Thinking Strategically About AI

A working guide for small nonprofit executive directors who cannot afford another hat.

Your answers save automatically in this browser - nothing is sent anywhere, and no email is required. Close the tab, come back next week, and your work will still be here.

AI strategy is your job. Good news: you already know how to do it.

You set direction for your board. You decide which programs live and which get sunset. You choose where every scarce dollar goes. That is strategic thinking, and you do it every week.

Deciding how your organization uses AI is the same kind of decision. It is not a tech task to hand to the youngest staffer because they know Canva. It is a leadership call about where your people's time goes - and time is the scarcest thing you have.

Strategy asks why. Operations asks how.

Strategic thinking about AI means deciding which problems are worth pointing AI at and why. Operational work is figuring out which buttons to press. Most organizations skip straight to the buttons, try a chatbot for twenty minutes, shrug, and walk away thinking AI is just Google with better manners.

Get the why right first, and the how gets dramatically easier. That is what this guide helps you do.

A reframe worth keeping: adopting AI is not admitting you can't keep up. It is the opposite. "I built systems that don't depend on any one person" is one of the strongest sentences an executive director can say to a board. This guide helps you earn it.

How to use this guide

Seven short sections, each with something to read and something to do. Work through it in one sitting (about 40 minutes) or one tab at a time over a week. Your answers build toward a one-page AI Strategy Snapshot you can print and bring to your board.

Your constraints are real. So are the openings they create.

Small nonprofits face pressures that big organizations write white papers about. Here is what changes when you look at each one through an AI lens.

Funding is tight.The most useful AI tools for a small nonprofit cost $0 to $30 a month. This is one of the few capacity investments that does not require a grant.
Knowledge walks out the door.Every departure takes know-how with it. AI makes documenting processes fast enough to actually happen - and a documented process survives turnover.
No tech expertise on staff.Modern AI works in plain English. The skill it requires is judgment about your own work, and nobody has more of that than you and your team.
Zero spare bandwidth.The fix is choosing AI projects that pay back more time than they cost, starting in the first month. If a project doesn't do that, it's the wrong project.
A board member just asked "are we doing anything with AI?"Gentle pressure is an opening. The ED who brings an answer before being asked twice looks like a strategist, not a laggard.
You've been burned by consultants.Fair. A pilot approach protects you: small, cheap, measurable, and easy to walk away from if it doesn't earn its keep.
Do this now

Quick readiness check

Check every statement that is true for your organization today. Be honest - this is for you, not your board.

Our mission and values are written down and current.
I can name the three tasks that eat the most staff time each month.
At least a few of our key processes are documented somewhere other than someone's head.
We have basic rules about handling donor and client data.
My board has discussed AI at least once, even informally.
At least one staff member is curious about AI rather than afraid of it.
I could find $30 a month for a tool if it clearly saved staff hours.
Someone (maybe me) could own a small experiment for the next 60 days.
Check the boxes above and your readiness read-out will appear here.

You do not need a 40-page strategic plan. You need the right-sized approach.

Strategic planning models fill textbooks. For a small nonprofit thinking about AI, they boil down to four. Pick the one that matches your situation right now.

Start simple (goals-based).Your organization is stable and your mission is clear. Name two or three goals AI should serve - "cut report prep time in half" beats "explore AI" every time.
Fix what hurts (issue-based).One specific problem is bleeding you dry: turnover, reporting overload, a grant backlog. Point AI directly at that issue and nothing else.
Stay loose (organic).AI tools change monthly, so rigid three-year plans age badly. Build shared understanding of your goals, run small experiments, and review what you learned each quarter.
Crisis mode (real-time).If you are in survival mode, set short-term objectives only and check in weekly. AI can still help - drafting, summarizing, triage - but keep the scope tiny.

Whichever you choose, one classic tool earns its place: the SWOT analysis. It takes 15 minutes and turns "we should probably do something with AI" into an actual picture of your situation.

Do this now

Your AI SWOT

Answer in plain sentences. Rough is fine - you can refine later.

Your mission is the filter. Run every AI decision through it.

Mission drift rarely announces itself. It creeps in one shiny tool at a time. Three questions keep you honest before adopting any AI use:

1. Does it serve the people we exist to serve?Directly or by freeing staff time for them. If neither, skip it.
2. Would we be comfortable explaining it to our donors?If describing how you use a tool makes you squirm, that's your answer.
3. Does it reduce work, or just move it around?Some tools shift labor instead of saving it. Count the review time before you count the savings.

The ethical questions deserve real answers

In nonprofit circles, concerns about AI's energy use, water consumption, bias, and privacy come up constantly. They should. These concerns are real, and anyone who waves them away is not being straight with you.

They are also manageable. You can choose lighter-weight tools and batch your use. You can treat every AI draft as a draft from a well-read stranger who doesn't know your community - reviewed by a human who does. You can keep sensitive data out of tools that aren't approved to hold it. Taking the concerns seriously is not a reason to opt out of the conversation. It is the qualification for leading it.

The rule that covers most sins: AI drafts. Humans review. Humans publish. Every time, without exception.

Do this now

Governance starter kit

Check each commitment you are ready to make. These five lines are a working AI policy - most small nonprofits need little more to start safely.

A human reviews every AI-assisted item before it leaves the building.
Donor and client personal data never goes into a tool we haven't approved for it.
One named person is our AI point of contact - questions and experiments route through them.
We keep a simple list of which tools we use and what we use them for.
We revisit these commitments quarterly as tools and rules change.
Your commitments will be counted here.

Don't start with your biggest problem. Start with your loudest groan.

The biggest problem is usually the worst first project - too tangled, too many stakeholders, too easy to stall. The right first pilot is a task that is repetitive, time-hungry, and low-risk. The kind that makes you groan when it lands on your calendar.

What a good first pilot looks like

Small: 3 to 5 hours of total effort. Measurable: you know the hours it takes now, so you can count the hours it takes after. Real: run on your actual work, not a demo. Documented: written down step by step, so it keeps working after the person who built it moves on.

Here is the pattern, using board reports as the example. Today: four hours of gathering numbers, hunting through email for that one figure, and formatting. With AI in the loop: you feed it your notes and data, it assembles a first draft in minutes, and you spend 45 minutes editing and fact-checking. You saved roughly three hours, this month and every month after. Three hours in a month may not seem like a lot, but it sure feels like a lot when that report is due.

Do this now

Pick your pilot

Skip the dashboard. Track three numbers and one feeling.

Impact measurement can swallow whole staff retreats. For your first AI effort, keep it to this:

Hours saved per month.On the pilot task, before versus after. Count your review time honestly.
Processes documented.How many things now exist on paper that used to live only in someone's head.
Staff using AI confidently.How many people use a tool without needing hand-holding.
Your stress level.Not on any official scorecard, but you will know. It counts.

One simple rule: if the pilot does not save more time than it costs, stop. That is not failure - that is data, and it cost you almost nothing to learn. Pick a different task and try again.

Do this now

Time-back calculator

Enter your numbers and your board-ready sentence will appear here.

Don't go it alone

AI changes fast. You do not need to keep up with all of it - you need one trusted newsletter, a couple of peer EDs to swap notes with, and a quarterly hour to review what's working. The peer piece matters most. The best AI decisions I've seen nonprofit leaders make started with a conversation, not a search bar.

Your AI Strategy Snapshot

Here’s everything you entered, gathered onto one page. Print it and bring it to your board, or stick it on the wall where the Post-it notes go to die, but don’t let it die because it’s some great stuff!

Click "Build / refresh my snapshot" and your answers from the earlier tabs will appear here.

Want a guide for the next step?

If you worked through this and thought "I could use someone to walk beside me on this journey" - that is exactly what I do. I'm MJ, and I help small nonprofits use AI in ways that serve people, purpose, and what matters most. No jargon and no pressure.

Let's talk