Complete Guide to SBIR/STTR Grants for Small Government Contractors

The federal government funds more than $4 billion annually in small business R&D through SBIR and STTR programs. That money goes to companies building things that didn't exist two years ago — technology in the pre-commercialization gap where venture capital won't go and traditional government contracting doesn't reach.

For small government contractors with a technical edge, SBIR/STTR is the most underutilized funding mechanism in the federal ecosystem. It requires no equity stake from investors, no government contracting past performance, and no existing agency relationship. The application process is competitive — roughly 10–25% depending on the agency — but it's purely merit-based. Your track record in the field and the quality of your technical proposal are the only things that matter.

This guide covers: what SBIR/STTR actually is, who qualifies, the three-phase funding structure, how to find active topics, what makes a proposal competitive, and why most applications get rejected — so yours doesn't.

What SBIR/STTR Is — and What It Isn't

SBIR (Small Business Innovation Research) and STTR (Small Business Technology Transfer) are federally-mandated R&D funding programs. Each year, the 11 agencies with the largest extramural R&D budgets must set aside a fixed percentage — currently 3.2% for SBIR and 0.45% for STTR — for small business awards. The agencies run their own programs, write their own topic areas, and review proposals internally.

Think of it this way: instead of a grant to fund business operations, it's a contract to develop a specific technology that has a credible path to commercial or government use. The government pays for R&D that benefits both the technology developer and the public. The company retains IP rights to what it builds.

The critical difference between SBIR and STTR:

• SBIR — the small business must perform the majority of the work. Phase I requires at least two-thirds (66%) of the work done in-house. Phase II requires at least half (50%). Research institution involvement is optional.
• STTR — designed to bridge small businesses and research institutions. Phase I requires the small business to perform at least 40% of the work and a research institution (university, federal lab, nonprofit R&D org) to perform at least 30%. STTR is the better fit if you have an existing research partnership and want to leverage it.

For most small government contractors, SBIR is the better starting point unless you have an established university or lab relationship. The eligibility requirements, topic structures, and review criteria are nearly identical between the two programs at most agencies.

SBIR/STTR eligibility is deliberately broad. The goal is to fund small businesses doing innovative R&D, not to restrict access to companies that already have government relationships. But there are clear rules, and violating them is the most common reason applications get disqualified before review.

Basic eligibility requirements

Who should NOT apply

Not every small business is a good SBIR/STTR candidate. The programs are designed for companies doing genuine R&D with a novel technical component. Service companies delivering existing solutions, businesses without technical founders or employees, and companies without IP rights to what they're proposing should not apply — the review panels can tell, and a weak Phase I submission disqualifies you from the same topic for 12 months at most agencies.

SBIR/STTR awards move through three phases. Each has a specific purpose, funding range, and completion requirement before the next phase is available.

Eleven agencies participate, but the funding is heavily concentrated. DoD and HHS (NIH) account for roughly 70% of total SBIR/STTR spending. The right agency for your technology depends on where it fits in the federal mission space.

Finding your agency topic

Each agency releases its topic areas through a solicitation — a public document that lists the specific technical areas they want to fund that cycle. Topics are typically released annually or semi-annually, with submission windows of 2–3 months.

The process:

1. Identify the agency whose mission aligns with your technology. Don't apply to NIH with a defense technology; don't apply to DoD with a healthcare innovation. The review panels evaluate relevance to agency mission, and misalignment is an immediate disqualifier in scoring.
2. Read the topic areas — not just the titles, but the specific subtopic descriptions. Each subtopic has a description of what the agency wants to fund, what they expect from proposals, and what commercialization path they anticipate. Write your proposal to the topic, not around it.
3. Register early on the agency's submission portal. System access, entity registration, and user account setup take time and can block last-minute submissions. Do this before the topic opens.
4. Check pre-release windows — DoD releases its topics in a pre-release period (typically 3–4 weeks before the solicitation opens) where companies can read but not submit. Use this window to filter for relevant topics and pre-draft your proposal. The full solicitation opens later; the pre-release gives you a head start.

SBIR/STTR proposals are scored by review panels — typically 3–5 experts in the relevant technical field. The review criteria are public, published in each agency solicitation, and consistent across cycles. Understanding what reviewers look for and structuring your proposal accordingly is the single highest-leverage activity in SBIR/STTR pursuit.

The four core evaluation criteria

1. Technical merit and approach (typically 30–40% of score)

Is the proposed technical approach sound, innovative, and likely to achieve the stated objectives? Reviewers want to see a clear problem statement, a credible technical approach with identified risks and mitigation strategies, and evidence that the team has the capability to execute. Vague claims about "cutting-edge AI" or "revolutionary technology" without specifics score poorly. Specific, detailed, technically grounded proposals score well.

2. qualifications of the team (typically 20–30% of score)

Does the team have the technical expertise, track record, and resources to perform the proposed R&D? For Phase I proposals, reviewers look for the principal investigator's publication record, prior R&D experience, and relevant domain expertise. Company size and prior awards matter here — a Phase I award on your record is a strong qualification signal for Phase II. If you don't have a PI with deep domain credentials, consider adding a subcontract or consultant with them.

3. commercialization potential (typically 20–30% of score)

Does the technology have a credible path to commercial application or follow-on government use? This is the most commonly under-developed section of Phase I proposals. Reviewers want to see a defined market, an understanding of who the customers are, a business model, and a sense of what the post-SBIR pathway looks like. Vague statements like "this technology has broad commercial applications" score poorly. Specific: "we have LOIs from two DoD program offices for integration into [specific system]" scores well.

4. budget and feasibility (typically 10–20% of score)

Is the budget realistic for the scope of work proposed? Reviewers check whether the scope matches the budget and whether the duration (6–12 months for Phase I) is realistic for the proposed work. Unrealistically low budgets signal that the team doesn't understand the technical requirements. Budgets that appear inflated signal that the team is padding. Accurate, well-reasoned budgets tied to specific technical tasks score best.

Common proposal weaknesses that kill scores

SBIR/STTR Phase I success rates range from roughly 10% at NIH to 20–25% at NSF and DoD sub-agencies. That sounds daunting, but it's worth understanding where applications fail so you can avoid the most common mistakes.

Where Phase I applications fail

The topic mismatch problem. A significant percentage of Phase I submissions are from companies whose technology doesn't fit any active topic. They apply anyway, hoping the technical merit of their proposal will carry them. It doesn't. Review panels score within topic areas — an excellent proposal outside the topic scores zero.

The weak commercialization narrative. NIH in particular has become increasingly strict about commercialization potential. A Phase II NIH proposal with no clear path to market or downstream application is getting harder to fund. The agency wants to see that public dollars produce public benefit — which means real commercial or healthcare outcomes, not "we plan to commercialize eventually."

The generic proposal problem. Many SBIR proposals are written to sound impressive rather than to address the specific topic sub-bullet. "Our company has a revolutionary AI platform" is not a Phase I proposal. "Our AI-based anomaly detection system, trained on [specific dataset], will achieve [specific metric] on [specific hardware constraint] within [specific timeframe]" is a Phase I proposal.

How to improve your probability

Get a Phase I before chasing Phase II. Phase II success rates are significantly higher for companies that have already received a Phase I award — even from a different topic. The Phase I award is a credential that signals technical credibility. If your first attempt doesn't result in a Phase I, apply to a different topic or agency. A failed first submission doesn't disqualify you from future cycles, but you should analyze your score debrief (most agencies provide them) before reapplying.

Align your commercialization story with agency priorities. DoD reviewers care about dual-use technology that has both commercial and defense application. NIH reviewers care about health outcomes. NASA reviewers care about space technology that also has commercial spinoffs. Frame your commercialization narrative for the audience, not for generic "market opportunity."

Build your technical team with topic-relevant credentials. PI credentials are a significant factor in scoring. If your PI is an expert in the technology but hasn't published in the relevant topic area, either find a co-PI with matching credentials or explicitly name consultants/subcontractors who provide that credibility. The proposal needs to demonstrate the team can do the work, not just that the technology is interesting.