How to Respond to a Funder's Evidence Request

3 min read ▶ Listen to this post

Start by matching your response to the evidence you actually have, not the evidence you wish you had. A funder asking "what's your evidence this works" is really asking one of three different questions — did it happen, did it change, did it change because of you — and the honest answer depends on which kind of data sits behind your program right now.

First, know which tier of evidence you're standing on

• Output data only — sessions held, people served, materials distributed. This answers "did it happen," not "did it work." Say so plainly rather than letting a busy funder infer more than the data supports.
• A pre/post number with no comparison group. This shows something moved. It can't show your program moved it, because nothing rules out the economy, the season, or simple regression to the mean doing the work instead.
• A comparison group of some kind — a matched group, a staggered rollout, a difference-in-differences design. This is where a genuine causal claim starts to be defensible.
• Randomized evidence. The strongest tier, and rarely required for every request — funders generally don't expect an RCT from a program your size, and asking for one when you don't have it can read as overcorrecting.

Write the response to match the tier, not to impress

If you only have outputs, lead with them honestly and pair them with a stated theory of change: what you believe those outputs cause, and why, even without outcome data yet. If you have a pre/post number, report it as exactly that — a before/after change — and name what else could plausibly explain it, rather than letting the funder assume causation you haven't shown. If you have a comparison group, this is the moment to lead with it; it's usually the single most persuasive thing in the response, and burying it under output metrics undersells it. In every tier, a funder who reads a lot of these requests can tell the difference between confident and overclaimed within a paragraph, and the overclaimed version is the one that gets follow-up questions you can't answer.

If the honest answer is "not much yet"

Say that directly, and pair it with a plan: what evidence you're building toward, on what timeline, and what it would take to get there — a comparison group you're already planning, a baseline you're collecting this cycle. A funder asking for evidence is very often trying to figure out whether you understand your own program's evidence gaps, not whether you've already closed all of them. A credible plan to close a gap reads better than a stretched claim that pretends the gap isn't there.

Two mistakes that undo an otherwise good response

• Reporting only the outcomes that improved. If you tracked five outcomes and one moved, say so — a response that mentions only the winner reads, to an experienced reviewer, as a sign the other four are being hidden rather than absent.
• Borrowing someone else's causal claim. Citing a national statistic or another program's published result as if it applies to yours conflates "this kind of program can work" with "our program worked here" — two different claims that need different evidence.

If you're drafting a response to a funder's evidence request and aren't sure which tier your current data actually supports, that's exactly the kind of gap we help nonprofits assess and close.

Related reading

• What Funders Actually Expect From a Program Evaluation — the proactive version of this question: what to build into an evaluation before a funder ever asks for it.
• When Does a Program Need a Control Group? — how to tell which evidence tier your program actually needs to be credible, before you're asked to defend it.

One email whenever we publish something new. No spam, unsubscribe anytime.