13 answers small businesses should know about SBIR

No. "Owned by US citizens" is one path, not the rule. Four ownership structures qualify — the webinar's flat framing is the single most common reason an eligible firm self-disqualifies.

• The four paths (13 C.F.R. 121.702): majority-owned by US-citizen/permanent-resident individuals; or by other qualifying small businesses; or by multiple VCOCs / hedge / PE firms (at agencies that elected the authority); or by an ANC, NHO, or Indian Tribe.
• VC-backed is viable — but no single VC/hedge/PE firm may hold >50% (unless it is itself a qualifying small business), and you must register with SBA before submitting.
• Tribally owned firms qualify outright under the ANC/NHO/Tribe branch.
• Check the reg before you conclude you're out.

Headcount isn't the test. The standard is 500 employees together with all affiliates — and a controlling investor can aggregate an entire portfolio into your count.

• The real standard: 500 employees together with all affiliates (13 C.F.R. 121.103), measured by control — actual or potential, direct or indirect.
• The trap: a controlling PE/VC stake can pull the investor's whole portfolio into your count, pushing you over 500 without hiring anyone.
• SBIR relief: a shared investor alone doesn't create affiliation with the fund's other portfolio companies — but common control, board control, or identity of interest still can.
• Family firms: entities owned by close relatives that do business together are presumed affiliated (rebuttable by a "clear line of fracture").

Being foreign-born or holding a green card does not disqualify you. Undisclosed ties to a country of concern do — and at HHS/NIH there is no second chance that cycle.

• Countries of concern = exactly four: China (PRC), Russia, North Korea, Iran. Anything else is speculation until State designates it.
• What triggers it: foreign affiliations, investments, licensing/JVs, patent filings in a country of concern, or a malign foreign talent-recruitment program (42 U.S.C. 19237). Foreign LP money in your cap table is a common invisible trigger.
• Who discloses: every owner and "covered individual." The Foreign Disclosure Form is mandatory at Just-in-Time — miss it and you're ineligible.
• No pre-award cure at HHS: a mitigation plan is pre-submission defense, not a rescue. Clean up relationships up front; reapply next cycle if flagged.

A sole proprietorship is eligible — but form an entity. And no, you don't assign the patent: under Bayh-Dole you elect to retain title.

• Entity: a sole proprietor can apply, but form an LLC (or C-corp to raise) — liability, reviewer/investor expectations, and clean IP/data-rights ownership. Get IP counsel before you build.
• Title: under Bayh-Dole you elect to retain title — you are not forced to assign (the government can restrict this only in defined "exceptional circumstances").
• US manufacturing: the preference (35 U.S.C. 204) attaches to an exclusive license to use or sell in the US, and it's waivable — not a flat 51% mandate.
• SBIR data rights: a 20-year protected period from award (DoD codified it in DFARS, effective Jan 2025), then Government Purpose Rights. Structure Phase III carefully.

TABA (Technical & Business Assistance) is real money for commercialization help — up to $6,500 on Phase I and $50,000 on Phase II — and the April 2026 reauthorization opened the vendor market.

• What it funds: IP, market research, regulatory/manufacturing plans, customer discovery, sales help — plus new uses: cybersecurity and foreign-risk screening.
• Vendor freedom (new, April 2026): all 11 agencies must offer it, you choose your own vendor, and there is no approved-vendor list.
• On top or taken out? Agency-specific — Phase I is commonly additive, Phase II commonly deducted. Your solicitation is authoritative.
• Constraints: no profit/fee, not in G&A, and you must already be an awardee. Check the free innovation network (FAST, SBDC, APEX) first.

Three honest reasons to skip it — and three myths that talk people out of it for the wrong reasons.

• Real — opportunity cost: if your fastest path to revenue or a priced round beats a 6–10 month, ~15–25% cycle, skip it.
• Real — fit: SBIR funds innovation the agency needs, not a marginal product improvement.
• Real — compliance capacity: grant accounting, and for contracts FAR/DCAA, are real overhead.
• Myth — "SBIR is one thing": a DoD contract-style SBIR and an NSF/NIH open-call SBIR are different programs. Pick the right door.
• Myth — "it's dead": reauthorized through 2031, with new anti-mill proposal caps from FY2027.
• Myth — "worthless because it won't scale me": non-dilutive first, then raise from strength.

SBIR runs in three phases. The third one is a genuine business asset most applicants never leverage.

• Phase I — feasibility. Prove the concept works. Smaller award, typically ~6–12 months. This is the gate to Phase II.
• Phase II — development. Build the prototype / full R&D. Larger award, typically ~2 years, normally open only to Phase I awardees (some agencies allow "Direct-to-Phase-II").
• Phase III — commercialization. Not paid from SBIR set-aside funds; it's work that derives from your SBIR. The superpower: an agency can award Phase III follow-on work (including production) sole-source, without re-competition, and your SBIR data rights follow.
• Durations and amounts are set by each agency's solicitation — confirm there.

It depends on the agency — and that single fact decides your compliance burden and your whole mindset going in.

• Grant / cooperative-agreement agencies (NIH, NSF, DOE, USDA and others): largely investigator-initiated — you propose your idea. Lighter deliverable structure; grant accounting.
• Contract agencies (DoD/DoW, NASA, DHS and others): the government defines the topic/need; FAR-based, deliverable-driven, tighter accounting expectations (DCAA-style).
• Why it matters: applying to a contract agency with a grant mindset (or vice-versa) is the most common reason capable firms bounce off. Match the door to how you work.
• Talk to the agency's program manager before you write — they'll tell you if you're a fit.

STTR is SBIR's sibling for teams built around a research institution. If you're a university spinout, STTR is often the better fit.

• The partner: STTR requires a formal partnership with a nonprofit research institution (e.g. a university or FFRDC). SBIR does not.
• The work split: SBIR — the small business performs at least 2/3 (Phase I) / 1/2 (Phase II). STTR — the small business does ≥ 40% and the research institution ≥ 30%.
• The principal investigator: for SBIR the PI must be primarily employed (>50%) by the small business; for STTR the PI can sit at either the company or the research institution — the key flexibility for academic founders.
• Same phases, same money mechanics — the eligibility answers in Part 1 apply to both.

Registration lead time is what kills first-time applicants. Start weeks before the deadline, not days.

• SAM.gov — get your UEI and an active registration. This can take weeks; it is the most common cause of a missed deadline.
• SBIR Company Registry at SBIR.gov — obtain your SBC Control ID (required by every agency).
• Agency systems — eRA Commons (NIH), Research.gov (NSF), DSIP (DoD), plus Login.gov where required.
• Rule of thumb: begin registrations the day you decide to apply — not after you've written the proposal.

There are statutory "guideline" award amounts, but they move — and a number from a blog is exactly what gets firms in trouble.

• Guideline amounts are set by statute and periodically adjusted for inflation by SBA. Agencies may award above the guideline (up to a hard statutory ceiling, with SBA approval) or below it.
• The solicitation is authoritative. Do not rely on last year's number — confirm the current Phase I / Phase II amount in the specific solicitation you're answering.
• Fee/profit: grants generally carry no fee; contracts allow a reasonable fee. TABA money (Q5) carries none.
• Non-dilutive: the award buys no equity and takes no ownership — its real value is de-risking the round that does.

Competitive, not a lottery — and a rejection is often a first draft, not a verdict.

• Odds: Phase I success runs roughly 15–25% by agency (~17% overall). NIH is the most competitive and the highest-volume; NSF and DoD land in the same band.
• Timeline: submission to award is commonly ~4–8 months, agency-specific — NIH runs standard cycles; DoD opens topic windows through its BAAs.
• Rejected? Get the reviewer feedback (e.g. NIH summary statements), revise, and resubmit next cycle. A large share of awardees won on a resubmission.
• Budget the calendar, not just the proposal — plan for a cycle, maybe two.

No — and the mass-proposal "mills" are exactly what the new anti-mill caps target. Use the free network first, and know the warning signs.

• Free help first: your state's SBIR/FAST program, SBDCs, APEX Accelerators, and the agency's own program managers (talk to the PM before writing).
• Legit paid help can be worth it for a first proposal — and post-award, TABA (Q5) can fund commercialization assistance from a vendor you choose.
• "SBIR mills" that mass-produce proposals are being squeezed by the FY2027 proposal caps in the 2026 reauthorization. Volume is not a strategy.
• Walk away from anyone who guarantees a win, wants equity for proposal help, or offers to be your principal investigator.