I've sat across from the staff sergeant with a working prototype, a modest Squadron Innovation Fund (SIF) budget, and nowhere to take it. More than once.
He'd found a fix worth hundreds of maintenance hours to his squadron, built it on his own time, and done everything the system told him to do. And then the system shrugged, because there was no money to scale it and no office whose job it was to catch it.
BLUF: The Air Force has built one of the best grassroots innovation engines in the Department of War (DoW) — the Spark cell network, the SIF, the Refinery, the connective tissue running up through the Strategic Studies Group (SSG) and Blue Horizons. Then it starves that engine of any money big enough to transition a capability, and gives it no guaranteed path to a program of record. Meanwhile the part of innovation that gets the stage lights — Spark Tank — rewards the best pitch to senior leaders and ends in a Pentagon announcement, not a fielded capability. We are funding the applause and starving the forge.
The engine and the game show
Start with what's good, because it's genuinely good. A Spark cell is grassroots innovation at the edge — Airmen in a real unit, chasing a real problem, with people whose whole job is to connect them to someone who can help. The Refinery can broker a path toward the right program office. The SIF puts a little money directly in a squadron's hands to try something. None of that existed in a usable form fifteen years ago, and the people who run it are some of the most committed I've worked with.
Here's the problem: none of it has a budget that can finish the job. The former AFWERX Spark has no real money of its own; a Spark cell is working out of whatever its wing can scrape together; the SIF is seed money, and even the newer SIF Marketplace mostly moves that seed money around faster. The SSG's own discretionary fund is tiny, and its also usually shared with Blue Horizons projects. So a good idea gets runway and encouragement and a demo — and then hits the wall every grassroots effort hits, because nobody owns the next dollar and no program objective memorandum (POM) line is waiting two years out on the future years development plan (FYDP) to catch it.
Now look at the part we celebrate. Spark Tank is a collaboration between the Air Force's office of continuous process improvement (CPI) and the former AFWERX Spark; finalists pitch, the public votes, and the winner is announced at the Pentagon in the spring after being showcased at the Air Force Association's (AFA) Warfare Symposium in Aurora.[1] It's a great show. But winners aren't chosen on validated warfighter requirements or on feedback from the operators who'd use the thing — they're chosen on who tells the best story to the people holding the trophy. And when the trophy is handed over, no money moves, no requirement is written, and nothing shifts on the POM. The applicable program office is told to "make it a program of record" with no budget to do it. So it dies on the FYDP like everything else — except this one got a standing ovation first. It's a self-licking ice cream cone.
We're brilliant at lighting sparks and hopeless at building the forge that turns them into weapons.
The fix
The easy answer is to say "put real money and a real landing zone behind the engine, and stop confusing the pep rally for the program." But real money has to come from somewhere. Reality is, it's going to take three moves. First, give squadron-level innovation a transition bridge — a small fund that triggers when a project has a validated warfighter requirement and a named program office willing to own it, not when it wins a popularity vote. I'll get to where that money comes from later in this series, because there's valid ideas. Second, make that part of the celebration: no innovation award, Spark Tank included, without a program executive office (PEO) landing zone and a POM plan named on the same day as the trophy. Use the Refinery to broker that PEO relationship at the start of a project, not after it's already won. Third, measure the whole enterprise by capabilities fielded to units, not by events held or ideas submitted. A Spark cell that gets one fix into the hands of a squadron is worth more than a symposium booth.
And one move that gives the whole thing teeth. When money moves on the SIF Marketplace, check it automatically — with a large language model (LLM), not a committee — against the combatant commands' own validated demand: Joint Urgent Operational Needs (JUONs), Urgent Operational Needs (UONs), and the unified combatant command (UCC) requirement lists. If a Spark cell is quietly building its own market across multiple wings (or the equivalent in the other services) to solve a problem the warfighter has already validated, that isn't a feel-good story — it's an indictment. It means the program office and the programmers — the J5/J8, or the service's A5/A8, G5/G8, N5/N8 — drew the cut line in the wrong place. So fire the trigger automatically: pull the prioritization decision from the requirements working group (RWG) and/or the Weapons and Tactics (WepTac) conference that produced that cut line in front of a small virtual panel — reps from the UCC's J5/J8 and the owning mission design series (MDS) staff — to explain the miss and fix it, up to and including a reprogramming. A cell moving SIF dollars between bases is the market pricing a demand the formal system refused to fund. Making that mid-year money move fast is exactly what a dynamic budget program activity code (BPAC) is built for — another thread for a later post.
What you can do Monday
- Unit commanders: Stand up or feed a Spark cell — the grassroots energy is real and it's cheap. But the day you fund a SIF project, name the program office that would own it at scale and call them. Don't let your troop's prototype die at the demo.
- PEOs / PAEs: When a Spark cell brings you a fix that already works in a unit, treat it as a requirement to adopt, not a distraction to deflect. The Refinery exists to broker exactly that handshake — answer it.
- Contracting officers: The SIF project that proves out needs a vehicle to scale — an Indefinite Delivery, Indefinite Quantity (IDIQ) or an other transaction agreement (OTA). Build it early so a working fix doesn't stall waiting on paperwork.
- DIU and the service innovation organizations: Spark is the best grassroots front end in the DoW. Give it a standing transition fund and a guaranteed bridge to a landing-zone PEO — connections alone are not a transition strategy. In a later post I'll describe how the DIU relationship to the service innovation organizations should be highly strengthened; this is the epitome of the design thinking portion of that relationship.
- OSD: Fund a squadron-innovation bridge that triggers on a validated requirement plus a named program office, not a pitch score. And change the metric: count capabilities fielded, not innovation events held.
- Congress: No new authority needed — the SIF and the service innovation funds already exist. In oversight, ask one question for every innovation award we brag about: how many reached the warfighter, and what was the dollar bridge from grant to program of record? If the honest answer is "a Pentagon announcement," we funded a game show.
The conversation happens on LinkedIn — come argue with me there.