Jill Aitoro breaks down the complicated problems facing the acquisition of new advanced technology from Silicon Valley – and what needs to change to fix it.

How should Pentagon leaders define a small business? On its surface, that seems easy to do, particularly if you acknowledge there are a couple of flavors – those with a modest workforce versus those with low revenue.

But if you caught Trae Stephens’ Twitter storm earlier this week, you would have discovered that the definition of a small business should require more thought. And that all the talk coming out of the Department of Defense adds up mostly to what he calls “innovation theater.”

“There is a huge difference between a low-growth/low-headcount business and a high-growth/product-focused tech startup,” the co-counder of Anduril tweeted. “We need to stop lumping these together.”

His comments came in response to Under Secretary of Defense for Acquisition and Sustainment Ellen Lord, who expressed concern recently with the ability of innovative small businesses to secure their systems. To Stephens point, small doesn’t mean less savvy. Not if they truly are innovative.

And the distinction for him goes even further. True innovative small businesses, he argued on Twitter and during a roundtable Defense News hosted in California last year, should demonstrate promise to become large businesses eventually. (During that roundtable several said the goal for the Pentagon should be to foster the next General Atomics).

Stephens pointed to the Air Force pitch day to prove his point that DoD is just not concerned with that kind of future growth potential. His analysis of the companies selected during that event, which essentially was like speed dating for contract awards, where companies could land funds through the Small Business Innovation Research program in as little as three minutes, showed that more than 50 percent had no venture funding of any kind. That’s a bad sign, Stephens said, since venture capitalists back the companies with the greatest potential for growth. In addition, most contracts were for cyber and services – not product development. Not to diminish the importance of cyber and services, but were those contracts breaking new ground? The work could probably be performed by any number of companies.

Also, there were three times as many companies from Northern Virginia than California and almost half were founded more than 10 years ago – not exactly the definition of a “Silicon Valley startup.” Now, innovation is not isolated to California, but it appears the winners here follow a pretty standard playbook for Pentagon small business contractors in both the scope of work and the makeup of the businesses.

What else? Contracts awarded were not recurring and averaged less than $100,000. Meaning, they will do little to ensure these companies have a trajectory to grow and really make some noise.

I reached out to Stephens to play devil’s advocate. Were venture-backed companies even at pitch day I asked him? Yes, he said. Without revealing company names, he said there were some strong venture-backed companies that did not get selected, “presumably because they didn’t play the inside game with hiring SBIR advisers and the like or weren’t known commodities in the source selection committees.”

He also said that “some non-zero part of this” is that high-growth companies are not targeting SBIR opportunities either. They don’t see them as beneficial, much in the same way that venture capital firms don’t.

I also questioned his theory that growth should always be the ultimate end game. Does the Pentagon always need another Palantir or SpaceX, or could 10 Silicon Valley developers in a room churning out algorithms to solve specific problems be a gamechanger? Wouldn’t that allow some of these startups to support the Pentagon without necessarily having to go all in on military as a market?

He agreed there’s a sweet spot in terms of contracts. If it’s too small, then little of consequence happens for the companies or the Pentagon. If it’s more than $10 million, then everything gets protested and killed by the primes, he said, pointing tocompanies like Capella Space, Improbable, and yes – Palantir as examples.

As for that “dual use” approach that so many are claiming is the future of technology, “It’s great that some commercially-focused businesses are interested in selling to the DoD, but there are also a lot of DoD-specific challenges that would be best served by new defense-focused tech companies.” Stephens himself is a partner at venture Founders Fund.

If this all sounds complicated, it doesn’t have to be. And Stephens isn’t trying to sound pessimistic or down on these grand efforts to unite the Pentagon and the tech community. (Nor am I.) But “innovation theater” won’t do it.

“The department needs to pick winners, with meaningful contracts, and give them access to the mission,” Stephens said. “There’s nothing specifically preventing them from doing that. They just don’t do it.”

And the Air Force’s pitch day was no exception.