Predictive AI for Nonprofit Fundraising Starts With Clean Data

Nonprofits are facing a critical moment: donor acquisition is slowing, retention is lagging, and fundraising teams are expected to do more with less. At the Fundraising.AI Global Summit, experts argued that one of the most powerful ways to reverse those trends could be using predictive artificial intelligence (AI) for nonprofit fundraising.

“The growth that's happening in predictive AI, in my view, is the primary area of adoption that's going to create the most differentiated results for nonprofits,” Chad Engelgau, CEO of Innovairre, said in the session, “From Data to Decisions: How AI is Powering Smarter Strategies for Fundraising,” last week.

Here’s how Engelgau and his fellow panelists explained how predictive AI can help nonprofits build smarter fundraising strategies. 

What Predictive AI Can Do

Engelgau explained how predictive models can help nonprofits identify donors’ preferred communication channels, improve donor engagement and reduce donor churn, regardless of their size. 

“AI models are way better at predicting outcomes — whether that's who's likely to churn, who's going to respond to this type of package, which channel are they more responsive [to] and what creative messaging are they most engaged in,” he said. 

The payoff is clear. Nonprofits using predictive AI have experienced a 20% to 30% increase in response rates knowing who to engage and when, Engelgau said.

Working with a lapsed donor and getting them to donate again creates greater lifetime value, and you already know that individual has an affinity for your charity versus somebody that's just sitting on a list file,” he said. “... Whether you have 10,000 records in your donor database or you have millions of records in your donor database, this predictive AI can leverage the data that you have in the seed files and integrate these list files into it and produce these results I'm talking about.”

Perhaps most importantly, Engelgau stressed that nonprofits no longer need to lean on data scientists with advanced statistical expertise to benefit from these kinds of tools.

“Just like ChatGPT has made us extremely powerful in terms of content creation categories, especially in the written word, these new tools allow us to do better predictions and move above and beyond the old techniques that we have so that we can embrace multichannel marketing … and consume all that massive amount of data,” he said.

How It’s Done

Many nonprofits struggle with data silos, or data scattered across multiple technologies and databases. AI is a great way to help pull that data together and share key takeaways from the complete picture of your donors, Jill Tennant, senior director of marketing for Microsoft Philanthropies at Microsoft, said. 

She shared how Make-a-Wish America embarked on such an effort to bring together its fragmented data. The nonprofit had donor records scattered across spreadsheets, fundraising databases and volunteer management systems. The fragmentation prevented staff from understanding each supporter’s full relationship with the organization.

By unifying donor profiles through Microsoft Azure, Make-A-Wish America could use predictive analytics to unlock more advanced insights, including smarter segmentation by giving history, previous engagements and preferences.

“It all starts with data,” Tennant said. “It's really about getting your house in order, if you will, from a data perspective, so that you can do some of these more advanced workloads in AI. We like to say that [generative] AI is not an AI strategy. That is certainly a place to begin, but then really you have to start with your data and getting this in order and leveraging data for your advantage when you're thinking about your fundraising.”

Predictive AI Requires Quality Data

Predictive AI, after all, will only work as well as the data it’s built on. The biggest value of experimenting with more AI capabilities, Cherian Koshy, vice president at Kindsight, said, is helping nonprofits break out of their bad habits. 

“Most organizations have enough data — it’s about what to do next with that,” he said. “And really a lot of organizations don’t trust the data, they don’t have time to use it or they don’t know how to interpret it.” 

Nonprofits must find the courage to change their use of technology despite their existing assumptions or workflows. Koshy called this line of thought “having brave data.” 

For organizations worried their databases are too messy to benefit from predictive AI, Koshy reassured them by encouraging them to start with simple, repeatable habits such as deduplication, standardizing entries and filling in missing contact details. He compared it to regularly brushing your teeth instead of waiting for a root canal. 

“If you think about what work you are doing and how important your data is, it really is the same as brushing your teeth or regularly cleaning your house. It's something that you have to do on a very regular basis, and it's something that we don't think about enough as nonprofit organizations.”