Your board asks for data you don't fully trust. Or they ask a question you can't answer. Or they push back on a number in your report and you're not sure how to respond.
These moments are uncomfortable. They're also more common than most EDs will admit publicly.
Here's how to handle them without losing credibility, and how to use them to build something better.
Why this is harder than it sounds
The instinct in these situations is to either defend the data ("these are the numbers we have") or deflect ("we'll look into that"). Both responses close the conversation without actually solving anything.
The real problem is that many nonprofit leaders are caught between two things: they know their data isn't perfect, and they're afraid that admitting it will undermine confidence in their leadership.
It won't. What undermines confidence is getting caught presenting numbers you can't explain, or promising to follow up and never doing it.
Start by being honest about what you know
Before your next board meeting, spend 30 minutes with your key metrics and ask yourself honestly: how confident am I in each of these numbers? Where did they come from? If a board member asked me to explain the methodology, could I?
Sort your metrics into three buckets:
High confidence. You know exactly where this number comes from, how it was calculated, and you'd stand behind it in an audit.
Medium confidence. The number is probably right, but there are assumptions baked in that you haven't fully documented.
Low confidence. You're not sure this number is accurate, or you know there are data quality issues that could affect it.
You don't need to share this inventory with your board. But having it in your head changes how you present data. High-confidence numbers get presented directly. Medium-confidence numbers get a sentence of context. Low-confidence numbers either get fixed before the meeting or get flagged honestly.
How to flag uncertainty without alarming anyone
There's a big difference between "our data is a mess" and "we're tracking this metric, and I want to flag that we're still refining our methodology."
The second framing is honest, professional, and doesn't invite panic. It also opens a conversation about improvement rather than closing one about a mistake.
Examples:
"This number comes from our CRM. I'm confident in the Q1 figure; we're still reconciling the Q2 data and I'll have a clean number for the next meeting."
"We're tracking donor retention this year for the first time. The methodology is straightforward, but I want to flag that we don't have prior-year data to compare against yet."
"This program participation figure is from our database. I know we had some inconsistent entry earlier in the year, so treat it as directionally accurate rather than exact."
None of these statements destroy confidence. They demonstrate that you know your data well enough to describe its limitations.
When a board member challenges a number
Someone asks where a figure came from. Or points out that it doesn't match something they remember from a previous report.
Don't guess. Don't get defensive. Say: "Good question. Let me pull the source data and follow up with you after the meeting." Then do it.
If you know the answer in the room, give it. If you don't, the worst thing you can do is speculate and be wrong. Taking it offline is a sign of rigor, not weakness.
Use these moments to build support for better data
A board member who asks hard questions about your data is an asset. They're telling you that data quality matters to them, and that they're paying attention.
Use the moment: "You're right that this number is harder to pin down than it should be. We've actually been thinking about how to improve our program tracking, and I'd love to get the board's input on where we should prioritize."
That turns a potentially awkward moment into a governance conversation about organizational capacity. It also builds the case for investing in data infrastructure, which is much easier to do when the board has personally experienced the problem.
The goal is confidence, not perfection
You're not trying to have perfect data before you can have confident conversations with your board. You're trying to know your data well enough to present it honestly, explain its limitations, and describe what you're doing to improve it.
That's a much more achievable standard. And it's one that builds genuine trust over time.
If you want to assess where your data stands before your next board meeting, our free Nonprofit Data Health Checklist takes about 10 minutes and covers the areas most likely to surface in board discussions.
If you work through it and want to talk through what you find, I'm happy to do a free 30-minute discovery call.
Joshua Barillas is the founder of Prismatic Consulting, a data services firm built exclusively for nonprofits. Learn more about our services or get in touch at hello@prismaticconsulting.us.