Of 100 first-time donors, only 20 will give again the following year. That’s according to the Fundraising Effectiveness Project, and it’s a stat that should make every nonprofit leader wince.
But is that a retention problem? Or a relationship problem?
Development staff already know what they should be doing. They should be staying in closer touch with donors, sharing better stories, and personalizing outreach. They should be mining their CRM for patterns and insights. They should be building relationships instead of just processing transactions. They know all of this. They just don’t have the time.
When I talk to EDs and development staff about why donors aren’t sticking around, the reasons are pretty consistent. Nonprofits struggle to tell compelling stories about their work. They send generic thank-you letters instead of personalized acknowledgments. They ask for money without first connecting with donors to understand what matters to them. And increasingly, donors care about that connection. They want to know where their money is going, what it’s supporting, and what difference it’s making. In today’s economy, people are selective. There are a lot of organizations asking for support, and donors are looking for reasons to feel invested, not just financially but emotionally.
The problem is that most development teams are stretched thin. They’re writing grants, planning events, managing volunteers, seeking donations, and trying to keep up with communications. Donor data is sitting right there in the CRM, but there’s little to no bandwidth to do anything strategic with it.
That’s where AI comes in. It’s not a replacement for human connection, but a tool that frees up time for it.
A Framework for Keeping the Humanity
When I coach organizations on incorporating AI without losing their authenticity, I use a framework called the 10-80-10 rule, developed by Dmitry Koltunov of Arbor, a storytelling platform for nonprofits.
The first 10% is human setup. This is where you set the tone, define your goals, and give AI the context it needs to sound like your organization. The middle 80% is where AI does the heavy lifting: pulling and analyzing data, drafting copy, generating variations, providing light personalization. The final 10% is human review and refinement. This is where you check for emotional resonance, add real-world specifics (a staff anecdote, a “because of you” story that couldn’t possibly be AI-generated), and make sure it sounds like you.
That final 10% is what turns an efficient campaign into a memorable one.
What This Looks Like in Practice
Here’s a simple example. Your organization just raised $25,000 in 24 hours after a pipe burst flooded your community center, and now you need to thank 50 donors.
You start by clarifying your segments. Who are these 50 donors? First-time givers? Longtime supporters? Major donors? You define your emotional goal (gratitude, impact, trust) and provide AI with some context: your organization’s voice, a few sample thank-you notes that feel authentic, details about the emergency and what the funds supported.
AI generates initial drafts tailored to each segment. One version for first-time donors: “You joined our community at a critical moment.” Another for longtime supporters: “Your ongoing commitment meant we could act immediately.” AI lightly personalizes, using mail-merge variables like first name, gift amount, and date. It creates tone variations: formal for institutional donors, conversational for individuals.
Then you review for emotional accuracy. Does this sound like us? You add personal touches, specific details about impact, maybe a P.S. that feels spontaneous. Every donor still feels the humanity. AI just saved you 90% of the time it would have taken to handcraft 50 notes.
AI Can Spot Patterns You’d Miss
This is where AI becomes a strategic partner, by identifying donor segments and patterns that humans would miss.
AI might detect that a percentage of your donors only give within 72 hours of local news coverage about your issue area. Humans might chalk that up to coincidence, but AI can cross-reference timestamps with local media coverage and find a clear correlation. These are reactive donors, driven by current events rather than campaigns. Now you can be prepared with thank-yous and follow-up stories ready to send within 48 hours of relevant news.
Or AI might flag donors who used to give three times a year but have recently dropped to once. These are donors who are still emotionally invested but may be financially stretched. Instead of pushing for another monetary gift, you could invite them to a volunteer shift or a behind-the-scenes coffee chat. The goal shifts from gift frequency to relationship preservation.
AI can also identify donors who give modestly but have the highest email open rates and share your social posts most often. Humans might overlook them because they’re low-dollar givers. AI can recognize them as donors who extend your reach, not just your revenue.
This is the kind of strategic intelligence that helps nonprofits work smarter.
What AI Can’t Do
AI cannot replace genuine human gratitude. And it can’t fake authenticity.
When I was a program officer, I used to send handwritten thank-you notes to grantees after site visits. These weren’t donors, but they were organizations who had given up their time to host foundation staff and board members. A handwritten note felt more personal than a mass thank-you email. I tried to include something specific about the visit, a detail that showed I’d been paying attention.
That wasn’t scalable. It wasn’t efficient. But it mattered.
Donors leave because they don’t feel seen. They’re not receiving personalized touchpoints (or maybe they’re not receiving any touchpoints at all). They don’t feel informed about what their donations are supporting. They don’t feel connected to the mission in the way they want to. I wrote recently about what happens when you walk your own donor journey and how quickly these gaps become painfully obvious.
AI can help you stay organized and strategic. It can help you identify who needs attention and when. But it can’t replace the phone call that rebuilds trust. It can’t manufacture the human moment that turns a one-time giver into a lifelong supporter.
The best use of AI is to help save time so you can create more space for those moments.
Protecting Donor Trust
Supporters share personal details (family ties, interests, giving history) with an expectation of confidentiality. When you start using AI to analyze and engage with that data, you’re taking on a new level of stewardship responsibility.
Not all AI tools handle donor data the same way. If you’re copying and pasting spreadsheets into public AI platforms, you’re taking some risk. Look for AI that’s built into your CRM or that offers enterprise-level data protections. Before you start using any tool, ask: How is our data stored? Who has access? What happens if something goes wrong? If your vendor can’t answer those questions clearly, that’s a red flag.
You should also think about internal policies. Who on your staff has access to donor data? What tools are they allowed to use? What are your procedures for disclosing AI use to donors when appropriate?
Transparency matters most when AI touches anything that affects voice, visuals, or values. If you’re using AI to create or alter donor-facing content like videos, photos, or narratives, disclose it. You don’t need to apologize for using AI. You just need to be honest about it.
The nonprofits that earn donor loyalty in the next five years will be the ones who say, “We use technology ethically, openly, and always in service of human connection.”
Where to Start
AI won’t build relationships for you. But it can help you stop doing busywork and start doing relationship work. It can help you see patterns in your data that would otherwise stay hidden. It can help you draft, organize, and communicate more effectively so you have more bandwidth to focus on what matters.
If you’re wondering where to start, pick one operational headache that’s eating up staff time every week. Maybe it’s thank-you letters. Maybe it’s pulling and analyzing donor reports. Maybe it’s segmenting your email list.
Start there. Use AI to take that task off your plate. Then take the time you just freed up and spend it on a donor conversation or a handwritten note. Because those are the things that keep donors around.
Image by June Laves from Pixabay