The hardest money to raise is the first fifty thousand dollars. Investors want evidence that the technical approach works; producing that evidence requires building something; building something requires money. Revenue does not exist yet, so a lender has nothing to underwrite. This is the gap that non-dilutive proof-of-concept funding exists to fill, and in the United States there is more of it than most founders realize — it is simply not advertised, and almost none of it comes from the places people look first.
What follows is the realistic map: what each source funds, what it pays, and what you have to have before applying.
First, be clear on what a proof of concept is
Funders in this space are buying the answer to a technical question, not a product. A proof of concept demonstrates that the underlying principle works — that the sensor can resolve the signal, that the chemistry is stable, that the mechanism survives the load. It is deliberately ugly and usually not a prototype at all, a distinction worth internalizing from proof of concept vs prototype.
Which means your application has to be framed as a hypothesis with a falsifiable answer, not as a business plan. "We will determine whether a passive thermal path can hold the sensor within ±0.5 °C over a 40 °C ambient swing" wins money. "We will develop our product" does not. If the honest answer to your central technical question could turn out to be no, you are writing the right kind of proposal.
SBIR Phase I: the main instrument
The Small Business Innovation Research program is by a wide margin the largest source of proof-of-concept money in the country. Federal agencies with substantial research budgets are required to set aside a percentage of those budgets for small businesses, and Phase I is explicitly a feasibility study — the government paying you to find out whether your idea works.
- What it pays. Phase I awards commonly run from about $50,000 to roughly $300,000 depending on the agency, and several award close to the statutory guideline. Phase II, for the ones that succeed, runs to roughly $2 million.
- How long. Typically six to twelve months of work.
- What it costs you. No equity, no repayment. You keep the intellectual property, subject to government use rights.
- Who runs it. Each participating agency runs its own program with its own topics and culture — the Department of Defense buys against specific solicited needs, the National Science Foundation funds open-topic deep technology, the National Institutes of Health funds health applications, and the Department of Energy, NASA, and USDA all run their own. Picking correctly is most of the battle; see NIH, NSF, DoD or DOE: which agency should you apply to.
- Eligibility basics. A for-profit US small business, majority US-owned, under 500 employees, with the principal investigator primarily employed by the company during the project.
The full mechanics — registrations you need before you can even submit, proposal structure, and how the phases connect — are in our SBIR grant application guide and in Phase I vs Phase II.
STTR, when the science lives in a lab
The Small Business Technology Transfer program is SBIR's sibling, with one structural difference: it requires a formal partnership with a nonprofit research institution, which must perform a defined share of the work. If your proof of concept depends on equipment, expertise, or a discovery that sits inside a university, STTR is often the better fit and the competition is somewhat thinner. The differences that matter in practice are in SBIR vs STTR.
NSF I-Corps, when the uncertainty is commercial
Sometimes what is unproven is not the physics but whether anyone wants it. NSF's Innovation Corps program funds structured customer discovery: teams go through an intensive curriculum built around interviewing a large number of potential customers, with a modest grant attached. Regional programs run shorter versions at lower cost and are open to applicants without a prior NSF award.
Two reasons to take it seriously even if the grant is small. It frequently kills bad ideas cheaply, which is a service. And completing it materially strengthens the commercialization section of a later SBIR proposal, which is often the weakest part of an engineer-written application.
University proof-of-concept centers and gap funds
Many research universities operate internal funds specifically for the stage between a lab result and something a licensee will look at. They go by different names — proof-of-concept centers, gap funds, translational research funds, commercialization awards — and typically pay in the tens of thousands of dollars.
Access usually runs through the technology transfer office, and eligibility usually requires some connection to the institution: a faculty co-investigator, a license to university intellectual property, or an affiliated incubator. If your technology touches academic research at all, that office is worth a meeting; the working relationship is covered in partnering with universities and research labs. Be careful about IP terms before you take the money, since university funds often come with license or revenue-share strings.
State and regional programs
This layer is the most underused. Depending on where you are incorporated, you may find state SBIR match or bridge grants that add money on top of a federal award or fund the proposal-writing effort itself; state technology development funds and university-affiliated seed programs; economic development grants aimed at manufacturing and job creation; and the federally funded Manufacturing Extension Partnership centers, which provide subsidized engineering and manufacturing help rather than cash. Amounts are smaller and competition is far lower, because fewer people know these exist.
Prizes, challenges, and corporate programs
Federal agencies run open prize competitions where the award is paid for a demonstrated result rather than a proposal — no equity, no reporting, and often no eligibility restrictions beyond being a US entity. Private foundations fund proof of concept in mission areas such as health, climate, and accessibility. Large corporations run open-innovation calls in their supply chains, though those usually come with commercial strings worth reading carefully.
How the sources compare
| Source | Typical amount | What it buys | Dilution |
|---|---|---|---|
| SBIR Phase I | $50,000 – $300,000+ | Six to twelve months of feasibility work | None |
| STTR Phase I | Similar to SBIR | Feasibility work shared with a research institution | None |
| NSF I-Corps | Tens of thousands | Customer discovery and market validation | None |
| University gap fund | $25,000 – $100,000 | Bench-to-demo translational work | None, but IP terms apply |
| State match or grant | $10,000 – $100,000 | Proposal support, bridge funding, equipment | None |
| Prize challenge | Varies widely | A demonstrated result, paid after the fact | None |
Plan around the timeline, not the amount
The catch with all of it is speed. Federal solicitations open on fixed cycles, and the interval from submission to money in the bank commonly runs six to nine months — see how long it takes to hear back on an SBIR application. If your runway is shorter than that, non-dilutive funding cannot rescue you and you are looking at a different conversation, which is where the comparison in grants vs investors becomes the relevant one. The founders who use this money well apply early, apply to several programs at once, keep working while decisions are pending, and treat every rejection as a free set of reviewer comments to rewrite against.
Projects House is a product development engineering firm. We help founders turn a vague idea into the shape a proof-of-concept funder recognizes: a specific technical hypothesis, a work plan with milestones and go/no-go criteria, a credible budget, and the engineering execution once the award lands. We are not a law firm, and anything touching patent filings should go through a registered patent attorney. To scope the technical plan behind a proof-of-concept application, reach us through the contact form.