More With More, Not More With Less

Being intentional with AI in nonprofit work

By Larry Taylor, Managing Partner at BTH

A personal note before we begin. While feelings about AI are still in flux, more people are now using it as a personal assistant — to help write an email, edit a document, or think through a task in front of them. The next steps can introduce new rounds of uncertainty: putting AI to work on the data your organization actually runs on, or handing it a routine task to free your time. However you feel about AI today, the surest way through that uncertainty is to be clear about the outcomes you want for yourself and your team. I’d start where any human-centered organization should — not with the technology, and not even with the tasks AI might take on, but with the everyday work of your people, and how you want it to change for the better. This post is about how doing right by you and your team is the same thing as expanding what your organization can do for the people it serves.

While this blog posting mentions our work with United Way, the same concepts apply to all of our nonprofit clients.

A development director at a mid-sized United Way starts her day at 8:15. Coffee, calendar, inbox. By 9:30 she’s reformatting a spreadsheet because a board member needs it Friday and the system it came from won’t export the way he wants it. By 11, she’s pulling donor history together from three different places because the gala invitation list has to go out tomorrow. At 1:00, she’s drafting a grant report from data she knows by heart — some of the language is the same as last quarter’s. Somewhere in there, between the assembling and the formatting and the reformatting, she’s also building the major-gift pipeline that will determine the organization’s funding baseline for next year.

All of it matters. The major-gift work is the reason she came into this profession. The rest is what the systems require of her to get there. For most nonprofit professionals, the balance between those two has been off for a long time.

Down the hall, the community impact manager is in the same shape. The CFO is in the same shape. Every one of them is navigating the same gap between what the systems give them and what their role actually requires.

This is the conversation worth having about AI in nonprofit work. Not the technology conversation. The one about what your team’s days really look like, and what your organization can do as a result.

Two kinds of tasks

Jensen Huang, the CEO of Nvidia, said something recently that’s a useful place to start: there’s a difference between the purpose of a job and the tasks of the job. Watching someone type all day doesn’t tell you what their job is — the typing is the friction between the person and the work, and the work is the judgment behind it.

That’s the spark, but it isn’t quite the whole picture, because every job is made of tasks — and plenty of them are worth doing. The more useful cut isn’t purpose versus task. It’s the tasks that drain your energy versus the tasks that give it back. Most people in mission work can tell you, instantly, which of their tasks are which.

The energy-giving tasks are the ones that let them apply their judgment and creativity. The development director on a major-gift call she’s been building toward for two years. The program manager redesigning a curriculum based on what she’s heard from families. The community impact lead sitting with a partner agency to think through what next year’s investment strategy should actually look like. These are the moments where the purpose of the job is visible in the doing. People go home from days like that feeling fulfilled.

The energy-draining tasks are the ones that exist around the purpose, not in service of it. They’re tedious, time-consuming, and rooted in constraints no one chose. The spreadsheet that has to be reformatted because two systems don’t talk. The hours spent figuring out which donors to call this week because the answer isn’t in any one place. The grant deliverables pulled together by hand because the program data lives one place and the financial data lives another. People go home from days like that depleted — and with the quiet sense that the work they were hired for is still waiting.

For many nonprofit professionals, the draining category has grown over time, not because the work changed but because the systems around the work multiplied. Most nonprofit tools were built to serve one department — finance, or development, or programs — and the work has long since stopped respecting those boundaries. The gaps between departments get filled by people: by hand, by exporting, by reformatting, by remembering.

The on-going shift in technology isn’t hype. John Doerr — a venture investor and philanthropist who has been calibrating technology hype cycles for fifty years, and who along with his wife has given more than a billion dollars to climate, education, and global health work — recently called AI “the biggest thing ever,” and “underhyped.” His point isn’t that the technology is magical. It’s that the most important change is in how work itself gets redefined.

What the AI conversation usually gets wrong

The story that’s been told about AI is that it will handle the data assembly, the drafting, the pattern recognition, the routing — and therefore the people who do those things should worry. But that gets the categories backwards. For the development director, the data assembly was never the job; the job was the judgment the assembly was supposed to enable. For the data analyst, pulling the data together from disparate systems was never the job either; the job was the pattern she could see once the assembled data is in front of her — the question she could ask that no one else had thought to ask. The judgment has always been waiting on the other side of the friction.

When AI handles the assembling, the development director starts her day at 8:15 with the gala list already built and the donor histories already pulled. She spends her morning on the calls she came into this work to make. The analyst spends hers on the questions she’s been wanting to ask of the data for months. The job is the same. The day is different.

The flywheel effect

When the friction subsides, the first lift is the obvious one: people get back to their individual work — the work that makes a difference, the work that gives them energy. The development director makes more calls. The community impact manager spends more time with partner agencies. The inbound grants manager closes a stronger renewal. That alone would be a meaningful change.

But there’s a second lift, and it’s the one that matters most for a United Way. When you’re buried in the friction of your own job, the first thing that quietly gets sacrificed is the work that connects you to everyone else. The handoff that takes a moment of thought rather than a forwarded email. The conversation with another department about a partner agency you’re both working with. The note in a record that helps the next person pick up where you left off. The thing a donor said in a stewardship visit that should change how the program team frames the spring appeal. The outcome trend the community impact team is seeing that should change how development tells the story to major donors.

This is the connective tissue of a mission-driven organization — and it is, when you look at it plainly, service turned inward. The same instinct that makes nonprofit people good at serving their communities is what makes them help a colleague, flag a pattern, set up the introduction. They serve each other internally so the organization can serve the community better. It’s the most natural thing in the world for these teams — and it’s the first thing the friction erodes away.

When AI reduces the friction, that connective work gets the room it needs. The development director sets up a site visit for a donor whose interest in early childhood has been growing — and brings the community impact director along so the donor’s questions get answered directly. The campaign manager follows up on something a workplace coordinator mentioned about employee interest in volunteering, and a new corporate partnership starts to take shape. A community impact analyst spends an afternoon with someone on the data side, and they find a pattern that reshapes how next year’s allocations get designed.

None of these are anyone’s job in a given week. All of them are how a United Way grows — more people served, more money raised, deeper partnerships, a clearer story to tell about what actually changed. The prevailing paradigm has been a nonprofit doing more with less — the intentionality here is to do more with more.

That’s the flywheel. My work helps you do yours; your work helps me do mine better. Data and processes start to connect because there are finally people with the time and attention to connect them.

The demand for nonprofit services has never been capped by need — the need has always been larger than any United Way could meet. What was capped was funding and bandwidth — and the complexity of working across systems that weren’t built to talk to each other consumed a real share of what bandwidth there was. Reduce the repetitive tasks, and a team doesn’t just do more of what it was already doing. It starts doing more of the work it was always meant to do together. That’s mission expansion.

The instinct behind our unified platform, One United Way — one organization, not a set of departments running in parallel — has been right for a long time. What’s been missing was never the vision. It was the time, and the connective work, to make it real. It is now technology becoming an enablement layer, rather than the recent past where the technology outright blocked or impeded these desired behaviors.

One step, then the next

AI in nonprofit work unfolds as a series of steps — one process, one team, one pilot — that build on each other. The first is usually one where the value is clear, the risk is contained, and the team can see exactly what changed. The second builds on what the first one taught. By the third or fourth, the team is moving with more confidence, and the kind of work they’re willing to take on starts to look different from what they would have considered at the start. Looking back, the organization is doing things it couldn’t have imagined when it began — not because the technology did anything dramatic, but because people who spent years compensating for the seams between systems finally had room to do the work they came here to do, and to do it together.

The conversation worth having

The AI conversation in nonprofit work is really about what your team’s days look like, and what your organization can do as a result. The tasks create friction; the friction can be eased. The work has always been the purpose. And the instinct behind One United Way has always been the right shape.

But time alone doesn’t connect anything. Free up a development director’s morning and she’ll spend it well — yet the handoff to her colleague, the shared note, the pattern worth passing to another team still need somewhere to land. The connective work needs a place where it actually holds: where what one team learns reaches the team that needs it, where a donor’s interest captured in one system shapes what another team does next. Building that place — and helping your people grow into the work it frees them for — is what we do at Beyond the Horizon. It’s also how we think about AI, which we’ve written more about here.

We’ll be continuing this conversation in person at the United Way West Regional Conference this August, leading a breakout session on exactly this. If your team is starting to ask these questions, come find us there.

Finally, our point of view is that AI is a tool — and when implemented with human-centric intentionality, it will improve the composition of the nonprofit professional’s day to day experience. 100% of BTH’s clients want to expand their missions and their capacities while also improving the composition of their employee’s work. Not more with less, more with more.

That’s what we mean when we talk about AI working for human good. It isn’t about machines doing what people do. It’s about people, finally, being able to do what they came here to do — and to do it together.