SBIR Phase I is one of the few federal programs that hands research money directly to small businesses with no strings attached to past performance. You don't need prior government contracts. You don't need a cleared facility. You need a technically credible idea, a qualified principal investigator, and a proposal that follows the structure reviewers expect.
The program distributes roughly $3–4 billion per year across 11 federal agencies. DoD alone runs multiple solicitation windows, and agencies like NIH, NSF, DOE, and NASA each publish their own topic lists. Phase I awards are typically $150K–$300K for 6–12 months of feasibility work. Phase II, if you win it, ranges from $750K to $2 million.
This guide covers what goes into a competitive Phase I proposal, how agencies score it, and the specific mistakes that sink first-time applicants.
What SBIR actually is (and isn't)
SBIR stands for Small Business Innovation Research. Congress created it in 1982 under the Small Business Innovation Development Act. Each agency with an extramural R&D budget above $100 million must set aside 3.2% of that budget for SBIR awards.
Mechanically, Phase I is a research contract, not a traditional procurement. You're not promising a deliverable product. You're promising to investigate whether your technical approach is feasible. The government pays you to find out if the idea works, and the IP generally stays with your company under the Bayh-Dole framework.
Eligibility: your company must be for-profit, majority-owned by U.S. citizens or permanent residents, have fewer than 500 employees, and have the principal investigator employed primarily by the firm at the time of award. That last point trips up academic researchers who try to spin out a company while keeping their university position as their primary employment.
The 11-section structure
Most agencies follow a standard proposal structure derived from SBA guidance. The exact section names vary by agency, but the underlying content is consistent. Here's what you need to cover:
1. Cover page and abstract. The abstract is often the only part a program manager reads before deciding whether to send a proposal to reviewers. Keep it under 200 words. State the problem, your approach, and the commercial outcome.
2. Identification and significance of the problem. Define the technical problem precisely. Use data. If you're proposing a faster corrosion sensor for aircraft, cite the current inspection intervals and failure rates, not general statements about corrosion being bad.
3. Phase I technical objectives. List specific, testable hypotheses. These are not project milestones. Each objective should be falsifiable: you should be able to say at the end of the period whether you achieved it.
4. Phase I work plan. Break down the technical tasks, assign time estimates, and identify who does what. Reviewers check that your team has the capacity to execute in the timeline. A 6-month Phase I with a single part-time researcher raises flags.
5. Related R&D. Describe what others have tried and why your approach is different. This is where you demonstrate technical literacy. Cite peer-reviewed literature, patents, and prior government-funded work. Agencies can check SBIRTECH and Research.gov to see if similar work has already been funded.
6. Relationship to future research. Explain how Phase I results feed into Phase II. Agencies want to fund a program, not an isolated experiment. The cleaner the logic chain from Phase I feasibility to Phase II development, the better.
7. Commercialization plan. This section carries more weight than most first-timers expect. See the dedicated section below.
8. Key personnel and bibliography. The principal investigator's qualifications are evaluated directly. List relevant publications, patents, and prior funded research. If your PI lacks a strong CV, pair them with a technical advisor with credentials in the field.
9. Facilities and equipment. Describe what your company actually has access to. If you're doing lab work, identify the specific equipment and whether it's owned, leased, or accessible through a university partnership. Vague references to "state-of-the-art facilities" get discounted.
10. Subcontractors and consultants. You may use subcontractors, but SBIR rules require that your company perform at least 2/3 of the work in Phase I. Document any planned subcontracts with clear scope and dollar amounts.
11. Prior SBIR/STTR work. If this is your first award, say so clearly. If you have prior awards, the agency will check your Phase II conversion rate and your commercialization outcomes. A history of awards with no Phase II conversions is a negative signal.
How agencies score Phase I proposals
NSF uses a two-criterion framework: intellectual merit and broader impacts. NIH uses significance, innovation, approach, investigator, and environment. DoD agencies vary by component, but most use some version of technical merit, commercialization potential, and team qualifications.
In practice, almost every scoring rubric comes back to three questions. Is the technical approach scientifically sound? Does the team have the skills to execute it? Is there a credible path to a commercial product or government application?
Phase I scores are assigned by ad hoc reviewers, usually 3–4 people with subject-matter expertise. These are not procurement officers. They're researchers and engineers. Write for a technically sophisticated reader who has no patience for marketing language.
The commercialization section is not optional
First-time applicants treat the commercialization section as an afterthought. That's a mistake. DoD program managers in particular will score a proposal down significantly if the commercialization plan reads like a generic market analysis.
A credible commercialization section does three things. It identifies a specific customer segment with a quantified market need. It describes the regulatory or procurement pathway to reach that customer. It names potential partners, prime contractors, or licensees where possible.
For a DoD-focused proposal, this might mean identifying a specific program of record that would benefit from your technology and describing the transition pathway. For a commercial application, it means stating your go-to-market model and identifying whether you're selling direct, through a distributor, or licensing.
The SBA's SBIR.gov site maintains a database of Phase II awardees with their commercialization outcomes. Reviewing companies in your technical space that went on to Phase II and beyond gives you a template for what plausible commercialization narratives look like.
Principal investigator qualifications
The PI does not need to be a PhD. SBIR rules say the PI must be primarily employed by your company during the period of performance and must devote meaningful effort to the project. What matters to reviewers is whether the PI's background matches the technical scope of the work.
If your proposal is about developing a machine learning algorithm for materials characterization, your PI should have documented experience in machine learning or materials science, ideally both. Publications, patents, prior funded research, and relevant commercial work all count.
If your team's technical credentials are thin for the specific topic, consider hiring a part-time technical consultant with a strong CV and listing them as a senior/key person. Their credentials can compensate, but the PI must still be your employee and must be doing real work.
Three action steps
First: pick the right solicitation topic. Go to SBIR.gov and filter by agency and technology area. Read the topic description carefully. Agencies write topics around specific problems they want solved. If your technology doesn't map cleanly to the stated technical objectives, don't apply. The mismatch will show in reviewer scores.
Second: request a pre-submission meeting with the topic technical point of contact (TPOC). Most agencies list a TPOC for each topic. A 30-minute call before you write confirms you understand the scope and signals genuine engagement. Program managers notice which companies bother to make contact.
Third: read at least 10 funded Phase I abstracts in your technology area on SBIR.gov. The abstracts are public. They show you the framing, the vocabulary, and the scope that agencies have already funded. If your proposal uses the same framing and is technically distinct, you're starting from a defensible position.
