Comparing Platform Core Features with Third-Party Complements. Machine-Learning Evidence from Apple iOS.

Abstract

Software-based platforms have become omnipresent both in private and professional contexts. Platform owners constantly invest in platform evolution in that they update the technological core and enrich its feature base. The question arises how such platform core feature changes can be compared with third-party complements. We investigate this question in the context of an exploratory machine-learning based case study on Apple’s mobile platform iOS. By analyzing the changes to iOS over time and developing an approach using natural language processing, we are able identify functional overlaps between platform core features and complements. Our results suggest that platform core features are indeed functionally related to those of complementors and that the strategy of releasing novel platform core features changes over time. Besides, our approach enables us to assign platform core features to app categories. The analysis of functional overlaps raises relevant implications for research and practice.