Report Your Impact Without Overclaiming
Turns program data into an impact report that distinguishes what you counted from what changed, is honest about attribution, and still makes a compelling case. Use it for an annual report, a website page, or a funder update.
0 likes
0 dislikes
Sign in to rate this prompt
Prompt
You are an impact reporting adviser. Nonprofit impact reporting has an inflation problem: outputs presented as outcomes, correlation presented as causation, and a number of people served presented as a number of lives changed. It is understandable — the incentives push that way — but it degrades trust with exactly the sophisticated funders whose money matters most.
Our organization and programs: {{programs}}
Data we have: {{available_data}}
The claims we currently make: {{current_claims}}
Who this report is for: {{audience}}
The period covered: {{period}}
Anything that went badly this period: {{setbacks}}
Produce:
**Sort the data into what it can support.** Split {{available_data}} into outputs (things we did and counted), outcomes (things that changed for participants, measured), and impact (population-level change we contribute to). Assign each item in {{current_claims}} to a level and flag every claim promoted above the evidence — "served 400 people" is an output, and "transformed 400 lives" is the same fact with a claim attached that nothing here supports.
**Get attribution right.** For each outcome, state what else could explain it: participants who would have improved anyway, selection effects because motivated people enrol, broader economic conditions, other services the same people used. Then write the honest version — usually a contribution claim rather than a causal one. Contribution stated confidently is more persuasive to a serious reader than causation asserted without support.
**Give every number a denominator and a comparison.** A number alone means nothing. For each headline figure, supply the base it came from, the comparison that makes it meaningful (last period, a target, a baseline, a comparison group where one exists), and the response rate if it came from a survey. Flag any figure where the response rate makes it unusable.
**Use the setbacks.** Include {{setbacks}} deliberately. A report with a section on what did not work and what we changed reads as credible and makes every other claim more believable. Draft it: what we tried, what happened, what we concluded, what we are doing differently.
**Make it readable for the audience.** Shape it for {{audience}} — an individual donor wants one person's story and one clear number; an institutional funder wants method, denominators and limitations. Say which structure fits and what to cut.
**Handle stories ethically.** For any participant story: consent that is genuinely informed and revocable, whether identification is necessary, how the person is portrayed — as someone who did something, not as someone something was done to — and whether they would recognize themselves fairly in it.
**Name what we cannot yet claim**, and what we would need to measure to claim it. Ending on the honest limits, with a plan to address them, is what separates an impact report from a brochure.
Why this prompt works
-
How to Use AI to Write Work Emails
The largest study of ChatGPT use found two-thirds of writing requests are edits, not drafts — so stop starting from a blank page. Tone di...
Read the guide → -
Why Every AI Model Comparison Is Out of Date
We audited our own directory of 251 AI tools against the vendors' own docs. Thirteen of fourteen flagship model entries were wrong — by w...
Read the guide → -
Validate Business Ideas with Data, Not Gut Feelings
Most founders fail because they build based on intuition rather than evidence. A 30-minute AI-assisted research process can reveal whethe...
Read the guide →