Re-Use and Patterns of Digital Innovation in Open Crowds

Abstract

This paper examines reuse patterns in transparent multi-staged contests, in which a crowd of developers can reuse others’ designs when developing a novel digital application. They compete for the best application, one that is of high quality in both code and feature. Transparency, defined as collective information access to (1) the design solution, and (2) and the quality of it, triggers variability in reuse choices. Depending on the relative performance, developers adapt their risk preference along a continuum of exploitative to explorative reuse. However, variability in reuse behavior is also shaped by an individuals’ design preference: to focus on either code or feature. To explore patterns of reuse and performance emerging from this micro-level heterogeneity, we develop an agent-based simulation model, informed by theory of reuse and variable risk preferences, as well as emerging empirical data. Our results reveal unexpected discontinuous performance patterns. Design preferences, a bias towards code or feature, lowers the collective performance of the crowd. However, transparency insulates those at the top from such negative effects. It produces high performing designs even if there is bias. Further, while transparency reduces variability and smooths the patterns among those at the top, it also creates unique opportunities for ‘newcomers’ to submit high-quality applications.