Key Facts
- Three different application processes — DOD (topic-driven, 20-page technical volume), NIH (open, 6-page Research Strategy), NSF (Project Pitch first, then 15-page proposal). Each is evaluated by different reviewer types with different priorities.
- Commercialization potential is a scored criterion at all agencies. It is not filler — reviewers have explicit commercialization scoring criteria and will mark down proposals that treat it as an afterthought.
- Phase I win rates: NIH ~14–18%, DOD ~10–20% per topic, NSF ~15–20% of full proposals funded. NSF Project Pitch has a ~25% invitation rate — you can fail before writing the full application.
- SAM.gov registration required before applying — can take 7–10 business days to activate. Do not leave this to the week before the deadline.
- Phase I does not guarantee Phase II. Phase II is a separate competitive application at most agencies. Phase I completion is necessary but not sufficient for Phase II funding.
Summary
SBIR Phase I awards fund feasibility research for U.S. small businesses across 11 federal agencies, with Phase I award sizes ranging from $150,000 (NASA) to $314,363 (NIH) over 6–12 months. The application process differs significantly by agency. DOD responds to specific technology topics in competitive solicitations. NIH accepts open submissions year-round tied to standard receipt dates. NSF screens all applicants through a 2-page Project Pitch before inviting full proposals. In every case, the core scoring criteria are the same: technical merit, commercial potential, and team qualifications. The proposals that fail — consistently, across all agencies — do so because they prioritize technical depth at the expense of commercialization clarity, or because they address the wrong level of maturity for Phase I.
The Fundamental Mistake: Writing a Research Grant Instead of an R&D Proposal
The SBIR program is categorically different from academic grants like the NIH R01. Academic grants fund curiosity-driven research aimed at generating knowledge. SBIR funds technology development aimed at creating a commercial product. This distinction is not just rhetorical — it affects every section of the proposal, and reviewers are trained to distinguish between them.
When an SBIR Phase I proposal reads like an academic research proposal — heavy on literature review, focused on understanding a phenomenon rather than solving a problem, vague on what the deliverable is — reviewers mark it down on commercialization, and often on innovation as well. The mental model to hold when writing: you are a small company that has a technical idea for a product. The government is paying you to determine whether that idea is feasible. At the end of Phase I, you should have a prototype, validated data, or a clearly tested hypothesis that tells you whether Phase II is worth pursuing. Your Phase I proposal should read like a plan for that feasibility test — not a research agenda.
DOD SBIR: Topics, Technical Volume, and What the Government Actually Wants
DOD SBIR works differently from every other agency. Each DOD solicitation (published three times per year for most components) lists specific technology topic areas — defined capability gaps that a military program office wants addressed. You are not proposing your own research question. You are proposing an approach to solving the topic's stated problem.
This has an important implication: topic fit is the first scoring gate. If a reviewer does not believe your technology addresses the topic, the rest of the proposal doesn't matter. Read the topic statement carefully, including the subtopics, focus areas, and references. Contact the technical point of contact (TPOC) listed in the topic — this is expected and encouraged. A 15-minute call with the TPOC before submitting can confirm whether your approach is in scope and occasionally surface which aspect of the topic the program office most wants addressed.
DOD Phase I technical volumes are typically 20 pages, divided into: technical approach, management plan, commercial potential, and sometimes a prior work or company capabilities section. The technical approach should describe: what you will do in Phase I (not Phase II), specific tasks and deliverables, a realistic timeline, and success criteria. Reviewers are specifically looking for whether Phase I has a clear, bounded scope. A common scoring problem: Phase I proposals that describe Phase II work. If your technical approach contains objectives that would realistically take 24 months and $1.5M, reviewers will identify the mismatch and score the proposal lower on feasibility.
The commercial potential section is not an afterthought. DOD SBIR exists to build the defense industrial base — the government wants technologies that will eventually be sold to the military or dual-use markets. A credible commercialization narrative includes a specific customer (the program office that issued the topic, if relevant, or a named prime contractor), an estimated market size with a source, and a realistic Phase III transition plan. Vague statements like "there is a large market for advanced sensors" score poorly.
NIH SBIR Phase I: The 6-Page Research Strategy
NIH SBIR/STTR Phase I applications are submitted through the standard NIH eRA Commons system and reviewed by dedicated study sections — typically special emphasis panels convened specifically for SBIR/STTR applications. Unlike DOD, there are no topics. You propose any research question relevant to NIH's mission in health and biomedical science.
Since 2010, NIH SBIR Phase I Research Strategy is limited to 6 pages — Significance, Innovation, Approach, combined. This is a dramatic constraint. Six pages forces precision: there is no room for background that reviewers already know, for tangential references, or for detailed methodology that belongs in the supplementary resources section. The standard NIH R01 advice applies here with more intensity: if a sentence does not directly advance your argument, cut it.
The Specific Aims page — a separate 1-page section — carries the same weight as in an R01. Reviewers read it first and form their overall impression before opening the Research Strategy. Your aims should be testable within the 6-month Phase I period, should directly establish feasibility for a Phase II product, and should produce clear go/no-go decision criteria. "If Aim 1 demonstrates X threshold of activity in our cell model, we will proceed to Aim 2 in vivo validation" is a Phase II-appropriate framing. Phase I aims should themselves be the decision criteria.
NIH SBIR reviewers score on the same five criteria as standard NIH grants — plus two additional criteria specific to SBIR: Commercialization Potential and Qualifications of the Principal Investigator. Commercialization here means potential to lead to a company-built product, service, or FDA-approved therapeutic. NIH SBIR is not for academic researchers who want to generate knowledge. If your Phase I proposal doesn't explain how positive results would lead to a Phase II product and eventually a marketable therapeutic or device, reviewers will score Commercialization below average regardless of the science.
GrantMetric Analysis
- NSF's Project Pitch is the most underrated gatekeeping mechanism in federal grants. Before NSF will accept a full SBIR proposal, you submit a 2-page Project Pitch describing the technology, the market, and the team. NSF program directors — not an external panel — review these pitches and decide within 5 weeks whether to invite a full proposal. Roughly 25% of pitches receive invitations. The pitch is rejected without feedback; there's no "pink sheet" equivalent. What NSF is looking for at the pitch stage: evidence that this is a technology startup (not an academic lab looking for bridge funding), a specific unmet customer need with a named target market, and technical differentiation from existing solutions. If your pitch reads like a research proposal or a business plan without technical specifics, it will not be invited. Write it as a distillation of both.
- The company qualifications section is where solo founders hurt themselves most. All SBIR proposals require a section on team qualifications. Reviewers evaluate whether the people described can actually execute the work. A single PI with a strong technical background but no business or regulatory experience will score lower on team qualifications at NIH than the same PI paired with an industry co-investigator with startup or commercialization experience. You don't need a full team — Phase I is small — but you do need to demonstrate that the key skills exist somewhere in your company or consultant network. A named business development advisor, a scientific advisory board member, or a letter from a potential customer all strengthen the commercialization picture and implicitly address team capability concerns.
- The time to start thinking about Phase II is when you write Phase I. The most competitive Phase II applications come from companies that designed Phase I as a deliberate setup. The Phase I success criteria should map directly onto Phase II hypotheses — if your Phase I shows a 3-fold improvement in sensitivity over baseline, your Phase II aims build on that specific threshold. Reviewers evaluating Phase II know the Phase I data; they're scoring whether the Phase II plan represents a logical, defensible next step. Companies that treat Phase I as a standalone project and write Phase II as a new proposal from scratch consistently score lower than companies that designed the two phases as a deliberate program from the beginning.
Commercialization Section: The Part Everyone Writes Last and Should Write First
Every SBIR Phase I application requires a commercialization section or narrative. At DOD, it's typically 1–3 pages. At NIH, it's part of the 6-page Research Strategy. At NSF, it's weighted heavily in the Project Pitch and explicitly evaluated in the full proposal. Across all agencies, it is the section most frequently cited in reviewer critiques as weak or insufficient.
A fundable commercialization section addresses three things: the specific problem your technology solves for a specific customer, the market size with a defensible estimate and source, and the pathway from Phase I results to a product that can be sold. On the customer specificity point: "government agencies" is not a customer. "The Army's Program Executive Office for Soldier — specifically PEO Soldier's body armor program, which has an annual budget of approximately $200M and has published requirements for 20% weight reduction at equivalent protection levels" is a customer. The more specific the customer description, the more credible the commercialization argument.
Market size estimates are often where credibility breaks down. Avoid top-down estimates like "the global sensor market is $45B and we expect to capture 1% within 5 years." This type of market sizing is immediately recognizable as arbitrary. Bottom-up estimates — "there are approximately 8,000 trauma bays in U.S. hospitals; at $12,000 per unit with a 3-year replacement cycle, the addressable U.S. market is approximately $32M per year" — are harder to produce but dramatically more persuasive. Use government procurement databases (USASpending.gov, SAM.gov contract awards) to anchor market estimates with real data.
Budget: Phase I Doesn't Need to Be Minimalist
Many first-time SBIR applicants deliberately underbudget Phase I to appear conservative or to improve competitiveness. This is a mistake on both counts. Reviewers do not score budget size — they score whether the budget is adequate to accomplish the proposed work. An underfunded Phase I that can't realistically complete its aims scores lower on feasibility than a properly funded one.
Budget key points by agency: DOD Phase I is capped at $275,000 for 6 months (some components allow extensions to 12 months at the same dollar level). NIH Phase I is up to $314,363 in direct costs for 6 months; this is a limit, not a guideline — you can request less. NSF Phase I is up to $275,000 for 6 months. These are direct cost caps; indirect costs (F&A) are paid on top of these amounts at the rate negotiated between your company and the government. Most SBIR-stage startups without a negotiated rate use the NIH de minimis rate (26% on modified total direct costs) or the DOD limited rate.
SBIR Phase I Application Checklist
- Register in SAM.gov at least 3 weeks early — activation takes 7–10 business days and is required before any federal award
- Verify SBIR eligibility: ≤500 employees, ≥51% U.S. citizen/permanent resident owned, PI primarily employed at the company at time of award
- For DOD: identify target topics at sbir.defense.gov; contact the TPOC to confirm your approach fits the topic scope before writing
- For NSF: submit Project Pitch first at seedfund.nsf.gov — do not spend time on a full proposal until invited
- Write the commercialization section before the technical section — this clarifies what Phase I needs to prove
- Include specific go/no-go decision criteria in the technical approach — what result at the end of Phase I tells you whether to pursue Phase II?
- After winning Phase I: immediately check your state economic development agency for SBIR matching grants (~20 states offer them)