Retain, reactivate or acquire: Can nonprofits reliably use community profiles as an alternative to past donation data?
研究发现非营利组织可以使用社区层面的捐赠者档案来预测捐赠行为，其准确性与使用实际捐赠数据相当，有助于制定针对活跃、流失和潜在捐赠者的策略。
• Community profiles can be an alternative to past donation data in predicting expected donations. • Profile data is critical for simultaneous active donor retention, lapsed donor reactivation as well as prospective donor acquisition. • Even in the presence of solicitation endogeneity, nonprofits can predict donation behavior using community-clustered profiles. • Behavior of donors with varying degrees of behavioral heterogeneity (in terms of durations) and different lifetime relationships with nonprofits can still be predicted using the profiles. • Nonprofits do not need comprehensive donor profiles, even limited information about characteristics of donors at the community level is sufficient for predicting their behavior. Nonprofits face the challenge of low response rates to solicitations, leading to unachieved fundraising goals. They face difficulty in retaining active donors, reactivating lapsed donors, and acquiring prospective donors. The challenge often stems from the need for more reliable data for predicting the expected behavior of different groups of donors. Although nonprofits have reliable data relating to past donations from active donors, the data on lapsed donors is limited, and data on prospective donors is nonexistent. We propose that nonprofits can use community-clustered donor profiles to predict the expected donations. Our results validate that predictions based on “actual donation data” and “community donor profiles” are equivalent in accuracy. Drawing insights from the nonprofit marketing and social psychology literature, we suggest that nonprofits can reliably devise targeting strategies for active, lapsed, and prospective donors using community-clustered profiles. We test these predictions using a donation incidence model and conduct several robustness checks.