What This Episode Answers
Most nonprofit leaders who feel like AI isn’t working for them are not using it badly, and they are not behind. The reason the output comes back generic is that the organization’s own knowledge — how decisions get made, who it serves and why, what a good grant report actually sounds like — has never been written down anywhere it could be handed over. Brooke Richie-Babbage explains why this is an onboarding problem rather than a technology problem, and why the friction a leader feels when re-explaining context is the clearest available map of where her organization’s design deficit sits.
Key Questions Answered on AI Micro Series
- Why does AI give my nonprofit generic, off-brand output?
- Am I behind on AI if it isn’t working for me yet?
- Is better prompting the thing that separates leaders who get value from AI from leaders who don’t?
- What context does an AI tool actually need to work for a specific organization?
- Why isn’t my organization’s knowledge written down anywhere?
- What is a design deficit, and how does it show up in AI use?
- How do I start documenting what my organization knows without building an operations manual?
About This Episode
Two nonprofit leaders open the same tool on the same day. One of them hands it a pile of notes and gets back an SOP she can use, and a task that used to take two hours takes twenty minutes. The other corrects the draft, re-explains the context, tries another prompt, and quietly concludes the whole thing is more trouble than it’s worth. The difference between them is almost never the prompt. In this episode, Brooke Richie-Babbage makes the case that the second leader is sitting on an onboarding gap, not a skills gap — and that the gap predates AI by years. She walks through the brilliant-new-hire frame, explains why the knowledge that runs most organizations lives in people rather than in structure, and leaves listeners with one exercise to do this week. Part one of a two-part micro-series.
What You’ll Learn
- Why generic AI output is an onboarding failure, not a capability failure — and what a brilliant new hire with no onboarding and a one-line instruction would produce in exactly the same situation.
- Where your organization actually keeps its knowledge — the layer underneath the SOPs that governs judgment calls, funder framing, and what “good” looks like, and why it is the last thing to get documented.
- A one-week exercise that turns the friction into a document — pick one recurring piece of work, and brain dump what you’d have to tell someone brilliant for them to do it the way you want.
Key Takeaways
- The difference between leaders who get value from AI and leaders who don’t is rarely the prompt.
Brooke Richie-Babbage argues it comes down to how much of the organization’s context was already written down before the leader started. That is a structural head start, and it can be built. - Generic AI output is a predictable result of no onboarding.
An extremely capable new hire given a task, one line of instruction, and no background will produce something competent that doesn’t sound like your organization and misses the judgment calls you’d make without thinking. Nobody would call that a bad hire. - The friction is diagnostic information, not evidence of failure.
Every time a leader finds herself re-explaining something, she has located a piece of her organization’s know-how that lives in a person rather than in a system. - This is a design deficit arriving through a new door.
A design deficit is the gap between what an organization is being asked to carry and what its structure was actually built to absorb. Organizations that grow faster than their structures end up running on knowledge that was never institutionalized, and AI makes that visible faster than anything else. - Writing it down pays a bigger dividend than better AI output.
The same document that lets a tool act on the organization’s behalf is the document that stops the organization from depending on any one person’s memory.
Frequently Asked Questions
- Why does AI give my nonprofit generic answers?
Because it has no context about the organization. Brooke Richie-Babbage compares it to handing a brilliant new hire a task with one line of instruction and no onboarding. The output will be competent and generic, because nothing about the mission, the community served, or how decisions get made has been supplied. - Am I behind if AI isn’t working for me yet?
No. Brooke Richie-Babbage is direct about this: leaders getting strong results are not further along technically, they simply have more of their organization’s context available to hand over. That is a documentation gap rather than a skills gap, and it is closable in a week of focused work on a single recurring task. - What is a design deficit in a nonprofit organization?
A design deficit is the gap between what an organization is being asked to carry and what its structure was actually built to absorb. Brooke Richie-Babbage uses the term for nonprofits that grew faster than the systems, roles, and structures underneath them, leaving the organization held together by the people rather than by its design. - Where does organizational knowledge usually live if it isn’t documented?
In people’s heads and in scattered Google Docs. How decisions get made, what good looks like in different contexts, how the organization sounds to donors, how the founder’s history gets used — this layer governs most daily judgment calls and is typically the last thing any organization writes down. - What should I document first to make AI useful for my organization?
One recurring piece of work, not the whole operations manual. Brooke Richie-Babbage suggests picking something like a grant report or a donor follow-up and listing everything you’d have to tell a brilliant new hire: who it’s for, what matters most, what judgment calls exist, and what excellent versus good enough looks like. - Do I have to turn all of this into a formal onboarding binder?
No. One of the more useful things about these tools is that the brain dump can go directly into the AI, which will then help structure its own onboarding. The output of that session is documentation the organization did not have before, produced without a formal documentation project. - How does documenting context reduce dependence on one person?
Once the undocumented layer exists outside somebody’s head, the organization stops needing that person present for the work to happen correctly. Brooke Richie-Babbage frames this as the real payoff: better AI output is the small prize, and relieving the bottleneck is the larger one.
Step-by-step: the brilliant new hire exercise
- Pick one recurring piece of work. A grant report, a donor follow-up, a partnership outreach email. One thing, not the operations manual.
- Imagine handing it to a brilliant new hire on day one. Someone fast, capable, who remembers everything you tell them and can execute exactly the way you want.
- Write down everything you’d have to tell them. Who is this deliverable for? What matters most? What context do they need about the community, the funder, the history?
- Name the judgment calls. What decisions exist inside this task, and how do you make them? How should different factors get weighed against each other?
- Define the standard. What does excellent look like here? What does good enough look like? Those are usually different, and the difference is rarely written anywhere.
- Put the brain dump straight into the AI tool. You do not have to format it into an onboarding document first. Hand over the raw list and let the tool help structure what it still needs.
Resources Mentioned
- Claude — the AI assistant Brooke uses herself.
- ChatGPT — named as one of the tools leaders are using.
- Gemini — named as one of the tools leaders are using.
Related Episodes
- Leveraging AI Micro Series Part 2: Stop Using AI to Save Time. Use It to Build Capacity
-
Leveraging AI Micro Series Part 3: The One Document That Makes AI Actually Know Your Organization
Full Episode Content (click to expand)
INTRO: Two very different experiences with the same tool
Everyone says that AI is supposed to give you hours back, and for a lot of nonprofit leaders, it does just that. In my own life and business, it has given me just so many days back, it’s been transformative. I am watching leaders hand ChatGPT or Claude a first draft of something. A set of notes, talking into Claude and saying, turn these into an SOP, right? A recurring task, something that used to take 2 hours now takes 20 minutes. One leader told me it’s like having an assistant she can hand something to who’s fully competent to do exactly what she wants. And I am also talking to a lot of nonprofit leaders, of people who are having a very different experience.
They are correcting drafts, they are re-explaining context, they’re trying another prompt, they are wondering if maybe they’re doing AI wrong and whether or not they are, it’s just not that useful. It’s more trouble than it’s worth.
So this episode is for that second group, and here’s what I want to say: you are not doing AI wrong. And you’re not as far behind as you think. When I look at the difference between those 2 groups, those 2 experiences with AI, it usually isn’t about who found the perfect prompt. It’s almost never about who found the perfect prompt. It’s about one thing the first group has that the second group doesn’t have yet, and it’s something you can build.
PART 1: The brilliant new hire
Here’s the way I’ve started thinking about AI. Imagine you hired the smartest, fastest person you’ve ever worked with, with deep expertise in a particular area. They write well, they learn instantly, every piece of context you give them they remember forever, they take on anything you hand them, and it’s their first day.
Now, what happens if you give that person a task with a one-line instruction and no onboarding?
They’ll do their best. They’re smart, they’re fast, they learn, but you’ve given them nothing. You’ve given them no context, you’ve given them no background. They’ll do their best and it’ll be fine, but it’s not going to sound like your organization. It’s not going to draw on your mission or your strategy or your context. It’s going to miss a lot of nuances about who you serve and why, who’s in your ecosystem, in your community. It’ll probably, if you ask it to help you make a decision, make calls you would never make.
Is that the new hire’s fault? No, that’s an onboarding problem, right? You would never onboard somebody that way.
That’s exactly what’s happening when AI hands you something generic or off-brand, or that just doesn’t sound or think right. It’s not that the tool isn’t capable, it’s that it hasn’t been onboarded into your organization.
PART 2: Why the onboarding doesn’t exist
So why has that happened? Why is this sort of gap there? Because in most organizations, onboarding doesn’t exist anywhere to hand over, not in that way, and definitely not for AI, not for new staff either. How your organization actually works, who you serve, how decisions get made, what good looks like in different contexts, your brand voice and how you sound often lives in various Google documents or in people’s heads. You know what the emails are supposed to sound like when they go out to your donors or supporters. You know what a good grant proposal sounds like. You know the way that you leverage the history of the founder or the community partnerships, the people on your team know that because they work with that information all the time. It’s not institutionalized anywhere, and that is really common with organizations that grow faster than the structures they were built on.
So I talk all the time about design deficits. This is a design deficit showing up through a new door.
Right, and this is the part I actually find really exciting: AI is showing the gaps more clearly than anything I’ve seen in a really long time. Every time you find yourself re-explaining something to Claude, that’s a piece of your organization’s know-how that’s ready to be captured. The friction you’re feeling is not a sign that AI won’t work for you. It’s a map of exactly what to build so it can.
PART 3: What would you tell your brilliant new hire?
So here’s how to think about this. What would you tell your brilliant new hire? This is the question that I would sit with this week. If you hired this brilliant new person tomorrow and handed them one recurring piece of your work, what would you need to tell them so they could do it the way you want them to?
Pick one thing, not your whole operations manual. The grant report or a donor follow-up, a partnership outreach, and then think about how you would onboard that person and notice everything that comes up.
What would you explain? What context would you give, right? Who is this deliverable for? What matters the most? What judgment calls are there to be made? How would you make those judgment calls? Now remember, this brilliant new hire can execute exactly the way you want them to. So what would you tell them to do? What does excellent look like? What does good enough look like? How do you want them to weigh different factors? Everything on the list that you brainstorm is valuable context. Big things, small things, that’s your organization’s know-how, and it’s the version of the know-how that often lives in your head. It is the version of your organization’s interstitial tissue that tends to undergird our design deficits, because these are the kinds of things that are the last things to be documented. Right? They’re not turning into SOPs. They’re just sort of the what people know, right, in the organization.
Once it’s written down, it doesn’t just make AI better, it makes your whole organization less dependent on any one person. It’s like the magical elixir that solves the bottleneck problem.
And the good news is you don’t actually have to take this brain dump for this brilliant new person, all of the things that you would use to onboard them, you don’t actually have to turn them into an onboarding binder. One of the best things that AI allows you to do is brain dump directly into AI, and it will tell you how to onboard itself.
CLOSE: What’s coming in part two
So the good news is, and I’m going to leave this with you because I’m coming back for the second part of this micro series next week. Here’s what I’m going to leave you with: the same tool, whether it’s Claude or ChatGPT or Gemini — I’m a Claude girl myself — the same tool that’s showing you the gaps, the design deficits in your organization, can help you close them faster than ever possible before. So next week, I’m going to talk about exactly how. What I think your first AI move should actually be, or if you’ve been experimenting, what your next AI move should be, and it’s one that builds capacity, not just speed.
That’s it for this week. I’ll see you back next week for more Mastermind.
Work With Brooke
Want to work together?
Apply for the Next Level Nonprofit Mastermind, a high-touch coaching and training accelerator for established organizations with $1M+ budgets that are ready to design for impact sustained at scale.
Budget under $1M? Start with the Fit Check. It routes you to the right structure for where your organization actually is: a System from the Systems Suite if you need one specific piece of operating infrastructure built, or an invitation to The Stability Flywheel Lab if you’re at the $750K–$999K threshold.
Diagnostic
If you’ve found yourself re-explaining your organization to an AI tool and getting generic work back, you’ve already located one of your design gaps. Take the free Design Deficit Diagnostic to find out where your organization’s biggest structural gap is — and what to fix first.
Connect with Brooke
LinkedIn: https://www.linkedin.com/in/bjrichiebabbage/
Instagram: https://www.instagram.com/bjrichiebabbage/
YouTube: https://www.youtube.com/@brookerichie-babbage
About the Host

Brooke Richie-Babbage is a nonprofit strategist and executive coach with 25+ years of sector experience. She is the founder of Bending Arc, a strategy and coaching firm that helps nonprofit CEOs transform operationally fragile organizations into resilient institutions designed to sustain impact at scale. Her work is built on a single conviction: resilience isn’t about grit — it’s about good design.
