Logo.png
Ë
By Evan Lyle • June 28, 2017

ML for Non-Giver Ask Amounts

Imagine you're a fundraiser at an event filled with potential donors. You meet someone from out of state who seems interested in your nonprofit. You exchange contact information but aren’t able to talk again in person that night. The next day, you're faced with a predicament: how do you tailor your outreach to this person to maximize the gift opportunity? If you suspect he or she has a large capacity to give, maybe the right approach is to first meet for lunch to establish a more meaningful relationship. But, if this prospect has never made a gift greater than $100 to any nonprofit, a different approach might be a better use of your time. The correct path is not always easy to discern, so how do you choose your plan of action? The wrong decision could leave money on the table or tarnish your organization’s reputation.

This problem boils down to one of estimation. How can you determine a good estimate of somebody’s capacity to donate to your nonprofit? This is especially difficult with individuals who do not have a giving history on which to base your estimate. Now, you're left making what is essentially a random guess. You might have some idea based on the way they acted or dressed, but that is pretty unreliable.

Enter machine learning. Forbes estimated in 2015 that there were 44 trillion terabytes of data worldwide, and that was two years ago. With this unbelievably vast amount of public data in the world, there is enough to characterize a donor based on minimal available information such as their name and address. However, there’s a problem. Though machine learning has existed for about 50 years, it is still difficult to make use of the technology—especially if your expertise is in fundraising and not computer science or statistics.

At Gravyty, we aim to bridge this knowledge gap.

We use numerous data sources and machine learning models to address the problem. For example, we pull data sets about your mystery donor's zip code. Then, we implement machine learning techniques to help put the pieces together for you. Once we have our estimated capacity or ask amount, we convey this information in a way that’s both easy to consume and act upon. Utilizing a service that applies cutting edge technology and analysis to empower fundraisers will make nonprofits smarter and more successful.