Defense Valley of Death Explained: Why Prototypes Stall
The defense valley of death is the gap where a technology works as a prototype but does not become a funded program of record.
A May 2026 essay from the Center for a New American Security (CNAS) argues that the first gap, from prototype to production, has narrowed, while a second gap, from production to programmed multiyear funding, is the harder one.
What the Valley of Death Means in Defense
In defense, the valley of death is the gap where a technology works as a prototype but does not become a funded program of record. A program of record, as we use the term on this page, is a funded, budgeted acquisition program with its own line in the Department of Defense (DoD) budget.
The May 2026 CNAS commentary by Brian Katz describes production buys as annual and discretionary, which is the stage before a program of record.
The Accelerate the Procurement and Fielding of Innovative Technologies (APFIT) program office says initial funding through the Planning, Programming, Budgeting and Execution (PPBE) process can take two years. APFIT says it funds validated technology that has stalled on those budget obstacles.
The APFIT office splits the problem into four valleys, running from research to scale, and our APFIT guide walks through each one. This page covers what happens after a working prototype exists, which the CNAS commentary treats as two separate problems.
The Gap Between Prototype and Production
The first gap, between prototype and production, has narrowed, according to Brian Katz's CNAS commentary published May 14, 2026. Katz credits pathways like those of the Defense Innovation Unit (DIU) and more flexible contracting. He describes the target of those efforts as the "long-discussed valley of death between prototype and production".
Practitioners also describe this gap through the way research, development, test and evaluation (RDT&E) money is organized into budget activities. A 2018 practitioner summary on Acquisition Talk says the Defense Advanced Research Projects Agency (DARPA) and the labs are typically funded through budget activities 6.1 to 6.3. Hand-off to a program office often happens at 6.4, and sometimes at 6.5. This is a practitioner description, not an official definition.
Two Gaps in the Defense Valley of Death
The second valley of death, in Katz's argument, is the move from production to programming. He calls it the harder barrier, and he says the strongest firms are the ones running into it.
Production, in his framing, means annual, discretionary buys worth tens of millions of dollars. Programming means multiyear, institutionalized funding of $100 million or more through programs of record.
Katz says firms that stop at production stay in the "innovation ecosystem" and do not enter the "acquisition ecosystem". The acquisition ecosystem is tied to the Program Objective Memorandum (POM), the Pentagon's five-year budget plan.
Three Obstacles Between Production and a Program of Record
Katz names three structural obstacles that keep production buys from turning into programs of record.
- Fragmented ownership: operational commands can validate a capability, but they rarely control the long-term budget lines that would pay for it.
- Zero-sum budgets: a new capability usually requires cutting an existing effort to make room for it.
- A veto gap: as decisions move up to service headquarters and Congress, more stakeholders gain the power to block them.
DIU and APFIT in the Gap
Katz credits DIU-style pathways with helping narrow the first gap, prototype to production. A DIU director has said that after a successful DIU prototype, scaling often requires budget dollars from the DoD partner customer. A March 2021 Defense One report said DIU is not built to solve the gap between emergence and budgeting.
APFIT works at a different point. It funds validated technology stalled by budget obstacles, and the APFIT office says initial PPBE funding can take two years.
A capability can be validated by an operational command and still stall, because operational commands rarely control long-term budget lines. Our guide to how defense procurement works covers the budget cycle behind this, and our Small Business Innovation Research (SBIR) and Small Business Technology Transfer (STTR) guide covers that program.
| Stage | How It Is Funded | What Comes Next |
|---|---|---|
| Lab research and prototyping | RDT&E budget activities, typically 6.1 to 6.3 for DARPA and the labs | Hand-off to a program office, often at 6.4 and sometimes at 6.5 |
| Production | Annual, discretionary buys of tens of millions of dollars | Stays in the "innovation ecosystem" unless the buy becomes programmed funding |
| Program of record | Multiyear, institutionalized funding of $100 million or more | Own line in the DoD budget, tied to the five-year POM |
What the Valley of Death Means for Investors
A prototype award or a production buy is a different thing from a program of record. Under Katz's framing, a production buy is annual and discretionary, while a program of record carries multiyear funding with its own budget line.
A production order is not proof that a company has crossed the defense valley of death. By the CNAS argument, a firm that stops at production stays in the innovation ecosystem, outside the POM.
Katz reports defense start-ups captured 1.3 percent of Pentagon contract obligations in 2025, up from 0.6 percent a year earlier, so the share a little more than doubled. When you read a company's contract news, ask four questions:
- Does the revenue come from a prototype award, an annual production buy, or a program of record with its own budget line?
- Which operational command validated the capability, and does that command control a long-term budget line?
- Has the capability entered the POM, the Pentagon's five-year budget plan?
- Would funding it require cutting an existing effort, the zero-sum obstacle Katz describes?
Published Figures and Open Gaps
No official figure in the sources behind this page says what share of defense prototypes become programs of record. We do not cite one, and we leave out SBIR Phase II to Phase III conversion rates for the same reason.
Two figures would sharpen the picture: an official transition rate from prototype to program of record, and the start-up share of contract obligations in years after 2025.
Before you weigh any company on our defense startups list, find out which side of the defense valley of death its revenue sits on: an annual production buy, or a program of record with its own line in the budget.
Guides FAQs
What is the valley of death in defense?
The valley of death is the gap where a defense technology works as a prototype but does not become a funded program of record. A May 2026 CNAS commentary by Brian Katz argues the first gap, from prototype to production, has narrowed. He says the harder gap sits between production and multiyear programmed funding.
What slows the move from prototype to funded program?
The APFIT program office says initial funding through the PPBE process can take two years, and a DIU director has said scaling a successful prototype often requires budget dollars from the DoD partner customer. Katz argues DIU-style pathways and more flexible contracting have narrowed this first gap, so the harder problem now sits later, between production and programmed funding.
What is a program of record?
A program of record is a funded, budgeted acquisition program with its own line in the DoD budget. Katz describes moving into programs of record as reaching multiyear, institutionalized funding of $100 million or more. That funding is tied to the POM, the Pentagon's five-year budget plan.
What is the second valley of death?
The second valley of death is the move from production to programming, which Katz calls the harder barrier. It means going from annual, discretionary buys of tens of millions of dollars to multiyear funding of $100 million or more. He names fragmented ownership, zero-sum budgets and a veto gap as the three obstacles.
How does DIU help companies cross the valley of death?
Katz credits DIU-style pathways with narrowing the first gap, prototype to production. A DIU director has said scaling after a successful DIU prototype often requires budget dollars from the DoD partner customer. A March 2021 Defense One report said DIU is not built to solve the gap between emergence and budgeting.