This repository contains the files for my APRA PD DAS 2019 talk titled Generating Insights from Clustering Large Donors.
This is the description of the talk:
Stakeholders frequently ask questions like “what other donors look like this group?” or “what is the pathway to elevated giving?” Clustering techniques within the class of unsupervised learning methods let analysts generate insights into these questions for their most generous donors, even without a dependent variable they are trying to predict.
In this talk, the presenter will discuss a variety of clustering techniques, how and when the techniques can be applied, and how the results from clustering algorithms can be used. Segmenting an organization’s most generous donors into distinct groups can allow fundraisers, researchers, and analysts approach these important constituents in more targeted and personalized ways.
Additionally, the presenter will comment on some of the larger trends in fundraising. These trends include the increasing concentration of giving, the continued deviation from the 80/20 Pareto principle, and the potential disruption in philanthropy caused by the ultra-wealthy.