Hidden Influences of Crowd Behavior in Crowdfunding: An Experimental Study

Crowdfunding continues to transform financing opportunities for many across the globe. While extensive research has explored factors related to fundraising success, less is known about the social signaling mechanisms that lead potential contributors to fund a project. Existing large-scale observational studies point to non-straightforward characteristics of prior contributions (aka "crowd signals") that forecast further contributions to a project, albeit without theoretical support for their effectiveness in predicting fundraising success. We translate empirical crowd signals based on variations in the amounts and timings of contributions into mock contribution scenarios to scrutinize the influence of essential signals on contributors' decisions to fund. We conduct two experiments with 1,250 online participants. The first experiment investigates whether high crowd signals, i.e., contributions of varying amounts arriving at unequally spaced time intervals, are making people more likely to contribute to a crowdfunding project. The second experiment further examines the effect of basic competition on the role of the crowd signals. Across both, we observe that high crowd signals attract 19.2% more contributors than low signals. These findings are robust to different project types, fundraising goals, participants' interest level in the projects, their altruistic attitudes, and susceptibility to social influence. Participants' unguided, post-hoc reflections about the reasons behind their choice to fund revealed that most were unaware of their reliance on any crowd signals and instead attributed their decision to nonexistent differences in project descriptions. These results point to the power of crowd signals unbeknownst to those affected by them and lay the groundwork for theory-building, specifically in relation to the essential signaling that is happening on online platforms.