Abstract: With the rapid growth of the smartphone andtablet market, mobile application (App) industry thatprovides a variety of functional devices is also growingat a striking speed. Product life cycle (PLC) theory, whichhas a long history, has been applied to a great numberof industries and products and is widely used in themanagement domain. In this study, we apply classicalPLC theory to mobile Apps on Apple smartphone andtablet devices (Apple App Store). Instead of trying toutilize often-unavailable sales or download volumedata, we use open-access App daily download rankingsas an indicator to characterize the normalized dynamicmarket popularity of an App. We also use this rankinginformation to generate an App life cycle model. Byusing this model, we compare paid and free Apps from20 different categories. Our results show that Apps acrossvarious categories have different kinds of life cycles andexhibit various unique and unpredictable characteristics.Furthermore, as large-scale heterogeneous data (e.g., userApp ratings, App hardware/software requirements, or Appversion updates) become available and are attached toeach target App, an important contribution of this paperis that we perform in-depth studies to explore how suchdata correlate and affect the App life cycle. Using differentregression techniques (i.e., logistic, ordinary least squares,and partial least squares), we built different models toinvestigate these relationships. The results indicate thatsome explicit and latent independent variables are moreimportant than others for the characterization of Applife cycle. In addition, we find that life cycle analysisfor different App categories requires different tailoredregression models, confirming that inner-category Applife cycles are more predictable and comparable than Applife cycles across different categories.

Keywords: data mining, life cycle, digital market, app,app store, regression