Non-dilutive is the word founders latch onto, and the effect is real: a federal research award funds work without selling stock. But the size of that effect is easy to overstate and easy to understand precisely, so it is worth doing the arithmetic once rather than arguing about it in the abstract.
A worked example
Take a hardware company that needs about $1.5M to get through its core R&D. Two paths, same technical outcome:
| Stage | Path A: equity-funded R&D | Path B: grant-funded R&D |
|---|---|---|
| Fund the R&D | Raise $1.5M at a $6M post-money valuation; sell 25% | Phase I and Phase II awards cover it; sell nothing |
| Founders after that step | 75% | 100% |
| Series A: raise $1.5M | At $12M post; sell 12.5% | At $12M post; sell 12.5% |
| Founders at exit | Roughly 66% | Roughly 88% |
At a $50M exit that difference is worth around $11M to the founding team. Nothing about the technology changed. The only variable was who paid for the research. If the mechanics of that dilution are unfamiliar, how each round costs you ownership and what a cap table tracks are the place to start.
Now subtract the costs
The table above is the optimistic reading, and it is the version founders repeat to each other. Three real costs never appear on a cap table, and together they can eat most of the advantage.
- Founder hours. A serious proposal, plus the Phase II proposal behind it, consumes several hundred hours of exactly the people who cannot be replaced. Those hours have an opportunity cost measured in customers not called.
- Reporting and accounting overhead. Federal money brings technical reports, financial reports, timekeeping discipline, and an accounting system that can survive scrutiny. For a small team this is real ongoing work, sometimes a part-time hire.
- Requirement drift. The most expensive risk, and the least visible. You spend two years building to an agency's specification, hit every milestone, and end up with something no commercial buyer wants at the price you can build it for. The award was free; the roadmap it bought was not.
There is also a timing effect the table hides: the grant path is slower. If it delays your Series A by a year, the higher valuation you were counting on may simply not be there, and a competitor may have taken the market.
When the math genuinely favors the grant
The case is strongest when the work is real research with a binary outcome, when your commercial market is years out anyway, and when a failed experiment would have made the round unraisable. It is weakest when the work is engineering execution on a known design, where speed matters more than ownership percentage. That tradeoff is the whole of grants versus investors.
One more line item people forget: award money is generally treated as taxable business income rather than a gift, which changes the net figure — see how SBIR funds are taxed. All of this is general information, not tax, legal, or securities advice; run your own numbers past your CPA and counsel before deciding.
Projects House helps founders scope the R&D that sits behind either path. Bring us your development plan through our contact form.