool with novel mechanisms for better observing user behaviors in the mobile context, by which the individual behavioral data can be captured at the activity level. The detailed design, such as prototype structure, data structure, specific classes or methods, and sample codes, will be provided. In addition, some potential research opportunities will be elaborated. Furthermore, the practical value of our proposed prototype will be discussed.