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.
How the money actually moves
Every programme in the catalog whose disbursement terms we hold, classified by how the money reaches the recipient. A programme can reimburse against milestones and is counted in both.
Reimbursement — you spend first, then claim it back
243 programmes · 34.4%
Paid up front as a lump sum
212 programmes · 30.0%
Released against milestones or tranches
135 programmes · 19.1%
Disbursed as loan proceeds at closing
57 programmes · 8.1%
Realised as a tax credit at filing
50 programmes · 7.1%
Paid directly to a vendor or contractor
2 programmes · 0.3%
706programmes
- Federal 185
- State 295
- Private and foundation 182
- Other 44
n = 706 of 736 catalogued programmes carry researched disbursement terms. A programme can carry two mechanisms, so the bars are not a partition; the ring is, and shows which funders those 706 programmes belong to.
The reimbursement trap
Reimbursement is the most common disbursement mechanism in US small business funding and the least advertised. A programme offering “up to $50,000” on a reimbursement basis is asking you to spend $50,000 of your own money and wait. That is a working-capital problem wearing a grant's clothes, and it is why the largest number on a page is often the wrong target.
The arithmetic nobody shows you. Take a $50,000 reimbursement grant with a 50% match. You fund the work, so you need the cash first. You claim half back after the spend is evidenced and accepted. If the first drawdown lands 30 to 60 days after the agreement is signed, the business has carried the full cost for two months or more — on top of a match it also had to find. 141 of 714 catalogued programmes (19.7%) require that match.
| Theme | All programmes | Federal | State | Private and foundation |
|---|
| Reimbursement — you spend first, then claim it back | 34.4% | 45.4% | 33.2% | 14.3% |
| Paid up front as a lump sum | 30.0% | 18.9% | 24.1% | 44.5% |
| Released against milestones or tranches | 19.1% | 30.8% | 19.7% | 6.0% |
| Disbursed as loan proceeds at closing | 8.1% | 8.6% | 8.8% | 8.2% |
| Realised as a tax credit at filing | 7.1% | 3.8% | 14.6% | 0.0% |
| Paid directly to a vendor or contractor | 0.3% | 0.5% | 0.3% | 0.0% |
Share of programmes at each funder level using the mechanism. Programmes per group: Federal n=185 · State n=295 · Private and foundation n=182. Click any column heading to sort.
Federal (n=185)
Reimbursement — you spend first, then claim it back45.4%
Paid up front as a lump sum18.9%
Released against milestones or tranches30.8%
Disbursed as loan proceeds at closing8.6%
State (n=295)
Reimbursement — you spend first, then claim it back33.2%
Paid up front as a lump sum24.1%
Released against milestones or tranches19.7%
Disbursed as loan proceeds at closing8.8%
Private and foundation (n=182)
Reimbursement — you spend first, then claim it back14.3%
Paid up front as a lump sum44.5%
Released against milestones or tranches6.0%
Disbursed as loan proceeds at closing8.2%
Federal money is slowest to reach you
45.4%45.4% of federal programmes reimburse and 30.8% release against milestones, against 18.9% paying up front.
Federal awards are also the largest in the catalog, which makes the float harder: the bigger the award, the more of your own cash the mechanism assumes you have.
Private money is fastest
44.5%44.5% of private programmes pay a lump sum and only 14.3% reimburse.
For a cash-tight business this inverts the usual advice: a small brand grant paying $10,000 into your account beats a state programme offering four times as much against receipts.
State money hides a tax-credit tail
14.6%14.6% of state programmes are realised as a tax credit at filing rather than as cash, against 0.0% privately.
A credit is real money and frequently the largest number available — but it arrives at your next filing, and only if you have the liability to offset.
Five cash-flow mistakes
1Reading the ceiling as the cheque. 34.4% of programmes reimburse. The advertised number is what you may eventually recover, not what arrives.
2Stacking a match on top of reimbursement without doing the arithmetic. 19.7% of programmes require a match; combined with reimbursement, the cash needed up front can exceed the award.
3Assuming a milestone structure pays early. 19.1% of programmes release against milestones, and the first tranche is usually the smallest.
4Counting a tax credit as working capital. 7.1% of programmes are realised at filing and only if you have liability to offset.
5Choosing on size alone. For a cash-tight business the 44.5% of private programmes that pay up front are often worth more than a larger reimbursement award they cannot float.
Questions founders actually ask
How long after winning a grant do you get the money?
It depends entirely on the mechanism, and most programmes do not pay on winning at all. 243 of 706 programmes (34.4%) reimburse after you have spent and evidenced the cost; 212 (30.0%) pay a lump sum on selection; 135 (19.1%) release against milestones. The mechanism, not the amount, decides when cash reaches you.
What does reimbursement actually mean for my bank balance?
You pay for the work first, from your own cash or a credit facility, then submit evidence and wait for the claim to be accepted. Until that claim clears, the grant has cost you money rather than provided it.
Which grants pay up front?
Private and corporate programmes, mostly: 44.5% of them pay a lump sum against 18.9% of federal programmes. This is the strongest practical argument for the smaller brand grants that larger lists dismiss.
Is a tax credit as good as a grant?
It is real money, and for state programmes it is frequently the largest number available (14.6% of state programmes work this way). But it arrives at your next filing and only offsets tax you actually owe, so it cannot solve a cash problem this quarter.
Can you tell me the terms for one specific programme?
Not on this page. Per-programme disbursement terms are the paid Win Brief; this page is the pattern across 706 programmes. Browse 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 706 carry researched disbursement terms read from each programme's own award conditions.
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 706 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.