Your team decided that writing federal proposals is not worth the effort. That was a real calculation and most of it holds. Here is the part that does not, and where it stops holding is where the money is.
In its report on the valley of death, the Defense Innovation Board asked startups and their investors what was wrong with selling to the Department of Defense. One answer came back so plainly that DoD published it verbatim:
“Writing large, complex proposals is not worthwhile due to the low probability of winning and lack of constructive feedback. Pitch decks, pitch events, with higher win probability and less preparatory work are preferred.”
Read that as a business decision, because that is what it is. Somebody weighed effort against probability, noticed that losing teaches you nothing because no feedback ever comes back, and concluded the proposal was a bad investment. Pitch events cost less and convert better. That reasoning is sound.
It is a claim about return on effort, so it lives or dies on where the money actually is. GAO looked at that. Other Transaction Agreements are the flexible instrument startups are pointed at precisely because they avoid heavy proposals.
Eight dollars go in as prototype for every dollar that comes out as production. The pitch-event path is genuinely good at getting you into the sixteen billion. It has never once put anybody into the two.
The reasonable assumption is that a prototype that performs gets noticed, and somebody inside the program office carries it forward. GAO checked whether that happens.
“DOD does not know the extent to which these prototype OTAs directly resulted in production awards.”
The department has no reliable method for tracking which production contracts came out of prototype work. Not a backlog, not a delay. There is no mechanism. So there is no file connecting what you built to what they should buy, and no analyst assembling one. If that argument gets made, you make it, in writing, every time. A pitch deck does not do that work and was never designed to.
If your company has ever held an SBIR Phase I or II, work that derives from it can be awarded sole source, at any dollar value, with no expiration. No further competition. Most companies do not know this. That is not the interesting part. The interesting part is from the same DIB report:
“Contracting officers and program managers are not aware that Phase III grants permission to award a sole-source contract for products matured under Phase I and II contracts.”
The government side does not reliably know either. This is not leverage you hold over a contracting officer. It is something you bring them, with the authority cited, early enough that a competitive requirement has not already been drafted. Once the solicitation exists, the window has usually closed. Confirm the current citation with your contracting officer or counsel before relying on it. Then raise it anyway.
Federal procurement runs about $700 billion a year. State, local and education procurement runs about $1.5 trillion, spread across more than ninety thousand buying entities. Roughly twice the size, a fraction of the attention, because there is no press release in a county contract.
It also has a structure federal does not. A cooperative purchasing award, through a vehicle like NASPO ValuePoint, Sourcewell or OMNIA Partners, is competitively awarded once and then usable by tens of thousands of agencies without any of them running their own competition. Sourcewell alone moved $5.85 billion in contract sales last year, up more than fifteen percent.
Getting on one is a long, compliance-heavy RFP response. It is exactly the artifact your team decided was not worth writing. It is also the one document where the math you ran comes out the other way. One response, correctly done, and you stop selling to agencies one at a time for the life of the contract.
The reason technically excellent companies lose is rarely technical. Section L tells you how to submit and Section M tells you how you will be scored. They frequently contradict each other, and where they do, you satisfy both. A page limit, an answer filed in the wrong volume, a missing cross-reference: each ends the evaluation before anyone reads your approach. One company lost on a graphic title page and GAO told them the agency was right.
Then there is the part specific to companies like yours. A proposal is a binding offer. Startup vocabulary is built on “10x faster” and “guaranteed uptime” and “fully autonomous,” and that language walks straight out of the deck into a contract you will perform at your own cost for years. The defensible version is almost never to promise less. It is to state the conditions.
Past the prototype, the document is not overhead on the sale. It is the sale. If you want to test the last point on something real, the commitment auditor runs in your browser, uploads nothing, and asks for no email.
One live bid, read in 72 hours · $7,500 →