Donor Propensity Modeling

The Donor Propensity Model (DPM) is a machine-learning model that estimates a score from 1 to 10 to predict the likelihood that an individual were to make a “major” donation. The DPM model considers various factors including demographic data, behavioral data, and historic contributions, donations, and engagement with an organization.

Major donors can be defined on your organization’s goals (e.g. those who can reasonably donate $15K+ over a three-year time period) and includes new and existing donors. The model can also retrospectively identify historically underrepresented constituents across gender, ethnicity, geographic location, and age.

The DPM model can be used to inform several aspects of a prospecting and fundraising strategy, including:

Donor Segmentation By grouping donors based on their propensity score, organizations can tailor communication and outreach efforts to maximize the potential value of each donor segment. High-value donors may receive more personalized communication and special recognition, while those from less historically represented groups can also be identified.

Donor Acquisition By estimating the propensity score of potential donors, organizations can prioritize development efforts to target those who are likely to provide the greatest value. This can help optimize marketing spend and increase the efficiency of donor acquisition efforts.

Donor Retention The model can help identify donors who are at risk of lapsing or reducing their giving levels. Organizations can use this information to develop targeted retention strategies, such as personalized outreach or special events, to improve donor retention rates and maximize long-term value.

AWS Redshift, Glue, S3, CodeCommit, SageMaker, EventBridge, QuickSight