Grant Report and Indicator Crosswalk
Every funder defines served differently, so the same programme has to report 1,000 people to one and 214 to another. Both are right.
Free download · No account needed
Reporting burden scales with the number of funders rather than the size of the programme, and the reason is not narrative. It is that the programme ran once and each agreement defines its indicators differently. Served means unduplicated individuals with a billable encounter to one funder, households to another, three-or-more contacts to a third, new clients only to a fourth, and encounters rather than people to a fifth. Every one of those numbers is correct, and none of them is the figure on your own dashboard.
So the artifact is a definition table rather than a report template. Indicator Crosswalk holds one row per funder indicator: the funder's definition quoted verbatim, the internal filter that produces it, the value for the period, and its ratio to your headline figure. The objectives come from whatever the proposal committed to. What is being saved here is derivation, not writing, and derivation by hand is why this quarter disagrees with last quarter.
In the worked example a single quarter of one programme produces 1,000, 812, 612, 447, 289 and 214 people, plus 3,146 encounters. The people figures span a factor of 4.7, and because spend is identical at $412,000, cost per participant spans $412 to $1,925 by exactly that factor. So anybody quoting a cost per participant without naming its denominator is quoting a number that has a 4.7x range built into it.
What is in the pack
Indicator Crosswalk
One row per funder indicator carrying that funder's definition quoted verbatim, the internal filter that produces it, and the value for the period.
Every filter run in one pass
All indicators for all awards derived together from one participant record, which is what stops this quarter's figure disagreeing with last quarter's.
Reporting Calendar computed, not typed
Each due date derived from the award's own cadence and interval, so a period end moving moves every deadline that depends on it.
Financial to Programme Reconciliation
Programme cost by award against the share of participants each award claims, which is what surfaces double-counted people and unfunded programme cost.
Variance Explanation with four causes separated
A definition change, a data timing effect, a real programme variance and a target that was wrong, kept apart so the section reads as management.
The no-filter row
An indicator the data cannot produce says so and stays blank. Nothing gets approximated, because an untraceable outcome figure is not a reporting gap.
How it works
- 1
Send the executed agreements
All of them, with attachments and any reporting guidance sent separately. The indicator definitions live in there and they are what makes each number different.
- 2
Quote every definition verbatim
Copied rather than paraphrased, then mapped to the specific field or filter in your data that produces it. Paraphrasing drifts toward the number you already have.
- 3
Run one pass over the period
Every filter, every award, at once. The spread gets reported with its ratios so it reads as expected rather than as a data error somebody has to hunt.
- 4
Relate the money to the people
Cost by award against participants claimed by award, then the narrative per funder from that funder's crosswalk rows and nothing else.
Frequently asked questions
Why not just report the same number to everyone?
Because none of them asked for it. Each agreement defines its own indicator, and reporting your headline figure against a definition that excludes 188 of those people is inaccurate rather than simplified. The crosswalk does not make the numbers agree. It makes the disagreement explicable, repeatable and defensible.
How are the deadlines worked out?
From each award's stated cadence and interval. Federal financial reports submitted quarterly or semiannually are due no later than 30 calendar days after the period, annual reports 90 days, and the final financial report 120 days after the period of performance ends. Those are outer bounds, so awards often state something tighter.
Is a wide spread between funders a sign my data is wrong?
No, and that framing matters. Every figure in the example traces to the same participant record and every one is correct. They answer different questions. A programme manager who sees 1,000 and 214 for the same quarter without being told this will spend a day looking for an error that does not exist.
What if a funder asks for something I cannot measure?
The row says no filter available, and so does the report. Then it says what the data can support and offers that instead. Approximating an outcome figure is the one move this pack refuses, because a number nobody can trace to the participant record is not a reporting gap.
Why does the reconciliation matter?
Because performance reporting has to relate financial data to programme accomplishments, and doing that honestly surfaces things no single report shows. In the example six awards claim 106% of the participants and 15.8% of programme cost is on unrestricted funds.
Does this complete the SF-425?
No. The federal financial report has its own accounting fields, including cash on hand, the indirect cost rate and unliquidated obligations, and tying source vouchers to those is separate work. What each agreement obliges you to file at all belongs in the award terms register. Where a report uploads through a portal, the same tagging rules apply as at submission.
Does a good report actually help?
It is most of what earns a renewal, and renewals are where the large grants are. Reading foundations' own filings in prospect research shows the pattern plainly: the biggest awards go to organisations already on the list, so a first grant is really an audition for the second.
Find out how many numbers your quarter really has
Send the executed agreements and one period of programme data. The first thing back is every funder's own definition next to the figure it produces.
Build my crosswalk