The American Red Cross (ARC) relies on fundraising and planned giving – when a donor leaves a bequest to ARC in their will, for instance – to fulfill their mission of alleviating human suffering in the face of emergencies.
BKV executes direct mail programs that harvest planned gifts by targeting current ARC donors. Historically, BKV and ARC have employed basic age and giving history criteria to determine who to target.
ARC was interested in using advanced data analysis and predictive modeling to move beyond these basic criteria and identify other factors that result in planned gifts, provided BKV could demonstrate an increase in direct mail program performance resulting from this analysis.
BKV compiled ARC’s internal donor history data – looking at donors’ date of birth, giving history, donation amounts, etc. – as well as external data in order to understand donor demographics, income levels, etc. We then built statistical models, using logistic regression to identify the factors that increase the likelihood of both a donor responding to a direct mail piece and actually making a planned gift.
Once the statistical models were built, BKV created donor scores that reflect the likelihood of a donor responding to our mailing and/or making a planned gift to ARC. We then executed a direct mail test using cells that employed our traditional mailing criteria and cells that used the predictive model results.
Those test cells where we applied predictive modeling outperformed the control cells, which were made up of traditional criteria (age, giving history, etc.), resulting in a 32% increase in response rate. Also, BKV generated lists of “high likelihood donors” (those with high donor scores) that the ARC’s gift planning officers are now using as they communicate and prospect. As a result, ARC has elected to rely more heavily on advanced analytics and predictive modeling to help their direct mail program become even more efficient and effective.