Counting what you did is not the same as showing what changed. Most organizations are good at the first and quietly stuck on the second.
Your data lives in five systems that do not agree. A donor CRM, program software, the accounting system, and a set of spreadsheets one person maintains. Producing any number that crosses two of them is a manual reconciliation, so it happens rarely and nobody fully trusts the result.
The measures you have count activity: people served, sessions delivered, units provided. A funder reading them learns what you did, not what changed because you did it. That gap between an output and an outcome is where most impact reports stop.
And some numbers are not safe to publish. Small denominators re-identify people. Unadjusted rates punish higher-acuity caseloads. A regulator reads the same page a donor does. That decision has to be made in advance by someone who has thought about it.
One measurement system, built on the logic model your funders already think in: what you put in, what you do with it, what that produces, and what changes as a result. Most organizations report the first three stages well. Funders increasingly fund the fourth.
The system runs across the platforms you already have. I am not selling you software and I am not asking you to migrate anything. The pipeline that pulls it together is delivery infrastructure on my side, not a product you license.
Then the part that is genuinely different: governed disclosure. One measurement system, several surfaces. A program team, a board, and a funder need the same underlying numbers at three different grains and cadences. Which metric reaches which audience is a decision we make together and revisit as programs change, and withholding something is a defensible governance choice rather than an embarrassment, provided the reason is written down. Where a number does go out, it goes with a methods note naming the numerator, the denominator, the exclusions and the case-mix caveat, and where possible benchmarked against a published standard rather than against nothing.
Efficacy is proven by the same number reported two years running, with its method stated, not by publishing everything you have.
This one has a defined front door: the Impact Baseline, four to six weeks.
A working session with your leadership team on what the organization is actually trying to change, for whom, and how you would know. This is a strategy conversation before it is a data one, and it is where most of the value is.
Deep work across the donor CRM, program systems, accounting and the spreadsheets. Most organizations are holding more evidence than they realize, in worse condition than they hope.
A document you can hand to a board or a funder: your strategic direction, your outcomes framework, and what your data proves today. Including the harder answer, which is which of those numbers is ready to leave the building and which one needs work before anyone outside sees it.
Which measures go to program, which to the board, which to funders, and which stay internal for now. Written down, with the reason, so the decision survives a change of staff.
The Baseline stands alone and is yours to keep. It is also month one of the Partnership, if the reporting turns out to be the thing worth keeping alive.
I led PQI and CQI through Council on Accreditation review, which is the formal discipline of proving quality with data rather than asserting it. Inside the same organization I built the KPI dashboards our decisions actually ran on, as CFO and COO, which means I have had to defend a number to a board rather than only produce one. I also built donor-advised-fund technology at giv2giv, so I have seen the reporting relationship from the grantmaker's side of it.
There is a full impact dashboard built for a fictional treatment foster care agency, where you can switch between the program, board and funder views of the same data and watch what appears, what disappears, and the governance reason it was held back. Every number in it is synthetic.
An Impact Report is a photograph. What organizations tend to want six months later is the same picture refreshed without anyone rebuilding it by hand, and someone to ask about the judgment calls as programs change. That ongoing arrangement is the Partnership, and continuing into it involves no new project and no re-onboarding.
Thirty minutes. Bring the number you wish you could put in front of a funder, and we will work out whether your data can already support it.