Note: This guest post is from Chris Riddell of Thrive Technology. Thanks, Chris!
In a time of deep uncertainty and instability, fundraising remains rooted in one constant: relationships.
You’ve built strong connections with your supporters. The data from those relationships is one of your greatest assets. By tapping into it, you can find clarity amid the chaos by focusing your efforts on the people most likely to engage, give, and grow with you.
This is exactly what machine learning does.
You’re already using it
You’ve already encountered machine learning, a form of artificial intelligence (AI), in your everyday life:
Netflix recommends shows based on what you’ve watched
- Your email filters out spam
- Your phone suggests words as you type
At its core, machine learning helps computers find patterns in data and make predictions without being explicitly programmed for every possible scenario. These systems “learn” from past data to make better guesses in the future.
Your fundraising opportunity
In fundraising, machine learning can uncover patterns that aren’t obvious to you: patterns that can inform decisions about who to engage, how to communicate, and where to prioritize outreach.
Most fundraising strategies rely on looking backward. You might segment donors by last gift amount, recency of giving, or event participation. While helpful, these approaches often miss more subtle signals – and they treat every donor in a segment the same way.
Machine learning helps you move beyond segmentation. Instead of grouping donors by broad categories, it allows you to predict the likelihood individuals taking a certain action. That might mean identifying which one-time donors are most likely to give again, which supporters might upgrade to a monthly gift, or which lapsed donors could be re-engaged with the right outreach.
Your team can focus on matters most – and avoid wasting resources on broad, low-impact strategies.
Five ways nonprofits are using machine learning to raise more money
1. Donor retention. Spot which donors are at risk of lapsing before they leave – and prioritize personalized outreach to retain them.
2. Major gifts prospecting. Identify hidden major gift prospects who might not show obvious signals yet, but share characteristics with your most generous supporters.
3. Campaign personalization. Tailor your appeals based on giving behavior, interests, and engagement history to better resonate with each audience.
4. New donor acquisition. Use models to score and prioritize acquisition lists, incorporating external data and factors, helping you focus on prospects who are most likely to respond.
5. Email optimization. Predict which supporters are most likely to open, click, or respond to a particular type of message, improving campaign performance without sending more email.
Frequently asked questions
“Isn’t this just for big organizations with data science teams?”
No! Tools and platforms have made these models increasingly accessible, even for small and mid-sized nonprofits. Many CRMs and email platforms now offer built-in predictive features, and consultants can help tailor models to your needs.
“Do we even have enough data?”
If you’re tracking donations, events, email activity, or volunteer participation, you likely have more data than you realize. Even a few years of donor history can provide enough info to start meaningful analysis. Typically, you’ll want at least 350-500 donors and 7,000-10,000 total supporters (donors and non-donors combined) to begin.
“Will this replace my team’s judgment?”
Absolutely not. Machine learning offers recommendations, not decisions. Your expertise – knowing your mission, supporters, and messaging – remains at the heart of your work. These tools simply help you spend your time and energy where they’re most likely to pay off.
Where do we start?
To see the benefits, you don’t need to jump straight into complex modeling. Start by asking focused questions like:
- Who is most likely to become a monthly donor?
- Which lapsed donors are worth re-engaging?
- How can we improve retention for first-time donors?
From there, explore the data you already have. Talk to your data leaders, analytics team, or technology partners about what’s possible. Consider starting with a small pilot project focused on one use case. Your goal is to learn, measure, and expand thoughtfully.
Another tool in your toolbox
Machine learning isn’t about replacing the heart of fundraising: relationships, passion, storytelling. It’s about giving your team smarter tools to focus on the right conversations, with the right people, at the right time.
Whether you’re a large organization or a small nonprofit, these approaches are more accessible than ever – and they can make a real difference in helping you achieve your mission.
Curious about how machine learning could support your fundraising goals? I’d love to explore what’s possible together. Reach out or leave a comment – let’s start the conversation.
Chris, I would absolutely love to learn more about this! I have been experimenting with Ai tools more and more in my consulting practice and intentionally “pulling the curtain back” so clients can see how I am using these resources to save time.
Alyson, I’d love to talk more about this! And, I’d love to learn about how you are using AI in your practice. Let’s connect!
https://wethrivetech.com/appointments
Also, I’m hosting a free webinar next Wednesday and repeated on Thursday, where we will explore machine learning as it relates the specific questions you and your clients have about how to grow your fundraising with machine learning.
http://bit.ly/4iKSYKQ
I’m unclear as to how an organization with 2-3 administrative staff, a few volunteers, and a database with a few hundred people in it might adopt this kind of strategy. Or are we simply too small?
Hi Pat — Yes, your organization may be too small to take advantage of this. However, I defer to Chris…