
How to apply for small business grants: what the eligibility line really asks, and why disbursement decides everything
Ownership tests, the difference between a grant and a loan programme, and the one field that decides whether a small business can actually use an award it wins.
Searching for "small business grants" returns three different things wearing one label: actual grants, loan programmes, and services that will help you apply for either. Telling them apart is the first job, and it is not always obvious from a listing.
This is about the second job — reading a real grant notice as a company rather than as a nonprofit, because the tests are different and the field that decides usability is a different field.
The eligibility line asks about ownership, not about mission
A nonprofit eligibility section asks what kind of organisation you are and what you do. A business eligibility section asks who owns you and where the work happens.
On federal research programmes the tests are specific and unforgiving. An SBIR award requires a small business majority-owned by US individuals or by other eligible small businesses, with the work performed principally in the United States. Venture ownership rules vary by agency and are the single most common reason a company that looks obviously eligible is not. STTR adds a required research partner and a prescribed split of the work, which is a constraint on the project rather than on the company.
On state economic development programmes the tests are usually location and headcount: where the business operates, how many people it employs, sometimes how long it has traded. These are stated in the notice, in prose, and they are what our eligible entity type field is recording when it says a notice is open to businesses.
Two things worth knowing about how those values reach a record. First, they are the source's own words mapped into a small vocabulary, and where a source category has no honest home it stays unmapped rather than being forced into the nearest one — a wrong mapping puts a record in front of a company it does not fit. Second, a notice that says nothing about who may apply produces Not stated rather than a list of every type it did not rule out.
Grants open to businesses is the slice to work from. The matching wizard filters it against what you tell it, and treats an unstated field as unknown rather than as satisfied, so a record never clears a filter on a field nobody confirmed.
Disbursement is the field that decides whether you can use the money
For a nonprofit with reserves, reimbursement is an inconvenience. For a company without them it is frequently disqualifying, and almost no grant listing records it as a field at all.
The arithmetic: a 500,000 dollar award paid in arrears requires an organisation that can carry 500,000 dollars of spending before the first payment arrives. A company with 150,000 dollars of runway cannot accept that award regardless of how well it fits, and discovering this after selection is worse than discovering it before.
Some programmes pay an advance, some mix the two — an advance to start and reimbursement thereafter — and some allow a working capital request under stated conditions. Where a notice describes a mixed arrangement, the record stores the arrangement rather than picking the simpler label.
We hold this as a stored, filterable field rather than a phrase buried in a description paragraph. There is one known limitation and it is more honest to state it than to let it sit: the reimbursement flag currently defaults to "not reimbursement" when a source says nothing, which means no and the source did not say are stored as the same value on this one field. Until that is fixed, treat an unmarked record as unknown and read the notice.
The other cost is time. Grant money is non-dilutive — no equity, no repayment — and it is not free: the price is the application, the reporting, and frequently a year between submission and cash. A 50,000 dollar award that costs six weeks of a founder's time is a decision, not a windfall.
Plan backwards, and register before you need to
The calendar problem for a company is the same as everyone else's with one addition: nobody in a small company has application work in their week already.
Working back from a submission date: registration, a budget that survives scrutiny, any partner commitments, and the internal decision about whose time this is coming out of. Federal submission requires an active entity registration that takes weeks the first time and lapses annually and quietly — a company that applied last year is not necessarily able to apply today.
The deadline planner works backwards from a deadline to the date each step has to be done. The award size checker does the other half: whether the award sizes in a slice of the catalogue are the size of your project at all, printing the number of records it computed from rather than a lone figure.
If the answer is that this cycle is not achievable, the useful work is the registration. It makes the next cycle a decision rather than a scramble.
What we cannot tell you
We cannot tell you whether a programme will fund a company like yours. We can tell you what its notice said about who may apply, and where the notice said nothing, that it said nothing.
We do not hold loan programmes, tax credits or equity programmes, so a search here is a search of grants specifically. That is a boundary rather than a completeness claim.
We publish no success rates and no fit scores. Where a notice states an expected number of awards, that sentence is on the record; where it does not, there is no honest figure to give and we do not manufacture one.
And we are not the only place to look. Skip, at helloskip.com, is built for small business owners and does something no database can: it awards grants of its own — 1,000 dollar Instant Grants and 10,000 dollar Skip Grants — alongside a free tier that states access to all public grants and funding. That deserves the credit. FindGrants offers free discovery with no account and sells a one-time application pack rather than a subscription, which is an honest answer to a real complaint about this category. Neither of them tells you when a field on a record was last confirmed against its source, which is the gap we are trying to close, and Instrumentl — the strongest prospect research tool in the category — states a scope that excludes small businesses entirely.
Sources
- Grants.gov search2 API, oppStatus counts — 1,025 posted, 561 forecasted — https://api.grants.gov/v1/api/search2 — read 7 September 2026
- GrantTrove catalogue slice for businesses, HTTP 200 — https://granttrove.com/grants/for/business — read 7 September 2026
- Skip pricing page — Starter free with access to all public grants and funding, Plus $29/mo, Pro $99/mo, and Skip's own $1,000 Instant Grants and $10,000 Skip Grants — https://helloskip.com/pricing — read 7 September 2026
- FindGrants pricing page — discovery free, Application Pack $99 per grant one-time — https://findgrants.io/pricing — read 7 September 2026
- Instrumentl stated scope, small businesses outside its stated scope — competitor registry entry checked 2 September 2026
Related
Field definitions
How we compare
In the catalogue
Read next
- How to apply for research grants: the institution applies, the indirect rate decides the budget, and the calendar is eighteen months
- How to apply for individual grants: the eligibility line that excludes most people, and what to do about it
- How to apply for K-12 grants: who the applicant is, what supplanting means, and the school-year calendar problem
Author

Series
Eligibility by applicant typeOne guide per applicant class, on how to read an eligibility section, calculate what a match would cost you, and lay out a timeline that fits your own calendar.
Categories
More Posts
3

A $75 million grant that funds nothing at $75 million
The headline number on a funding announcement is usually the programme pool — the total shared across every award. The most one applicant can receive is a different field, and on a real record in our catalogue it is not published at all.


How to apply for tribal grants: set-asides, sovereignty, and the eligibility values that get mapped wrong
Tribal governments and tribally designated organisations face an eligibility landscape nobody else does: dedicated set-asides, a recognition test that is binary, and listing vocabularies that flatten the distinctions.


How to apply for research grants: the institution applies, the indirect rate decides the budget, and the calendar is eighteen months
Research funding runs on rules that are invisible in a listing: who may hold an award, what an indirect cost rate is applied to, and how long the gap between submission and money actually is.

Newsletter
Join the community
Subscribe to our newsletter for the latest news and updates