The three games, in one table
Everything on this page is one row of a bigger pattern. Across all four dimensions we measure — why funders reject you, what they demand, how they pay and what they score — federal, state and private programmes behave like three separate markets that happen to share a word. This is the whole cluster in one view; each row links to the page behind it.
| Dimension | Federal is most likely to… | State is most likely to… | Private is most likely to… |
|---|
| Rejects you for | documents 38.9% | eligibility 34.9% | an unfinished form 33.0% |
| Demands from you | SAM.gov registration 47.1% | payroll records 40.1% | a pitch video 18.4% |
| Pays you by | reimbursement 45.4% | a tax credit 14.6% | a lump sum 44.5% |
| Scores you on | technical merit 52.4% | jobs created 45.3% | financial viability 48.4% |
Each cell names the theme that funder level raises most distinctively, with the share of programmes at that level that raise it. Percentages are within a level, not across the row.
The one-sentence version. If you are cash-tight, start private. If you are creating jobs, start state. If you are building something technically hard and can float the spend, start federal. Applying in the wrong lane is not a near miss — it is a different competition.
The criteria programmes publish, ranked
Each bar counts programmes whose published evaluation criteria or scoring rubric raise that dimension. A programme scores several at once, usually with weights it does not disclose.
Eligibility screening before any scoring
302 programmes · 42.4%
Financial viability of the business
285 programmes · 40.0%
Technical merit or innovation
215 programmes · 30.2%
Jobs created or economic impact
187 programmes · 26.2%
Community or social benefit
175 programmes · 24.5%
Fit with the funder's stated priorities
173 programmes · 24.3%
Team and management experience
169 programmes · 23.7%
Readiness to start and execute
143 programmes · 20.1%
Whether the funding is genuinely decisive
22 programmes · 3.1%
713programmes
- Federal 187
- State 300
- Private and foundation 182
- Other 44
n = 713 of 736 catalogued programmes carry researched evaluation criteria. Published criteria rarely disclose weights, so these are the dimensions scored, not their relative importance. The ring partitions those 713 programmes by funder.
The top entry is not a criterion at all. 42.4% of programmes apply an eligibility or completeness screen before anything is scored, which is why eligibility is also the commonest reason applications fail. Getting scored is a privilege you qualify for.
The same application, scored three ways
An application that wins federal money is often a poor state application and a hopeless private one, because the three funder types are not scoring the same qualities. The gaps below are the largest in this entire dataset.
| Theme | All programmes | Federal | State | Private and foundation |
|---|
| Eligibility screening before any scoring | 42.4% | 50.3% | 48.0% | 22.5% |
| Financial viability of the business | 40.0% | 35.3% | 37.0% | 48.4% |
| Technical merit or innovation | 30.2% | 52.4% | 28.7% | 11.5% |
| Jobs created or economic impact | 26.2% | 17.1% | 45.3% | 4.9% |
| Community or social benefit | 24.5% | 28.9% | 15.7% | 34.6% |
| Fit with the funder's stated priorities | 24.3% | 21.4% | 24.0% | 26.4% |
| Team and management experience | 23.7% | 35.3% | 13.7% | 33.5% |
| Readiness to start and execute | 20.1% | 36.4% | 15.3% | 13.2% |
| Whether the funding is genuinely decisive | 3.1% | 0.0% | 6.3% | 1.1% |
Share of programmes at each funder level scoring the dimension. Programmes per group: Federal n=187 · State n=300 · Private and foundation n=182. Click any column heading to sort.
Federal (n=187)
Eligibility screening before any scoring50.3%
Financial viability of the business35.3%
Technical merit or innovation52.4%
Jobs created or economic impact17.1%
Community or social benefit28.9%
State (n=300)
Eligibility screening before any scoring48.0%
Financial viability of the business37.0%
Technical merit or innovation28.7%
Jobs created or economic impact45.3%
Community or social benefit15.7%
Private and foundation (n=182)
Eligibility screening before any scoring22.5%
Financial viability of the business48.4%
Technical merit or innovation11.5%
Jobs created or economic impact4.9%
Community or social benefit34.6%
Federal scores the project
52.4%52.4% of federal programmes score technical merit or innovation against 11.5% of private ones — the single widest gap in the dataset.
36.4% also score readiness to execute and 35.3% score the team. Federal money funds work, and the case you make is about the work.
State scores the economy
45.3%45.3% of state programmes score jobs created or economic impact against 4.9% privately.
State funding exists to produce local economic outcomes, and an application that does not quantify them is arguing in the wrong currency.
Private scores the business
48.4%48.4% of private programmes score financial viability and 33.5% score the team, while only 11.5% score technical merit.
Brand and foundation programmes are asking whether this business will still exist and reflect well on them, which is why traction beats novelty in that lane.
One criterion nobody expects. 22 programmes (3.1%) score whether the funding is genuinely decisive — whether the project would proceed anyway without it. It is rare overall and concentrated at state level (6.3%), where it appears in incentive programmes. Where it appears, an application that sounds too comfortable can lose for that reason alone.
Five ways a strong application still loses
1Polishing before qualifying. 42.4% of programmes screen eligibility before scoring. Prose cannot rescue an application that never reaches a reviewer.
2Writing one application for three lanes. Federal scores technical merit 52.4%; private scores it 11.5%. The same document cannot lead on both.
3Ignoring the funder's own words. 24.3% of programmes score fit with stated priorities, which are published and rarely mirrored back.
4Asserting impact. 26.2% score jobs or economic impact and 24.5% score community benefit. Both are usually adjectives in the application and numbers in the rubric.
5Sounding too comfortable. Where a programme scores whether funding is genuinely decisive (3.1% overall, 6.3% at state level), a confident “we will do this anyway” is a losing sentence.
Questions founders actually ask
What do grant reviewers actually look for?
First, whether you qualify at all: 302 of 713 programmes (42.4%) screen eligibility or completeness before scoring. Among scored criteria the commonest are financial viability (40.0%), technical merit (30.2%), jobs or economic impact (26.2%) and community benefit (24.5%).
Do federal and private funders want the same things?
No, and the gap is the widest in this dataset. 52.4% of federal programmes score technical merit against 11.5% of private ones, while 48.4% of private programmes score financial viability against 35.3% federally.
How much do published criteria tell me about weighting?
Very little, which is a real limit of this data. Programmes publish the dimensions they score far more often than the weights they apply, so these figures show what is scored, not what matters most inside the room.
Is a strong business enough to win?
In the private lane it is close to the whole case — 48.4% score viability and 33.5% score the team. Federally it is secondary to the project itself, where 52.4% score technical merit and 36.4% score readiness.
Can you tell me how one programme scores applications?
Not here. A specific programme's criteria, and what its reviewers have rejected before, are the paid Win Brief; this page is the pattern across 713 programmes. See what carries a full win layer in the database.
Methodology, scope and what this data cannot tell you
Computed from the GrantCompass verified US catalog of 736 funding programmes, of which 713 carry researched evaluation criteria taken from published scoring rubrics, guidelines and reviewer guidance. Published criteria rarely disclose weights, so these are the dimensions scored, not their relative importance.
Themes are not mutually exclusive. A programme's own text routinely raises several concerns at once, so a programme can count toward more than one theme and the shares deliberately sum above 100%. They are not a partition and must not be read as slices of a pie.
This records what programmes say, not what happened to applicants. GrantCompass holds no applicant-level outcome data: nothing here is a rate over real applications, and no figure on this page describes how likely you are to win. It describes what 713 funders tell applicants, read consistently and counted.
Absence is not permission. Where a programme's record does not mention a requirement, that is a fact about the record, not evidence the requirement does not exist. Counts are always of programmes that state a thing, never of programmes that fail to.
Thin cells are suppressed. A share is only published for a funder level with at least 25 programmes behind it, which is enforced in the generator rather than in review. Local and municipal programmes are therefore absent from several breakdowns.
Per-programme detail — what one specific programme requires, scores and rejects on — is the paid Win Brief and is never published here. Everything on this page is an aggregate across programmes, or a public field (award ceiling, level, type, status, application effort) shown per programme.