Decision guide for boards and leadership teams · updated
10 questions every nonprofit board should be able to answer — and most cannot
By Norton Lam · AI consultant · Twin Cities, MN
Ranked by how few boards can answer them cold. Each one comes with why it is hard to answer from inside the organization, and what changes once you can. Best used by sending them to your board members and your directors separately and comparing what comes back: the disagreements are more useful than the answers. Four of them you can settle yourselves this quarter without spending anything.
TL;DR: you do not need a consultant to tell you the organization is stretched. You need to know which stretched thing costs the mission the most. These ten questions are the ones whose answers live in the gaps between functions, which is why nobody inside a single function can answer them and why they sit unanswered for years. If your board and your staff cannot agree on priorities, the disagreement is almost always one of these ten going unasked.
These are not trick questions and they are not a governance maturity quiz. They are the ten I have watched capable nonprofit leaders stall on, and they stall for the same structural reason every time: the answer requires comparing one function against another, and everyone inside owns exactly one function. That is not a failing of the team. It is the shape of the org chart, and a volunteer board sees it only in summary. Four of these you can settle with attention rather than money, and I have marked which. The fastest way to use this page is not to answer them alone: send them to each board member and each director separately, collect the answers in writing, then read them side by side. Where two people answer the same question differently, you have found the agenda for your next board retreat. If you are a small staff with no leadership team to poll, the signs version is the better read.
Question 1 · The one nobody can answer
What does each program actually cost to deliver, all in?
Not what the grant covers. What it costs you to run, including the share of the executive director, the bookkeeper, the rent, the insurance, and the week of reporting at the end of every quarter.
Why it is hard from inside: the answer is spread across finance, programs, and people, and no single person owns more than one piece. Fund accounting is built to satisfy funders, not to tell you the fully loaded cost of delivering a service to one participant.
What changes when you can answer it: you stop defending the overhead ratio and start having a real conversation about which programs the organization can afford to keep at their current size. Almost every hard nonprofit decision gets easier once this number exists.
Question 2
What are the three things to fix first, in order?
Most leadership teams can produce fifteen things worth fixing in ten minutes. Almost none can produce three in a defensible order, with the reason each one outranks the next.
Why it is hard from inside: ordering requires comparing a development problem against a staffing problem against a systems problem, and the only people qualified to rank each one are the people who own it. Everyone ranks their own first, honestly.
What changes when you can answer it: the board stops relitigating priorities every meeting. An order with visible reasoning survives contact with the director who lost the argument, which is the only test that matters.
Question 3
If a program and the back office need the same dollar, which wins and why?
This is the ranking question again, sharper, because it is the version you face in a budget meeting. One unrestricted dollar, two legitimate claims: more service now, or the infrastructure that makes service cheaper next year.
Why it is hard from inside: unrestricted money is scarce and the mission argument always sounds better in the room. So infrastructure loses every time, quietly, until the year it fails and takes a program with it.
What changes when you can answer it: you get a rule instead of a wrestling match, and the person who lost can read why. That is the difference between a decision people accept and a decision people work around.
Question 4
What are we paying for every month that nobody uses?
Ask for one list of every recurring charge with the name of the person who owns it. You can do this yourselves this week.
Why it is hard from inside: nothing is hidden and no single person sees it all. Cards are spread across staff, renewals are automatic by design, and cancelling requires knowing that a second system already does the job.
What changes when you can answer it: it is usually the fastest money available anywhere in the organization, and it tells you something larger. How this list looks is how decisions get made when nobody is watching.
Question 5
What happens in the year our largest funder does not renew?
Name the largest single source of revenue as a percentage, and say what the organization does in the year it does not renew. Not whether it will. What you do if it does not.
Why it is hard from inside: everyone knows the number and nobody owns the scenario. Development is measured on raising this year’s money, not on reducing next year’s dependency, so the work that would fix it never competes for anyone’s time.
What changes when you can answer it: diversification stops being a slogan in the strategic plan and becomes a ranked project with a cost, an owner, and a date. Boards are good at this question once someone puts it in front of them.
Question 6
Which programs subsidize which, and did we choose that?
Every multi-program nonprofit has programs that cover their costs and programs that do not. The cross-subsidy is a legitimate choice. The question is whether it was one.
Why it is hard from inside: it requires question one answered for every program at once, and the flows run through shared staff and unrestricted revenue that no report separates out.
What changes when you can answer it: you can defend the subsidy on purpose, to the board and to funders, or you can resize it. What you cannot do is keep funding it by accident and calling the result a budget.
Question 7
Who is the single point of failure?
Name the person whose two-week absence would actually hurt. Not inconvenience. Hurt. Usually it is the ED, the one person who writes the grants, or whoever knows how the restricted funds are really tracked.
Why it is hard from inside: that person is excellent, so the dependency reads as commitment rather than risk. Nonprofits are especially prone to calling it culture, because the person in question is usually proud of it.
What changes when you can answer it: writing down what one person knows is often the highest-return project in the organization, and it is the prerequisite for a leadership transition that does not become a crisis.
Question 8
What is the one compliance or contract risk that would genuinely hurt us?
Not a list of twenty risks. The one. Usually it is a grant agreement clause nobody has reread since signing, a lapsed filing, a classification question about contractors, or a data practice that would not survive a donor asking about it.
Why it is hard from inside: risk lives in the documents nobody reads after signing, and the person who would spot it is the person you do not employ. Volunteer board expertise helps, and it is episodic by nature.
What changes when you can answer it: most real risks are cheap to fix and expensive to meet. Finding it is the work. Fixing it is often a phone call, a policy, and a line in the minutes.
Question 9
What did the board decide last year and never finish?
Every board has a decided-but-never-done list. It is the honest measure of execution capacity, and it is usually longer than anyone wants to read into the minutes.
Why it is hard from inside: nobody keeps the list. Each stalled item has its own reasonable explanation, usually capacity, so the pattern is invisible unless someone lines them up across a year of minutes.
What changes when you can answer it: you find out whether the real problem is deciding or doing. If it is doing, another plan is the last thing the organization needs, and that is worth learning before you buy one.
Question 10
What does our own data already know that nobody has read?
Your systems have been recording for years. Donor history, lapse timing, program outcomes, volunteer hours, event returns, three filed 990s. Almost none of it gets read across systems.
Why it is hard from inside: the answers live in the gaps between tools, and joining them up is a project nobody is assigned. Each system reports on itself competently and says nothing about the others.
What changes when you can answer it: the findings come from your own numbers instead of sector benchmarks, which means your board cannot dismiss them as generic. This is the part AI has genuinely changed: reading everything at once got cheap.
Three honest notes
Before you go lookingThree notes
Not knowing is normal, and it is not a governance failure
No board running a real organization has all ten of these on hand. Staff are close enough to run the thing, which is exactly what makes the whole-organization view hard to hold, and a board meets for a few hours a quarter. The signal is not how many you can answer. It is how long it would take you to find out.
Four of these you can answer yourselves this quarter
The recurring-charges list, the single point of failure, the decided-but-never-done list, and the funder concentration question all need attention rather than money. Start there. If those four produce the picture you needed, you are done and you have spent nothing.
An answer nobody acts on is worse than not asking
Once these are in the minutes, they are on the record, and a known risk nobody addressed is a harder thing to explain later than a risk nobody had named. Before you go looking, decide who reads the answers, in what month, and what they are authorized to change without another vote.
Frequently asked questions
Common questions6 answers
We have a strategic plan. Why is that not enough?
A strategic plan sets direction. It almost never ranks the operational work underneath it, which is where the money and the staff hours actually go. Most organizations can name their five-year vision and cannot name which three things to fix first this year, in order, with a reason.
Our board says they can answer all ten.
Then ask for the answers in writing, separately, before anyone sits down together. Ten answers that contradict each other is the most useful thing your board will learn this year, and it costs you an email.
Which one should we bring to the board first?
The funder concentration question. It is the least political of the ten, every committee has a real stake in the answer, and disagreement about it is informative rather than defensive. Nobody has to be wrong for it to be worth an hour.
What if the executive director and the board chair answer the same question differently?
Expected, and that is the finding. Opposite answers about the same number usually mean two systems count it differently, or that the board is reading a summary the staff would not recognize. Fix the counting before you adjudicate the disagreement.
Can we do this internally instead of paying someone?
The four marked as free you should do internally, this quarter. The rest require ranking one function against another, and everyone inside owns exactly one function. Nobody ranks their own program second, honestly, and it is unfair to ask them to.
What does a proper answer look like when it is done?
One ordered list. The reason each item outranks the next, weighed by mission impact, a named owner per item, and a date. If what comes back is a pile of findings with no order, you paid for observation when what you needed was judgment.
Where to go next
The other way in
10 signs your nonprofit needs a full look
The same decision approached from symptoms instead of questions. Useful if you want to hand something to a board member or a funder who has not been inside this conversation yet.
The Nonprofit X-Ray answers all ten at once: a full leadership roster of AI executives across finance, development, programs, communications, technology, people, legal, and governance, a recorded debate over what outranks what, and one ordered plan with an owner and a date on every line.
Could not answer six of them?That is the normal number
The ordering is the part worth paying for.
Ten answers is ten findings. What changes anything is one ranked list across all of them, weighed by mission impact, with the reasoning visible so the director who lost the argument can read why. That is what the Nonprofit X-Ray produces, fixed fee by operating budget, with a finish date. If you would rather start smaller, a free Discovery Session covers how the organization runs and what is in the way.