Optimizing Giving Day: A Case Study in Using Machine Learning for a Constituency-Wide Campaign

Autor: Pelletier, Marianne M.
Zdroj: Journal of Advancement Analytics; 20240101, Issue: Preprints p3-11, 9p
Abstrakt: Abstract:Fundraising data science has moved into the annual giving and engagement realm, providing improvement in segmentation and a pathway to using artificial intelligence tools. This expansion has democratized the use of data science tools and broadened the use of methods that retailers currently employ, which is one reason why direct mail and phonathons are being rebranded as “direct marketing.” Ithaca College’s Advancement Office conducted its first Giving Day campaign in 2021. Though the campaign was successful, with $1.5 million garnered against a $1 million goal, the chair of the Philanthropic and Engagement Committee asked the Advancement staff to boost totals for 2022 to $2 million. Vice President Wendy Kobler contracted Staupell to build out optimized ask amounts with probability to give for all constituents. In this paper, we explore this single case study—combining an optimized ask amount and probability both to predict and set ask amounts and priority order for Ithaca College’s 2022 Giving Day campaign. Discussion includes methods used, forecasted versus actual results, issues and benefits from the project, and how staff used and reacted to the project. The paper shows that data science played a key role in the campaign’s success and discusses both wins and opportunities for improvement in a further study. This case study can be used as a model for any annual giving campaign.
Databáze: Supplemental Index