Understanding Content Contribution Behavior in a Geo-Segmented Mobile Virtual Community: The Context of Waze

Abstract

We examine content creation in a geo-segmented, crowdsourced social mobile virtual community app, Waze. We conceptualize a virtual and spatial factor, virtual crowdedness (defined as the density of Waze users in a particular geospatial location) and examine its role in encouraging user contribution. We posit that the relationship between virtual crowdedness and user contribution is driven by the tension between audience effects and bystander/content saturation effects. We analyze a panel dataset of user contributions on Waze from New York City to test our hypotheses. First, our findings indicate that although virtual crowdedness has a positive influence on total number of contributions, the magnitude of the influence decreases as virtual crowdedness increases. Second, the concave-down, increasing relationship is more pronounced for rush hours with high physical crowdedness than for non-rush hours with low physical crowdedness. A variety of robustness checks and alternative analyses based on matching estimators and spatial econometric models further support the main conclusions while mitigating concerns about endogeneity and spatial autocorrelation. Our findings provide several key practical implications for platform designers in that they should allow for users to visualize density of usage as well as improve the design of social features for encouraging user contribution to the mobile virtual community.

Keywords: mobile virtual community, user contribution, virtual crowdedness, prosocial behavior, free-riding behavior, matching estimators, spatial econometric analysis

JEL Classification: H42, L86, M15, C33, C55