The Pentagon Doesn’t Need More Startups: It Needs the Right Startups

A Signal Problem, Not a Startup Problem

I’ve seen it first-hand: the Department of War doesn’t have a shortage of startups clamoring for its attention. What it has is a signal problem. Amid all the noise, flashy pitch decks, endless Small Business Innovation Research (SBIR) proposals, and hyped press releases, a few signals of true value break through. We don’t need more noise in the form of:

• Slideshows instead of software. Slick PowerPoints are not solutions.
• SBIR pitch decks with no follow-through. Too many teams chase Phase I/II grants and never field a product.
• Hype without a prototype. Grand promises mean nothing without working tech to back them up.

What the Pentagon does need are the quiet professionals; the builders who aren’t showboating on the conference circuit because they’re busy solving real problems. These are the folks who actually visit the front lines, who understand the concept of operations (CONOPs) for a mission, and who ship working code or hardware fast instead of stalling in R&D. The best ones don’t waste time on vanity pitches; the best ones don’t pitch, they solve.

Noise vs. Substance in National Security Innovation

Spend enough time on national security, and you develop a nonsense filter. I’ve sat through dozens of startup pitches promising AI-enabled magic or revolutionary widgets. Most went nowhere. It’s not because the technology was impossible, but because the approach was all noise and no substance. Consider the data: the Pentagon’s top 20 SBIR award winners, so-called “SBIR mills,” consumed $3.4 billion in Phase I and II contracts, often producing little more than elaborate research reports. These firms have perfected the art of winning government funding while delivering minimal operational value to warfighters. They churn out polished proposals that impress review boards more than they solve problems. In other words, they excel at noise.

The result of this noise-heavy system? Promising capabilities get stuck in “pilot purgatory,” never moving beyond staged demos. Cultural and bureaucratic inertia inside DoW means even great tech can languish for years without a real contract, stranded in endless trials. The statistics are sobering: only ~5% of companies make it from SBIR Phase II to Phase III, i.e., from prototype to an actual program of record. That means 95% of taxpayer-funded prototype projects never transition to fielded capability, a huge waste of time and money. It’s no wonder many innovators give up. The signal (the truly impactful solutions) is drowned out by a tidal wave of well-intentioned pilots, paperwork, and PowerPoints that never deliver.

As a national security operator turned startup founder, I find this situation untenable. On the operator side, I’ve experienced the frustration of wading through vendor hype while lacking the tools I actually need. On the startup side, I’ve had to fight through a labyrinth of contracting just to put a proven solution in a soldier’s hands. The core problem isn’t a lack of ingenuity or patriotism; it’s a system that too often rewards sizzle over steak. When five-year acquisition timelines collide with 18-month tech cycles, guess what breaks? The acquisition system, and by extension, the warfighter’s trust. Private investors see this dysfunction and many shy away, perceiving defense as too unpredictable and slow to yield returns. In short, the noise is not just an annoyance; it’s actively driving away talent and capital that the Pentagon desperately needs.

Lessons from the Field: Quiet Builders Over Hype

So what does substance look like? In my experience, the startups that actually make a difference share a few traits. Call them “quiet builders” if you will. They’re often not the ones headlining tech conferences or boasting on social media. Instead, they’re busy doing three things:

• Getting boots on the ground (Visit the AOR). The best defense startups physically go to the Area of Responsibility (AOR), whether that’s an airfield in the Pacific or a remote ops center in the Middle East, and learn first-hand what troops and commanders face. There’s no substitute for seeing the bandwidth constraints, the dust and heat, the split-second decisions in person. If your product is meant for a platoon leader in a conflict zone, your engineers need to eat some dust there. I’ve seen startup CEOs show up at forward operating bases, not for a photo op, but to observe and iterate. Those are the teams that build tech that actually works when it counts. (Conversely, teams that never leave the comfort of HQ tend to build solutions that crumble outside a lab.)
• Knowing the mission (Understand the CONOPs). Quiet builders do their homework on the Concept of Operations. They deeply understand how the military user will deploy their tool in the real world. They learn the workflows, chain of command, and tactics, and they respect them. This means a drone startup, for example, grasps how an intelligence ISR platoon conducts surveillance missions, or a software team understands the steps a targeting officer takes to call in fire support. By knowing the CONOPs, these companies avoid the common pitfall of pushing a neat piece of tech that doesn’t fit into any actual mission thread. They craft their product to slot into existing workflows (or, if it’s truly revolutionary, they work with operators to adjust the workflow). This is the polar opposite of the hype companies that expect the military to invent a use for their gadget.
• Shipping fast and iterating (Deliver working code/hardware). The Pentagon has seen enough demos to last a lifetime. What makes people sit up and take notice is a functional prototype delivered on an operational timeline. Quiet builders don’t ask the DoW for a multi-year science project. They show up with a Minimum Viable Product in months, get it in the hands of users, and then rapidly iterate based on feedback. They fix bugs in weeks, not years. That speed and responsiveness build tremendous credibility. I recall a case where a small software team demoed a new targeting aid for special ops: within a month of initial demo, they had a dozen feature tweaks implemented because they sat with our JTACs (Joint Terminal Attack Controllers) nightly to see what worked and what didn’t. That kind of turnaround engenders trust. While others are still refining their slide decks, quiet builders are pushing version 2.0 to the field. The message spreads quietly among operators: “This thing actually works, let’s get more of it.” At that point, the startup no longer has to pitch; the solution is speaking for itself.

Perhaps most importantly, the quiet builders focus on solving problems, not on self-promotion. Sure, they can brief a general when needed, but they’d rather let an NCO or a pilot vouch for their product after using it in action. That authenticity is the ultimate signal in a world awash with noise. It’s no coincidence that many of these successful teams have ex-military folks in key roles or on call; they bridge the cultural gap. They don’t treat DoW end-users as a mystery market to be conquered; they treat them as mission partners.

Why the System Rewards Noise (and How to Fix It)

If everyone agrees we need more of the above behavior, why is the noise still so prevalent? The answer lies in misaligned incentives and structural hurdles that are hard to change but must be confronted. For one, the acquisition bureaucracy often measures the wrong metrics. It’s easier to count activities (meetings held, whitepapers written, small contracts awarded) than outcomes (capabilities fielded to units). A culture of risk aversion pervades many program offices; nobody gets fired for hiring a well-known defense prime, but bringing in a startup feels risky to career bureaucrats. Thus, the default is to tinker at the edges, fund a study, run a small demo, and declare victory even if it never scales.

The SBIR program, while well-intentioned, inadvertently became a poster child for these skewed incentives. It’s literally structured as Phase I: write a report, Phase II: build a prototype. Too many companies stop at the report. In fact, a recent Senate report highlighted that a handful of companies have turned SBIR into a business model, gorging on repeat grants while delivering almost nothing beyond paperwork. These “SBIR mills” game the system by exploiting loopholes (spinning up new shell companies to skirt funding caps, for example) and excel at the compliance paperwork and grant-writing, but not at transitioning tech to operations. They thrive because DoW has historically measured success in the SBIR program by outputs (contracts awarded, papers published) rather than real-world impact. The good news is that this is starting to change: there’s growing recognition in Congress and the Pentagon that SBIR awards should go to firms producing genuine operational capabilities rather than endless research. Proposed reforms would cap lifetime SBIR winnings to choke off the mills and favor those aiming for deployment. It can’t happen soon enough.

Another structural issue is the “valley of death” between prototypes and programs of record. A startup can pour its heart into a successful pilot with a unit, but if no Program Executive Office (PEO) picks it up and funds it long-term, that pilot withers. Why does this happen? Often because of budget cycles and misalignment. A unit might love your widget in 2025, but the budget for new programs was decided in 2023, and there’s no line item for it until maybe 2027. Meanwhile the startup runs out of cash. It’s a fundamental disconnect: the DoW’s timeline to budget and buy is out of sync with the speed of innovation. The Pentagon is trying to address this with things like the Defense Innovation Unit and the Office of Strategic Capital, but change is slow. In the meantime, founders must anticipate this gap. That might mean securing bridge funding or dual-use commercial revenue to survive the waiting period, and it definitely means lobbying for a transition path as aggressively as you pitch your technology. Align with a Program of Record early, or find a champion who will fight to create one for you.

Lastly, let’s talk about culture. Many folks in uniform or government truly want to see new solutions, but they’re understandably jaded. They’ve seen bright ideas come and go, with nothing to show on the ground. To overcome this, startups need to build trust the hard way, by delivering, not just demoing. Conversely, our defense acquisition professionals need to be empowered to take calculated risks on new entrants. Some of the most successful rapid adoptions of tech I’ve seen came when a combatant command or a service leader gave cover to try something new in real ops. We need more of that top-cover. And we need more processes that fund “test as you fight,” get the prototype out of the contractor’s hands and into an actual military exercise or deployment quickly. Nothing separates noise from signal like the crucible of real operations. If it works, fund it; if it fails, cut it and move on. This sounds obvious, but it’s amazing how many pet projects linger simply because no one ever put them to a true field test.

Signals for Investors: How to Spot the Real Winners

Venture capitalists and defense investors also have a role in amplifying signal over noise. With so many new defense-focused startups appearing, how can an investor tell which are the right startups poised to break through? A few litmus tests from the trenches:

• Domain Immersion: Look for teams that have spent serious time with their end-users. If a founder can casually reference spending a month aboard a Navy ship to trial their system, or their lead engineer has rotated through Army field exercises, that’s a green flag. It means they likely built something soldiers or analysts actually want. If instead all their “user discovery” was done via PowerPoint at the Pentagon, be wary.
• Prototype in Hand: A working prototype or at least a live demo counts for a lot. A startup that shows up with a ruggedized device or software running on real (even limited) data is far ahead of one bringing only notional architecture diagrams. Beware the startup that claims a classified secret sauce but can’t show a basic unclassified version. That’s often smoke and mirrors. The serious ones know how to demonstrate value even within security limits.
• Traction (Even Tiny) with DoW End-Users: This could be as simple as a letter from a unit commander saying, “We used this, and it filled a critical gap,” or a small purchase order via an OTA (Other Transaction Authority). If a Special Operations team spent its own limited exercise funds to try a tech, that’s a strong signal of value. By contrast, multiple Phase I SBIRs with no follow-on Phase III or production contracts are a red flag; it suggests the tech never convinced someone to buy in for real. Investors should dig into whether a startup’s touted “pilot with the Air Force” was a one-off science experiment or a stepping stone to wider adoption.
• Team Composition and Commitment: Pay attention to who’s on the team. Are there veterans or individuals with deep defense expertise in key roles? Do they understand testing and evaluation, deployment logistics, and certification requirements? A team that has a Pentagon ex-official for connections and a combat veteran for user perspective is covering both noise and signal: connections alone won’t save a bad product, and empathy alone won’t navigate the procurement maze; you need both. Also, check if the team is in it for the long haul. If their slide deck emphasizes exit strategy after winning “a couple of contracts,” they might not stomach the long defense sales cycle. The teams talking about programs of record, platform integration, and accreditation, not just quick wins, are more likely to stick it out and succeed.
• Focus on Mission vs. Hype: Ask the startup what military problem they solve. Then ask a few people in uniform if that problem is real. If you get blank stares, the startup might be a hammer looking for a nail (e.g., trying to foist a blockchain solution when no one asked for one). The strongest companies are mission-driven: they can articulate, for instance, “We reduce the time to target acquisition for an artillery battery by 50%,” or “We enable secure comms in a satellite-denied environment for small units.” They talk about mission outcomes, not just tech buzzwords. Substance impresses insiders; buzzwords impress only the ignorant. As an investor, if you can’t find the substance, neither will the customer.

By tuning your due diligence to these signals, you’ll not only pick better investments, but you’ll also encourage the right behavior in the ecosystem. Founders will learn that having concrete results and warfighter testimonials matters more than having retired generals on their advisory board (another noise red flag, by the way, if that’s their primary go-to-market strategy).

Call to Action: Send Up a Signal, Builders

The United States is entering an era of renewed great-power competition where real capability will decide outcomes. In the next fight, logistics convoys will be under constant attack, networks contested, swarms of autonomous systems deployed. This is not theoretical;, it’s the scenario we must prepare for. As one Army analysis put it, “contested logistics is, in fact, nothing new… it is the standard state of logistics in large-scale armed conflict between industrialized states.” In other words, we have to relearn how to sustain and fight under constant pressure. That demands serious innovation, not innovation theater.

So here’s my blunt appeal to the community: if you’re a builder solving a hard national security problem, we want to hear from you. To those quietly working on things like contested logistics, AI-enabled command & control, autonomous ISR (intelligence, surveillance, reconnaissance),; or counter-C5ISRT (countering adversaries’ command, control, communications, cyber, computers, intel, surveillance, recon, and targeting networks), this is your moment. The nation doesn’t need more startups, but it absolutely needs your startup if you’re fixing these kinds of critical gaps. Don’t get discouraged by the noise around you. Your focus on substance is your edge. Keep engaging directly with the mission, keep testing your tech in real-world conditions, and keep pushing it into the hands of those who need it. There are people inside the system who recognize your value and are fighting to clear the path for you; seek them out (and ignore the gatekeepers who just want more slides and studies).

For the defense leaders and investors reading this: the onus is on us to tune our receivers to find these signals. Trim the noise in our processes, whether it’s by reforming SBIR allocations or giving acquisition offices the freedom to fast-track something that’s proven its worth. We must reward those who solve over those who just pitch. If we do that, we’ll not only save time and money, but we might just field game-changing capabilities before it’s too late.

In the end, this is a call to refocus on what matters: delivering useful tools to the warfighter at the speed of relevance. No more innovation theater, no more pilot purgatory. Let’s back the people who quietly move the ball forward. The best solutions might arrive without fanfare, but when the next crisis hits, they’re what will tip the scales. Now is the time to find them, fund them, and field them. Less noise, more signals. Less pitching, more solving. Let’s get to work..