2024 Apra Data Science Challenge
Challenge participants learned by doing while competing in Apra’s fifth annual data science competition! The job was to forecast giving trends for a fictional organization and to identify the donor segments with the most growth potential for the upcoming year.
Challenge Context
It’s that time again! Hard to think that after a quick interregnum, your organization is ready to start a new fundraising campaign. Your mission, should you choose to accept it, is to instruct Executive Staff on the goals of that campaign.
Research Questions
Below are some questions you and your team may choose to consider as you work through this challenge. These are only meant to be suggested starting points, and you are encouraged to take your analysis in any direction you see fit:
- What gift bands are predictable and steadily performing? Are any of them performing worse over time?
- What does the donor base look like? What types of donors show steady, predictable patterns?
- Are there donor segments that are not under management where we might want to assign a gift officer?
2024 Data Science Challenge Submitted Posters
The Nature of Giving | Kelly Douglas, Jennifer Tan, California Institute of Technology
Philanthropy is an ecosystem that thrives under the right conditions. This project takes a three-pronged approach to catching gifts in their natural habitat. First it provides context on giving behavior and patterns through a dashboard, walking through environmentally themed insights. Next, it explores the fundraising team and donors and their designations through a pair of relationship maps, as nature is full of connections. It concludes with a cluster analysis based on the butterfly lifecycle to help the organization segment and approach donors appropriately. These efforts help philanthropy to take flight and reach new heights.
Tools: Tableau, Kumu, R
Data Science Challenge Poster
Data Science Challenge Explanation
Dungeons and Donors | Jenna Majzoub, Olive Buchholz, Peter Reiker, Bill Tran, Saint Louis University
In this project, we build a 5-year campaign for this organization using data visualizations, time-series modeling, and a donor pyramid. We found that this organization has highly generous donors that require assignment and interaction and put forth recommendations accordingly.
Tools: R, Python, Tableau
Data Science Challenge Poster
Data Science Challenge Presentation
Campaign Roadmap | Zach Doering, Trevor Price, University of Idaho
Overall Findings
- The organization has opportunities for more blended donations, deferred donations, and multiyear pledges, all which have opportunities for additional gift officer assignments.
- Major gift and direct mail participation are trending upwards, as well as totals in all gift level bands. Direct mail steadily continues to build the pipeline.
- Total giving is primarily from people, but there is opportunity with corporations and foundations, including with gift officer assignments.
Tools: Power BI, PowerPoint, Excel
Data Science Challenge Poster
Data Science Challenge Dataset
Gift Analytics | Edima Udo
Overall Findings
- Continue to leverage Mail and Email as Gift Channels & interaction touch point
- High number of one-time givers
Tools: R and RStudio
Data Science Challenge Poster
Data Science Challenge Explanation