Every founder asks the same question before committing six weeks to a proposal: what are the actual chances? The honest answer is that agency-published selection rates for SBIR Phase I generally land somewhere in the range of ten to twenty-five percent, varying by agency, by topic, and by cycle. Some narrow topics at small agencies draw a handful of proposals and fund one. Some NIH institutes and popular DoD topics draw many more and fund a similar number.
That headline is also the least useful statistic available, because the pool is not uniform. A meaningful share of submissions are administratively rejected, badly mismatched, or written in a weekend. Your real competition is the subset of serious, well-matched proposals — against that group your odds are worse than the headline, and against the whole pool a genuinely strong application does considerably better than one in five.
The numbers that matter more
Three secondary statistics shape planning better than the topline rate:
- Phase II selection rates are much higher. Only Phase I awardees compete, the pool is small, and agencies want their initial investment carried through; forty to sixty percent is common. Phase I is the hard gate, and its real value is buying the right to compete for the far larger Phase II award.
- Resubmission rates are strong. Most eventually funded companies were declined at least once. Where reviewers give substantive comments, a revision that addresses them directly is a materially different proposal.
- Rates vary hugely by agency. USDA, DOT, NIST, and EPA draw far fewer applicants than NIH and NSF. Where you apply may move your odds more than anything you write.
That last point is the highest-leverage decision in the whole process, and it starts with reading the eligibility and phase structure carefully in our SBIR grant application guide.
Why proposals lose
Debriefs across agencies repeat a short list of causes.
Administrative failure
Expired registrations, missing forms, exceeded page limits, budget over the cap. These proposals are never reviewed on merit — preventable, and still common. Start registrations weeks early and have someone other than the author run the compliance checklist.
Weak topic match
At the mission agencies this is the dominant cause. A founder sees keyword overlap with a topic and writes a proposal that is really about their existing roadmap. The topic author, who wrote the requirement because of a specific operational problem, does not see that problem addressed. Partial matches almost never win.
A commercialization plan that convinces nobody
Reviewers score commercial potential, and at some agencies it carries weight equal to technical merit. Top-down market sizing, no evidence of customer contact, no named competitors, and a funding path that ends at "we will apply for Phase II" — that combination sinks otherwise excellent science. How the scoring sheet actually weights these criteria is laid out in how federal grant review panels score proposals.
An overreaching work plan
Proposing to design, build, test, and validate a full system in a six-month Phase I signals that you do not understand the phase. Reviewers read ambition beyond the budget as naivety. A Phase I should answer one crisp feasibility question with a measurable pass/fail criterion. The scope difference between phases is spelled out in SBIR Phase I versus Phase II.
Vagueness
"We will optimize the system architecture" is not a task. Reviewers reward specificity: this material, this test method, this acceptance threshold, this named piece of equipment.
What measurably improves your odds
- Talk to the topic author during the pre-release window. At topic-driven agencies this is the strongest move available: you learn what problem sits behind the topic and what has already been tried. Once the solicitation opens, that channel closes.
- Include letters of support from real prospective customers. Two weeks of effort, disproportionate effect. A specific letter describing the buyer's current frustration reframes how a reviewer reads everything else.
- Generate preliminary data. Even modest bench results — a breadboard measurement, a materials test, a small user study — separate you from proposals that are entirely prospective. Reviewers fund evidence over intention.
- Bring in credible expertise. A named consultant, clinical advisor, or university collaborator with relevant standing addresses the "can this team execute" question directly. If a research institution genuinely holds the core science, the STTR variant may fit better — see SBIR versus STTR and partnering with universities on product R&D.
- Write the milestones as pass/fail. "Demonstrate flow rate within plus or minus 5 percent across the operating range" reads as a plan. "Investigate flow characteristics" reads as a hope.
- Have a non-specialist read it. If a technical generalist cannot state your innovation and your customer after one read, revise.
- Apply again. Persistence is genuinely the highest-correlation behavior among funded companies, and the debrief tells you what to fix.
Should you hire a grant writer?
Sometimes. A good SBIR consultant knows agency formatting, review culture, and the compliance minefield, and can raise a proposal's floor substantially. What they cannot supply is technical insight or customer knowledge — a proposal written entirely by an outsider reads like one. The realistic model is a founder-written technical core plus professional help on structure, commercialization framing, and compliance. Arrangements and costs are discussed in should you hire a grant writer for an SBIR proposal. Be wary of anyone charging a success fee as a percentage of the award; agencies have views on that and so should you.
Weigh the cost of trying
A first SBIR proposal typically costs six to ten weeks of founder time, more when preliminary data has to be generated — real opportunity cost that competes with building the product and finding customers.
The math improves on the second and third attempt, because the market analysis, competitive table, team bios, and technical background carry forward. Companies that treat SBIR as an ongoing program rather than a lottery ticket have far better lifetime odds: they submit several proposals a year across agencies, and their hit rate compounds.
If the answer is that you cannot afford ten weeks right now, that is a legitimate answer, and the alternatives are compared in grants versus investors and mapped more broadly in grants for inventors. Non-dilutive money is excellent value per dollar and expensive in time; equity is the reverse.
Projects House supports founders on the technical side of an application — feasibility scoping, preliminary bench data, cost and manufacturability estimates, and a development plan a reviewer will believe. If you want the engineering half of your proposal to hold up under review, reach us through the contact form.