Predictive Analytics for Major Giving
Predictive modeling is considered the best bet for a high return on investment in major giving analytics. It is a process by which a nonprofit can integrate transactional giving data with large amounts of donor profile information and then predict with accuracy what actions donors are most likely to take in the future. Past studies have shown that predictive modeling, when compared to simple wealth screening, is more accurate at identifying future major giving opportunities.
Looking at the combination of likelihood and capacity is critical to predictive modeling, as the wealthiest prospect may have little interest in your mission, and the strongest advocates may not have the capacity to provide large gifts. This combination of models highlights individuals who have strong past giving patterns – including longer consecutive giving histories and increasing gift sizes year over year – but the giving component is only part of the equation, highlighting one of the strengths of a modeled approach. The best prospects also have other common traits that correlate with major giving:
- Larger amounts of publicly identifiable assets
- Strong credit history
- High level of equity in their homes
While these factors might seem obvious, the strongest predictive models use rigorous statistical techniques to identify the less obvious factors that might be indicative of predicting likelihood to give to your organization. Since no two organizations are the same, no two predictive models should be either.
A predictive model can statistically determine how each aspect of a donor’s profile and giving history contributes to likelihood and capacity and then weights those aspects accordingly when ranking and scoring the donors. This might not work for all organizations in all circumstances, rather, its purpose is to determine which donor attributes predict a specific giving behavior at your organization.
For more information on this topic, please visit: Demographics Indicators