Most grant applications are won or lost in the needs statement. The program design matters, the budget matters, but the first question every reviewer scores is whether you demonstrated, with credible evidence, that your community needs this investment more than the other applicants' communities need it. That is a data problem, and it has a repeatable solution.
This guide covers where the citable numbers live, how to turn them into statements of need a reviewer can verify, and how to handle the small-county data gaps that rural applicants hit constantly. It is written for the person who writes the grant, not for a data team, because in most of the organizations that need this guide, those are the same person.
What reviewers actually score
Read any federal scoring rubric (HRSA, USDA, SAMHSA, and most state block grants publish theirs) and the demonstrated-need criteria reduce to five properties your evidence either has or lacks:
- Recency. Data from the current or prior release year. A 2019 statistic in a 2026 application invites the question of what you are hiding.
- Comparison.A rate means nothing alone. “14.2% of adults have diabetes” is a fact; “14.2% of adults have diabetes, against 11.3% statewide and 10.1% nationally” is evidence of need.
- Direction. A worsening measure is stronger evidence than a static one. Five years of trend in one sentence outweighs a paragraph of adjectives.
- Local specificity.Numbers for your actual service area, not the state, not the region, not a national statistic with “and our community is no exception” appended.
- Traceable citations. Source and year on every claim, in a form the reviewer could check in thirty seconds. Some of them do.
The six sources that cover most needs statements
You can assemble a defensible needs statement for almost any U.S. county from six public sources. All are free, all are federal or federally recognized, and all are names grant reviewers already trust.
| Source | What it gives you | Cadence | Best for |
|---|---|---|---|
| County Health Rankings & Roadmaps (University of Wisconsin Population Health Institute) | Roughly 90 measures per county: health outcomes, health behaviors, clinical care, social and economic factors, physical environment | Annual release | The backbone of most needs statements. One citation covers many measures, and reviewers recognize it instantly. |
| CDC PLACES | Model-based prevalence estimates for chronic conditions (diabetes, COPD, asthma, depression, heart disease and more) at county, place, tract, and ZIP level | Annual release | Disease burden claims. The only consistent source for chronic disease prevalence in counties too small for direct survey estimates. |
| Census Bureau, American Community Survey (5-year estimates) | Poverty, income, insurance coverage, housing, broadband, language, disability, age structure | Annual release of rolling 5-year estimates | Social and economic need. The 5-year file covers every county regardless of size. |
| CDC WONDER | Mortality by cause, county, age, race, and year, from death certificates | Updated as final and provisional files post | Mortality claims. Note that small counts are suppressed; multi-year and multi-county groupings often unlock them. |
| HRSA shortage designations (HPSA / MUA-P) | Federal designations for primary care, dental, and mental health shortage areas, with scores | Continuously maintained | Access claims. Many federal programs score or require these designations directly, so cite them by name and score. |
| EPA EJScreen and related EPA data | Environmental indicators (air toxics, PM2.5, proximity to facilities) joined to demographic indicators | Annual-ish updates | Environmental justice framing, which an increasing number of federal funders ask for explicitly. |
Two habits make these sources work harder. First, cite the data year, not just the source: “CDC PLACES, 2024 release” reads as competence. Second, keep the number of distinct sources small. A needs statement drawing cleanly on four sources beats one that name-drops eleven, because every additional source is another methodology the reviewer has to trust.
The small-county problem
If your service area is rural, you will hit suppression: the practice of withholding estimates when the underlying counts are too small to be reliable or private. Mortality data suppresses small counts entirely. Infant mortality, one of the most powerful measures in any maternal health or early childhood application, is published for only about a third of U.S. counties in a given year. Survey-based measures get wide confidence intervals or vanish.
Three honest ways through:
- Combine counties. A measure suppressed for each of four small counties is often publishable for the four combined. Funders think in service areas anyway, so a combined-area figure usually fits the application better than county-by-county fragments would have.
- Combine years.CDC WONDER and similar tools will compute multi-year rates that clear suppression thresholds a single year cannot. Label the period plainly (“2018–2022 combined”).
- Substitute the measure, and say so.When infant mortality is suppressed, low birth weight usually is not, and it is an accepted needs proxy in the same domain. Reviewers respect a sentence that says “county-level infant mortality is suppressed for populations this small; we report low birth weight, which is available for every county in the service area.” It signals you know the data rather than that the data defeated you.
Turning numbers into need
The same statistic can read as trivia or as evidence. The difference is three moves, applied in one or two sentences.
Benchmark it.Every headline rate gets a state and national comparison. If your number is worse than most counties in the country, say how much worse in plain words, not percentiles: “a diabetes rate higher than roughly nine in ten U.S. counties” lands; “92nd percentile” makes a tired reviewer do math.
Count the people.Rates are abstractions; reviewers fund people. Multiply the rate by the population it covers and write both: “18.9% of adults, roughly 3,900 residents of the service area, live with diagnosed diabetes.” This single habit does more for a needs statement than any other editing pass.
Show the direction.Where a measure has worsened across recent releases, put the arc in the sentence: “up from 15.1% five years ago.” If it has improved, do not hide it; pair it with the measure that has not, and let the honest contrast carry your credibility.
Mistakes that cost points
- Citing a statewide number for a local program, or a national statistic with no local anchor at all.
- Stale vintages, especially when the current release is one search away and the reviewer knows it.
- Cherry-picking so aggressively that the picture is implausible. A community that is worst at everything reads as a data problem, not a needs case.
- Jargon transplanted from analytics tools: percentiles, index scores, and composite measures without a plain statement of what they mean for residents.
- Uncited claims sitting next to cited ones. One naked statistic can make a reviewer re-check all the others.
If you want the assembled version
Everything above is doable by hand, and for a single-county application it might take a practiced writer a day. If you would rather not spend the day: every U.S. county's environmental, health, access, and social data is free to browse on our county pages, and for $10 we will generate a verified data report for your exact service area in a needs-statement format: state and national benchmarks, trends, residents affected, and a citation on every number, with each claim checked against the source data before the PDF is assembled. It exists because grant writers kept asking us for exactly that document.
Either way, the principle is the same. Reviewers do not fund adjectives. They fund communities whose need is legible in numbers they can trace. Make the tracing easy and the rest of the application gets read in a better light.