How to Read Commercialization Evidence Without Overclaiming
The SBIR evidence problem is not that there are no metrics. It is that the available metrics answer different questions. A sale, a government procurement, an outside investment, a patent, a spin-off, and a Phase III record can all be relevant while describing different events in a technology’s life. This chapter compares those indicators, interprets the recent evidence on SBIR revenue reliance and multiple award winners, and then examines how the 2026 reauthorization tries to make transitions easier to support and observe. The practical result is a rule for matching the strength of a commercialization claim to the evidence actually available.
The metric must match the claim
A commercialization claim should begin by naming its outcome. “The technology generated a patent” is a claim about an inventive output. “The firm received an investment” is a claim about outside capital. “The product was purchased by a government agency” is a claim about a particular procurement relationship. “The technology was commercialized” is broader: it implies that the result crossed from development into a form of use, sale, licensing, or procurement that is sufficiently concrete and durable for the context being discussed.
The broader wording cannot be supported by simply adding more unrelated indicators. Five patents do not equal five customers. A new investor does not equal a purchase order. A spin-off does not demonstrate that the original technology is operating in a market. A company’s revenue share from SBIR awards does not reveal whether a particular funded device was sold. These indicators may be connected, but they should not be added together as though they were points on one universal scale.
Sales and procurement are usually closer to direct evidence of market use than a proposal, a commercialization plan, or a technical result. Even they need a boundary. A first sale establishes market contact for a product or application. A government contract establishes a procurement relationship under a defined mission and set of terms. Neither alone proves repeat demand, broad private adoption, or economics that can sustain the supplier. The reader needs to know what was sold, to whom, when, and whether the event continued.
Investment, customer testing, and production capacity are often transition signals. They show that another institution has entered the project’s path and that some uncertainty is being tested outside the original research setting. An investor may be financing years of additional work. A customer may be evaluating a prototype without agreeing to buy it. Greater production capacity may prepare a company to fill orders that have not yet arrived. Those signals can be important evidence of progress while remaining weaker than realized and durable use.
Patents and spin-offs measure yet another kind of result. A patent can protect an invention or help make licensing possible, but it does not show that the invention became a product. A spin-off can demonstrate organizational or entrepreneurial impact, but the new firm may still be developing its technology. These outcomes may be valuable even when no direct sale is visible. They should be described as patenting or organizational outcomes, not silently translated into commercialization.
Expertise, scientific results, and societal value are especially important in specialized public missions. GAO notes that SBIR-funded work can create expertise that later supports a different product, or produce value for research scientists and society that is not captured by sales.