An SBIR Award Is a Beginning, Not a Market Result
The subject of this series is the Small Business Innovation Research (SBIR) program and the path its small-business awardees must travel from federally funded research to real-world use. The central question is simple but easy to answer badly: what does an SBIR award actually establish, and what would still need to happen before anyone could reasonably say that the technology was commercialized? The useful judgment is not to dismiss public R&D funding, but to separate evidence of funded work from evidence of customers, production, procurement, and sustained use.
That distinction matters because the word commercialization often appears close to the word award. A company may announce that it has won federal support for an innovative material, sensor, medical device, or software system. Readers can then slide, almost without noticing, from “the government selected this project for funding” to “the technology is on its way to market,” and finally to “the technology has succeeded commercially.” Those are different claims. Each requires a different kind of evidence.
The inference hidden inside an award
An SBIR award is meaningful. It says that a participating federal agency has chosen to support a defined research and development effort under a particular set of program rules. That decision can reflect technical merit, relevance to an agency mission, and the promise of an innovation. It can give a small firm money, time, expertise, and a credible signal to other funders. For a young company working on a difficult technology, those are not minor benefits.
It is also important to identify what the award decision is for. A federal agency may be trying to solve a research or mission problem, not certifying that a private market is ready for a product. A technology can be highly relevant to a government need while having only a few potential buyers. It can be scientifically valuable while requiring a manufacturing process that does not yet exist. It can attract public support because private investors consider the timeline too long or the technical risk too high. The agency’s decision and the market’s later decision may be reasonable responses to different questions.
But an award is still a funding event. It does not, on its own, show that a prototype can be manufactured at the required quality, that a customer will pay for it, that a government buyer will procure it, or that sales will cover the cost of continued operation. Nor does the award tell us how long the remaining work will take. A laboratory result can be technically sound and still need years of testing, redesign, certification, scale-up, integration, or customer education.
The distinction is important in both directions. If a sales record is not yet visible, that absence does not prove that an SBIR-funded project has no value. The work may produce a scientific capability, a patent, expertise that supports a different product, or a technology serving a specialized public mission rather than a large commercial market. Conversely, the existence of an award does not prove that any of those outcomes occurred. Good analysis must avoid both premature celebration and premature dismissal.
This is why the right unit of attention is not the award by itself. It is the relationship among the awardee, the agency, the funded technology, outside capital, potential customers, and the later source of operating revenue or procurement. The same award can be an important step for one project and an endpoint for another—not because the label changes, but because the evidence after the award does.
What SBIR actually funds
SBIR was established in 1982 to bring small innovative businesses into federally funded R&D and to encourage technological innovation, federal research capability, broad participation, and private-sector commercialization. The Small Business Administration provides program oversight and guidance, while participating agencies make the awards. The scale is substantial: according to the Government Accountability Office (GAO), agencies made more than 5,000 SBIR awards worth nearly $4 billion in fiscal year 2023.
That scale helps explain why a single award announcement can be rhetorically powerful while remaining analytically incomplete. The program is a large public R&D system made up of many projects, agencies, technologies, and time horizons. An aggregate award total tells us that public resources are being committed. It does not tell us how many projects are at feasibility, prototype, customer-testing, production, or commercial-operation stages. A serious claim must identify the project’s place in that larger system.
The program’s three-phase architecture is the first protection against overclaiming. In Phase I, a small business investigates whether an idea is technically feasible and has commercial potential. In Phase II, the business continues the research and development, which may include building or improving a prototype. Those stages can produce valuable technical evidence, but they are not the same as a customer order or a functioning business model.
Phase III is the point at which the program describes commercialization as the objective. Yet Phase III does not receive additional funding through the ordinary SBIR Phase I and Phase II budgets. It relies on other sources, such as a non-SBIR agency program, a government contract, private capital, licensing, or sales. That boundary is easy to overlook. A technology can move through federally funded research and still face its most consequential test after the award-funded work ends.
The phases should therefore be read as bounded statements. “Phase I award” means that a feasibility question has been selected for investigation. “Phase II award” means that an agency is supporting continued development after that earlier stage. “Phase III” points toward commercialization through a different funding and contracting environment. None of those labels alone answers whether the technology was produced reliably, adopted by a customer, or economically sustainable.
The sequence is also a pathway, not a conveyor belt. A project may need more technical work, wait for outside funding, change its target application, or pursue a different government program. A firm can make progress on one dimension while remaining blocked on another. The useful question is not simply whether the project moved to a later label, but what new uncertainty that movement resolved and which uncertainties remain.
The missing handoff
The most difficult step is often the handoff between R&D and use. SBIR support can end before a promising technology has attracted a private investor, a non-SBIR government contract, or a customer willing to pay for the next stage. The firm may need to finish engineering, validate performance in a real operating environment, build production capacity, protect intellectual property, or adapt the product to a particular buyer. These activities consume time and money, but their payoff may arrive much later.
The handoff has at least three dimensions. Technical work must become reliable outside the original research setting. Someone must provide the capital or contract that pays for the next round of work. And a customer or public buyer must have a reason to change behavior, accept risk, integrate the technology, and keep using it. An award can help with the first dimension without settling the second or third. Even a promising prototype may not yet have the documentation, production rate, price, or support model that a buyer requires.
GAO’s study of the Commercialization Assistance Pilot Program describes this vulnerability as a “valley of death.” The phrase is not a claim that every awardee fails at the same point. It names a recurring exposure: a small firm can make real progress and still lose the runway needed to carry that progress into the market. The problem is especially visible in deep-tech work, where scientific and engineering challenges take years to resolve, and in niche markets, where there may be only a few potential customers.
That context changes how an outcome should be described. A company that uses additional funding to scale production has advanced beyond a laboratory result, but it has not necessarily achieved durable sales. A company that receives an outside investment has gained a transition resource, but the investment is not proof that customers will adopt the product. Initial sales are stronger evidence of market contact than a patent, while repeat sales, a production system that can meet demand, or a sustained procurement relationship tell us more about commercial durability. The evidence becomes stronger by answering more of the questions that an award leaves open.
The policy surrounding SBIR reflects this problem. CAPP was created to give some eligible firms an additional Phase II continuation, potentially bridging the period before investors, sales, or other government funding arrive. The 2026 reauthorization extends SBIR, STTR, and CAPP-related authorities through 2031 while adding new transition-oriented tools, including strategic breakthrough funding, Phase III education, and more detailed data about phase designations and follow-on contracts. Those changes are important because they treat commercialization as a transition that needs support and measurement. They do not, however, turn an award into proof that the transition succeeded.
A more disciplined question
When an award announcement makes a commercialization claim, begin with three questions. What exactly was funded? What happened after the funded stage? What definition of commercialization is being used?
The first question identifies the phase, the technology, the awardee, and the agency relationship. The second looks for evidence of technical completion, prototyping, testing with customers, new investment, production, sales, procurement, licensing, or sustained use. The third prevents unlike outcomes from being quietly treated as equivalent. A patent may demonstrate an inventive output; a government contract may show procurement; a spin-off may show organizational impact. Each can matter without being interchangeable with commercial adoption.
The timing of the evidence matters as much as its type. A report written while R&D is still underway cannot establish a later sales outcome. A new investment may be a strong sign that someone sees potential, but it may also fund years of additional development. A first customer may validate a use case without showing that the firm can serve a broader market. These are not reasons to discount partial progress. They are reasons to describe it at the level the evidence supports.
This series follows that distinction through SBIR’s phase structure, the funding and capability gap, CAPP’s design and limited implementation, the evidence about awardees and unsuccessful applicants, the problem of measuring commercialization, and the current statutory response. The aim is a practical mental model: an award is a reason to investigate the next transition, not permission to assume its result.
The key lesson is modest but consequential. SBIR can fund the work that makes commercialization possible, and it can sometimes help a small business reach the next stage. But the award itself establishes neither a finished product nor a market result. To understand what happened, we have to follow the transition.