Sympathy to the Seemingly Needy : A Large-Scale Field Experiment on Social Influence and Non-Social Signals in Medical Crowdfunding

While existing research has examined the role of social influence on various user behavior, such as product adoption, reviews and purchases, there is a lack of studies on how social influence interacts with inherent non-social cues, such as inherent attributes. In this study, we attempt to fill in this gap, using medical crowdfunding as the research context, where information asymmetry between donors and fundraisers potentially hinders donation decisions. We first evaluate how donors leverage inherent fundraising case information, such as patient age, gender and type of disease, to assess the validity of the case. Then, through a large-scale randomized field experiment, we examine the impact of social influence on users' willingness to donate and how this impact changes, depending on the presence of non-social signals with varying signal strength. Our results show that the weaker non-social signal, patient gender, influences the likelihood to donate only for cases lacking stronger non-social signals. More specifically, when social influence is present, donors' willingness to donate increases but the impact of the weaker non-social signal is reduced. In contrast, for cases with multiple strong non-social signals, neither social influence or patient gender changes the willingness to donate; for cases with a single strong non-social signal, social influence improves the likelihood to donate, but patient gender does not. Overall, our findings indicate that the informational value of social influence is dependent on the presence of alternative information sources, such as inherent strong or weak non-social cues. These results also provide insights on how to leverage social influence, considering the relative informational value compared with inherent non-social cues. The informational value of social influence is particularly important for cases lacking the strong non-social signals to appear in strong need of help. Our results also provide insights on improving chances of funding success for medical crowdfunding