Crafting magic: Building a predictive model based on donor affinity

At the University of Waterloo (UW), the Annual Giving programmes aim to scale fundraising efforts while providing a personalised experience for donors. By analysing collected data, institutions can gain insights into donors’ interests, enhancing engagement and philanthropic contributions. Limited resources, however, often hinder the effective implementation of mass personalisation. This paper explores the challenges and opportunities faced by the UW in improving response rates for bulk appeals, such as Giving Tuesday, Renewal and Short Lapsed campaigns. The traditional ‘Last Gift’ segmentation method, while straightforward, is cumbersome and relies heavily on manual data preparation. To address these issues, we developed a prototype donor affinity model (DAM) that leverages donor data to predict and align fundraising efforts with donor interests. This model aims to enhance the effectiveness of our annual giving programmes by moving beyond last-gift analysis to a more comprehensive understanding of donor behaviour. This paper focuses on improving response rates for bulk appeals, starting with Giving Tuesday, using a donor affinity approach. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.