The Other SBIR Problem: Phase III Has No Money and Nobody Tells You

Every defense startup founder knows about Phase I and Phase II. Smaller awards, proof-of-concept work, the beloved non-dilutive capital that lets you defer your Series A conversation for another eighteen months. The SBIR program is genuinely useful for early-stage deep tech companies. Nobody argues otherwise.

Photo by Mikhail Nilov on Pexels.

Phase III is where things go wrong. Quietly, methodically, and in ways that most founders don't see coming until they've already structured their company around an assumption that turns out to be fiction.

Here's what the program documentation says about Phase III: it's the commercialization phase, where agencies are supposed to buy the technology they've been funding. No dollar cap. No application required. Sole-source justification available. Sounds like the finish line.

What the documentation doesn't say is that Congress appropriates zero dollars specifically for Phase III. None. The funding has to come from a program office's existing budget, which means a program manager has to choose your technology over something already in their spending plan. That choice requires internal advocacy, acquisition support, and a contracting officer willing to exercise the sole-source authority before the fiscal year closes. Those three things rarely align without someone pushing hard from inside the building.

The result is a graveyard of Phase II graduates with working prototypes and no path to revenue.

Venture investors who haven't done this work often treat a Phase II award as validation that a program office wants to buy the thing. That inference is understandable and wrong. A Phase II award means a technical evaluator found the proposal compelling and a small business program office had budget to fund it. It says almost nothing about whether an operational program manager has discretionary funds available, or whether your technology fits into a current program of record, or whether the relevant contracting shop has capacity to execute a sole-source action before October 1st.

Those are separate questions. Most investors never ask them.

The companies that survive Phase III tend to share a few specific characteristics. They identified a program office sponsor before they submitted their Phase II proposal, not after award. They understand the difference between a technical sponsor (the person who wanted the research done) and a budget sponsor (the person who controls the dollars to buy a fielded system). And they treat the Phase II period of performance as a business development window, not just a technical execution window.

The mermaid below maps the funding logic most founders discover too late:

graph TD
    A[Phase II Award] --> B{Program Office Budget Available?}
    B -- Yes --> C[Sole-Source Phase III Contract]
    B -- No --> D{POM Cycle Timing?}
    D -- Next Year --> E[Bridge Financing Required]
    D -- 2+ Years Out --> F[Pivot or Die]
    C --> G[Program of Record Entry]
    E --> G

Notice the branch that says "Bridge Financing Required." That's where most of the carnage happens. A company finishes Phase II in Q2, the program office is interested but won't have available funds until the next fiscal year at the earliest, and the founder now needs six to twelve months of runway from somewhere. If they've been burning through non-dilutive capital without building a parallel venture relationship, that bridge conversation happens from a position of desperation rather than leverage. Valuations suffer. Founders give up more than they should.

Some of this is fixable at the portfolio level. Investors who specialize in defense tech should be asking Phase II companies hard questions about program office relationships before writing checks, not after. Specifically: who is the O-6 or SES-level champion inside the acquiring command? What POM cycle does the relevant program office operate on? Has the company had a conversation with a contracting officer about sole-source eligibility, or are they relying on the SBIR program office to broker that relationship?

If the answers are vague, the Phase II award is a technology asset without a buyer. That's a different investment than founders typically represent it to be.

One more thing worth saying plainly: the DoD has tried to fix this problem. The SBIR Reauthorization Act of 2022 included provisions intended to strengthen Phase III pathways, and some agencies have gotten better at internal coordination. Progress is real but uneven. The Air Force Research Laboratory handles Phase III differently than the Army's contracting ecosystem, and neither operates the way the statutory language suggests they should.

Uniform execution remains a project for future congressional sessions. In the meantime, the gap is real, and companies that plan around the ideal version of the program instead of the operational version tend to run out of money proving a point nobody asked them to make.