Predictive Modeling

While predictive modeling for fundraising is growing in popularity and is becoming more widely used, its definition and applications still are not universally familiar for fundraising professionals.

Predictive modeling for fundraising typically involves the use of statistical techniques to determine which constituents are most likely to exhibit a particular behavior. In many cases, this behavior is related to giving. For example, many nonprofits use predictive modeling to identify the prospects most likely to

  • make a major gift

  • make a planned gift

  • respond to a particular type of solicitation

  • make a gift via a particular channel (by mail; online; over the telephone; etc.)

  • give to a particular project or initiative

Organizations use these types of predictive models to

  • identify new prospects

  • prioritize prospects

  • segment prospect groups to increase solicitation response rates (and save money on broad-based appeals)

  • help fundraisers optimize their portfolios to focus on the best prospects

  • determine which prospects to screen for wealth and capacity