Bellwether and the herd? Unpacking the u-shaped relationship between prior funding and subsequent contributions in reward-based crowdfunding

Abstract

Building on the insights from observational learning and other social influence research, this study challenges the existing literature that proposes a linear relationship between prior funding and subsequent contributions in the crowdfunding setting. Instead, we propose a U-shaped relationship, illustrating a negative relationship between prior funding and subsequent contributions when funding amounts are small and this relationship becomes positive when prior funding amounts are large. Consistent with the rational herding perspective, we assume that individuals do not mindlessly mimic one others' behaviors, but incorporate quality signals into their decision-making. Signals indicating a high quality project such as video quality, as well as the preparedness and passion of the entrepreneur enhance this U-shaped relationship whereas indicators of situational urgency weaken this relationship. Based on a sample of 11,019 daily observations of 333 Kickstarter projects, we find general support for our hypotheses. Our post hoc experiment further supports the mechanisms underlying the U-shaped relationship. This paper extends our understanding of the relationship between past actions and subsequent behavior, and contributes to the literatures of observational learning, crowdfunding, and visual information. Executive summaryWe challenge the extant literature that has proposed, but inconsistently documented, a positive effect of prior funding on subsequent crowdfunding contributions. We argue that such inconsistency is due to the lack of integration of insights from other social influence research, such as observational learning, threshold models, and bystander effects. This research has been limited when postulating that crowdfunding backers simply imitate others' actions in order to mitigate uncertainty producing a linear effect, yet such an effect depends on the intensity of others' actions and may be curvilinear. This research overlooks the fact that backers do not merely passively imitate others' behaviors, but incorporate quality signals sent by entrepreneurs to make such decisions. Thus, it is important to account for potential moderators that may influence the complex relationship between prior funding and subsequent contributions.Based on observational learning research and related models, accrued prior funding may be perceived differently by potential backers: individuals may view accrued funding as a sign of a decreased funding need on the part of a campaign's creators, because others are already backing the campaign or, alternatively, they may see it as a cue of high product quality. These conflicting effects may lead to a U-shaped relationship between prior funding and subsequent contributions at a collective level. At low funding levels, prior funding may be negatively associated with subsequent contributions because such levels indicate there is already support, but that the campaign may not be of high quality. At medium levels, prior funding may have no effect on contributions, aside from halfhearted interest of other backers. At high levels, prior funding may be positively related to subsequent contributions as this sends a robust cue of project quality and convinces prospective donors to contribute. Further, we derive a U-shaped relationship which may be moderated by visual media-based quality signals and situational urgency.To test these hypotheses, we conducted two studies. One was a field study of 333 Kickstarter projects, with 11,019 daily observations of crowdfunding activities from mid-February 2013 through mid-April 2013. We used the Kicktraq website to track Kickstarter projects and recruited 390 participants via Amazon Mechanical Turk (MTurk) to evaluate the video pitches on Kickstarter. Our results suggest a negative relationship between prior funding and subsequent contributions when funding amounts are small. This relationship becomes positive when prior funding amounts are large. Thus, we find a U-shaped relationship between prior funding and subsequent contributions, with an inflection point when prior funding amounts reach between 72% and 80% of the funding goal. This is strengthened in the presence of visual media-based quality signals, such as video quality, as well as the preparedness and passion of the entrepreneur. However, indicators of situational urgency may weaken this U-shaped relationship.We followed up with a post hoc experiment to document the two mechanisms by which prior funding influence subsequent contributions. We found that increases in the prior funding amount increased perceived product quality but decreased perceived urgency. These perceptions were found to jointly influence participants' proposed pledged amounts, supporting an explanation of a complex relationship between prior funding and subsequent contributions.This study is among the first to offer a more complete picture of the relationship between past actions and subsequent behavior under market and institutional constraints (i.e., when market actors observe the censored behaviors of other actors on crowdfunding platforms) and contributes to the literatures of observational learning, crowdfunding, and visual information.