Predicting Behavior of Current and Future Donors: Are Geographic Profiles an Alternative to Past Donations?
(A2020-64517)

AUTHORS

Shameek Sinha, IE Business School, IE University; Vijay Mahajan, Red McCombs School of Business, The University of Texas at Austin; Frenkel ter Hofstede, Red McCombs School of Business, The University of Texas at Austin; Sumit Malik, IE Business School, IE University, Spain

KEYWORDS

Non-profit targeting; Current vs. Future Donors; Profile vs. Behavioural Data

ABSTRACT

With competing nonprofits soliciting donations for alternative causes, abysmal response rates to solicitations makes meeting fundraising goals challenging. This results from failure to retain current donors and lack of potential acquisition targets as future donors. While nonprofits have past donation data from current donors, no such data is available for future donors. We propose that nonprofits could use geographic profiles of future donors to predict their expected behavior in absence of past donations. Specifically, we find that donations and profile-based predictions do not differ in statistical accuracy. This implies that profiles, which explain the observed heterogeneity in donations, can be used to predict expected behavior of donors. Further, prediction of donations might improve with the use of geographic profiles compared to donation data. This suggests that there is geographic clustering of pro-social behavior driven by the concentration of donor within geographic regions.