Say a division head has a product that has to reach customers before a competitor's does, and her company runs an accelerated approval path for projects that can be fielded within five years. The path comes with fewer reviews and a shorter chain of signatures. Her project qualifies on the first application. Two years in, the team is still finding out whether one of its core components can do what the proposal said it could, and the calendar that justified the fast path has no room for the answer.

The Pentagon runs a pathway built on a five year premise, and the Government Accountability Office published its latest accounting of it on July 2. The middle tier of acquisition pathways for rapid prototyping and rapid fielding are intended to be completed in five years, and the Defense Department plans to invest at least $49 billion across 23 of its largest programs on them.1 GAO found that between 2018 and 2025, 18 of the 40 programs that entered the pathway did so with immature technologies, which GAO defines as a technology readiness level of 6 or lower.1 Among those, GAO counts eight still on the pathway, and the technologies for seven of them remain immature.1

The label travels and the unknowns stay

The department's own policy says the pathway is not intended for programs that are primarily focused on technology development.1 GAO's conclusion is that its use "has largely been to develop immature technologies."1 Nine programs that entered with multiple immature technologies have made limited progress since 2018, and some programs reviewed that never planned a minimum viable product are ending the effort "without delivering a residual operational capability, sometimes after significant time and cost have been expended."1 GAO is a government watchdog with no product to sell, and this is a single department's portfolio, so I would not stretch it into a law about all fast programs. What it does show is a mechanism that any large organization can reproduce with its own fast designations.

Granting a program the fast pathway shortens its schedule on paper, and the unknowns inside it keep whatever length they had.

My read of why this happens is that the pathway is a decision about process, and the maturity of a technology is a fact about the work. A committee can approve a shorter review cycle in an afternoon. It cannot approve a component into working. So the learning that was supposed to happen before the program started happens inside it, on the clock the fast designation set, where nobody has a separate line item, a separate owner, or a separate date for it.

Where the schedule absorbs the surprise

The same report puts the expected cycle time of the Pentagon's major conventional programs, from program start to delivery, at just over 12 years.1 GAO expects that figure to keep climbing, in part because several programs have held their delivery dates fixed while their interim milestones slip.1 I find that detail more instructive than the headline. A date that stays put while the steps beneath it move is a plan nobody has reopened, a pattern I looked at in an earlier piece on estimates that go stale. GAO found it in those large conventional programs, and my guess is that fast programs are at least as exposed, because they are the ones with an executive's name attached to the date.

I suspect anyone who has sponsored an accelerated initiative will recognize the second half of the pattern. When the surprise arrives, the schedule is the only thing with any slack in it, and it is also the one thing the designation promised would not move. The team ends up reporting green against a date it can no longer reach, because admitting otherwise means reopening the decision that made the program fast.

What the gate should ask

GAO recommends that Defense policy require programs meant to deliver quickly that integrate multiple technologies, including those on the rapid pathway, to start with only mature technologies or to develop the immature ones separately, and the department concurred.1 Fourteen of 23 selected programs on the pathway also told GAO they do not intend to develop a minimum viable product.1 Put together, those two findings describe a fast program that has not decided which part of itself has to work first.

For a senior leader outside defense, the useful move comes before the designation. Ask the sponsor to list what in the project is still unknown, and then to separate it. Whatever still needs learning gets an owner and a budget of its own, with a clock allowed to run long, while the work that only needs building can take the fast lane. A pilot that contains research is a research project with a deadline attached, and it should be funded and governed as one.

Which of the projects in your portfolio carry a fast designation today, and has anyone separated what still has to be learned from what only has to be built?