Using Machine Learning to Understand and Manage the Transformation of Peer Donors to Organizational Donors

ABSTRACT

Peer‐to‐peer fundraising has become a popular funding approach for nonprofit organizations, generating quick revenue and a promising opportunity for donor base expansion by transforming peer donors into organizational donors following their peer donation. This study uses survey data from 706 participants to examine peer donors' transformation likelihood and its determining factors. Additionally, it evaluates the capacity of machine learning to predict which peer donors are most likely to transform. The results reveal that, among peer donors who lack prior affiliation with the nonprofit organization, the transformation likelihood is 14.1%, indicating a transformation rate of approximately one in seven peer donors. Regarding the determining factors, post‐donation communication with peer donors after their initial donation increased the odds of transformation threefold, while established nonprofit‐related factors, such as trust in the organization, exhibit no influence. Moreover, applying the random forest approach allowed for the prediction of the transformation with an accuracy of 79% slightly outperforming logistic regression. This study assists in identifying the donors most likely to transform, enabling fundraising managers to allocate efforts effectively and maximize fundraising success.