Finding innovation patterns of mobile services using update histories: a data-driven taxonomy approach

The mobile application market is highly competitive; thus, many companies struggle to secure a competitive advantage. As a means of continuous innovation, updates are important means to enhance visibility, increase the number of downloads and meet customers’ requirements. Despite the importance of updating and maintaining mobile applications, there have been very few attempts to analyse updated patterns. Previous studies mostly discussed the effect of updates on performance indicators such as customer satisfaction and number of downloads, neglecting how to analyse innovation patterns of mobile services. This study employs text mining for patch notes, develops data-driven indicators to represent the update and innovation patterns and identifies some clusters for the update patterns. As a result, four types of innovation patterns are suggested: fast & responsive, young & innovation-focused, sustaining & safe improvement and silence & avoiding. For each cluster, how innovation patterns differ is analysed and discussed.