Research, ideas, and leadership for a more secure, peaceful world
Overcoming Systemic Challenges to Fielding Next-Generation Capabilities
Author
Steven Varshavsky
The U.S. defense ecosystem faces two critical challenges that create a dangerous lag between technology development and battlefield deployment—one our strategic competitors are exploiting:
Empower Innovation Infrastructure
Reform Procurement Pipelines
Strengthen Public-Private Translation
Scale Agile Acquisition Models
Bottom Line
The tools exist—but the Department must act. Closing the gap between Silicon Valley innovation and national defense capability is not just a policy challenge, it’s a strategic imperative.
The United States no longer operates in a world where its adversaries are significantly less advanced or capable; rather, it has re-entered an era of great power competition. At the same time, the People’s Republic of China’s defense budget has grown by over 7% annually, rapidly modernizing its military and diminishing the technological advantage once held by the U.S.1 Credible deterrence will require bridging the divide between next-generation technologies and the defense ecosystem’s ability to implement and field them—an obstacle known as the Valley of Death. These inefficiencies not only stifle new market entrants from competing with entrenched firms but also slow the path to full-scale deployment. The pace at which the United States overcomes these hurdles may ultimately determine the balance of power in this strategic competition. Specifically, the methods by which it will overcome these hurdles are through reinforcing innovation on-ramps, encouraging programmatic excellence, and incentivizing battlefield testing.
This problem can be conceptualized using supply and demand. Valleys of Death appear as market failures in industries where supply and demand meet at suboptimal levels; this is either because innovation is not properly incentivized—leading to stagnated supply, or because purchasing incentives aren’t aligned properly—leading to incoherent demand signals. This is especially true in the quasi-monopsonistic market for defense-focused technology in the United States. Despite this, venture firms are beginning to take notice of the market opportunity; with one firm stating that ‘the opportunity set for founders cannot be ignored—annual defense spending between the U.S. and partner nations stands at $1.6 trillion, with the U.S. alone spending half of that global budget. In fiscal year 2021, the U.S. government paid $400 billion of that to contractors, with $117 billion going to artificial intelligence, autonomy, space, advances manufacturing, energy and material science.’2 In this paper, I aim to examine both supply and demand side inefficiencies as well as key incentive realignment strategies aimed at bolstering innovation and untangling bureaucratic procurement and operational challenges.
Technology is advancing at a pace that outstrips our commonplace understanding of that advancement. This is especially true of the technology sets deemed vital to advancing US national security interests. The US Defense Department is severely limited by its reliance on the defense primes, which make up 50-60% of the Department of Defense (DoD)’s contract obligations and stifle innovation through specialization and economies of scale in traditional kinetic weapon systems.3 These large prime contractors are optimized for stable purchase orders and for navigating government bureaucracy; they cannot pivot fast enough to handle the speed at which novel, commercially-driven technologies are entering the market.
Examples of this inefficiency abound; the Future Combat Systems program aimed to modernize the U.S. Army with a network of manned and unmanned vehicles, sensors, and weapons—was canceled due to technical issues, rising costs, and integration problems. The Expeditionary Fighting Vehicle (EFV), created to replace the old amphibious assault vehicles for the U.S. Marine Corps—was also canceled due to reliability issues in testing.4 U.S. over-reliance on large, conglomerated defense contractors has proven to be problematic not only when analyzing existing exquisite weapon systems, but also may be threatening to national security as the DoD looks to incorporate bleeding edge technology like AI, autonomous weapon systems, or hypersonic.
Another supply-side challenge that suppresses the number of companies that the DoD even gets to see propose solutions to its modernization and pacing challenge, is the misaligned incentive structure between the venture capital industry and the DoD. According to Bessemer Venture Partners, “government contracts are lumpy, bumpy, and unpredictable. This leads to revenue streams that are hard to forecast, with often very large Q3 results, since the US government fiscal year ends in September.”5 This has become a shared sentiment among the venture capital industry as it looks to take advantage of a once-in-a-generation opportunity to invest in the DoD: if investors feel that they can’t time return on investment, they will reduce the deal flow to defense tech companies.
Government contracting often offers an atypical return profile to private capital, such as venture capital and growth equity firms that prefer two to three year acquisition and resale turnaround time tables, and who have paradoxically become the primary backer of defense tech start-ups. Given this mismatch in timing, many startups choose to partner with established contractors who can facilitate their entrance into the market due to their long-standing relationships and ability to navigate the DoD acquisition process.6 This is exemplified by many partnerships on advanced manufacturing and emerging technology sets, including Shield AI’s inked deal with Boeing and General Motors Defense’s deal with Anduril. In these instances, large defense primes and start-ups are using each other as stalking horses to either enter or maintain their dominance in defense marketplaces. This may work in the short run to ease investor expectations but can act as a negative externality in the incentivization of innovation.
The Defense Innovation Board cautions that culture can be at odds with forward movement connected to National Defense Strategy priorities.7 Given that the lead time at minimum for new priorities to be introduced, bought, paid for, and delivered in the DoD is two years through the traditional POM process, market innovation will rapidly leapfrog warfighter access to that innovation. ‘Congress recognized this discrepancy and provided the Department of Defense (DOD) with at least 26 authorities related to budgeting and financial management that allowed DOD flexibility in its use of funds to support research and development (R&D), innovation, and modernization activities, but provision does not equate to use; further, use does not itself equate to maximized potential.8
Only about 5% of companies graduate from a Phase II to Phase III Small Business Innovation Research (SBIR) or Small Business Technology Transfer (STTR) program, posing another “valley of death” in selling to the defense department. However, unlike the previous phases, Phase III is technically not a direct part of the SBIR program. Instead, it represents the transition of the technology into the commercial market. Funding for Phase III is actually expected to come from sources other than the SBIR program (e.g., government contracts or private investment). This means that companies have to secure funding from a program of record upon gaining access to Phase III in order to maintain commercial viability and scale their technology to a broader warfighting user base.
Funding, however, is not the exclusive component of this innovation value chain that is detached from the mission. The DIB remarked that innovation flies blind and is detached from mission need. The report goes on to recount how in two separate training exercises six years apart, warfighter feedback was not implemented by the requirements process. This highlighted a very real problem in stakeholder communication: functionally, requirements officers, acquisition officers and PEOs, and the warfighter operate in almost entirely separate spheres.9
The current Planning, Programming, Budgeting, and Execution process lacks agility, limiting the DoD’s ability to respond quickly and effectively to evolving threats, unanticipated events, and emerging technological opportunities. To get a weapon system on contract, procured, and fielded to the warfighter takes 5 years using existing contracting policies and procedures. This integration timeline effectively ensures that the DoD operates on a five-year delay in an environment where new technology advances by the day, as evidenced by Google’s recent unveiling of the world’s first Quantum Computer, wherein only months later, Microsoft and Amazon launched their own prototypes in an effort to compete.10
Conversely, even if companies manage to find a program of record or an Indefinite Delivery, Indefinite Quantity contract vehicle through a Program Executive Office (PEO)11, they will often only have access to Research, Development, Test, and Evaluation (RDT&E) earmarked funds, which have a two-year timeline before adjustments are made. In many cases, if a program isn’t able to move to procurement within this window, these funds will be in danger of reappropriation in the next National Defense Authorization Act. Further, the over-index on prototypical technology creates a very wide-mouthed and thin-spouted possible productization funnel. ‘With prototypes ineligible for most of the defense budget’s “colors of money”, productization either gets shifted to private investors (exacerbating the loss of talent) or to larger companies via mergers and acquisitions. The vicious cycle then continues.’12 This friction in scaling affects startups abilities to create a viable go-to-market strategy even if their runway has been somewhat extended through private capital.
Rapid capability development and testing funds must continue to be authorized, deployed, and protected so that the United States can take advantage of rapidly evolving technologies within the same month that they are released, not the same decade. Additionally, these funds must be shielded from reappropriation, so as to keep investors’ incentives aligned with DoD objectives. The government has a much larger role to play in this first mile of innovation intake through any number of the authorities granted to incentivize innovation.
Of the more prolific authorities supporting defense innovation, SBIR/STTR program fosters technical merit, feasibility, and commercial potential in small-business R&D. Structured in three phases with increasing technological maturity and funding, it serves as a key entry point for startups into the defense pipeline. However, one of the most persistent criticisms is the difficult transition from Phase II to Phase III, which often forces companies to seek external funding—pushing many out of the defense ecosystem. While this incentivizes commercial viability, it weakens retention of innovative firms. To address this, Congress should establish a program of record matching service between Phase III companies and PEOs delivering similar capabilities, ensuring talent remains within the defense market and strengthening SBIR’s role in innovation.
One of the most notable successes in integrating cutting-edge technology is the Defense Innovation Unit (DIU). In FY 2022, DIU issued 36 solicitations—a 38% increase from the previous year—and received 1,636 proposals from commercial offerors.13 Its creation sent a strong signal to Silicon Valley that the DoD is a viable customer. By leveraging Other Transaction Authorities, DIU circumvented traditional procurement barriers, enabling startups to generate revenue while extending their development runway. Regardless of administration, DIU strengthens long-term readiness and should be expanded in both funding and authority. Elevating DIU to a PEO within its existing framework would provide direct control over allocated funds while allowing Congress to maintain portfolio-level oversight—ensuring accountability without issuing a blank check in appropriations and connecting investment to procurement for innovation.
If the DIU’s value proposition is in finding and integrating emerging technology, the Office of Strategic Capital’s, a brainchild of Defense Secretary Lloyd Austin, is capital deployment. As previously discussed, the most valuable resource for startups entering the defense ecosystem is time, and the Office of Strategic Capital’s ability to provide secured loans through appropriated funding gives them the runway to scale. Expanding these authorities without appropriating a single additional dollar could be achieved by allowing it to utilize a broader range of financial instruments—grants, contracts, loans, and equity arrangements—enabling it to recoup and reinvest returns into further technological development and deployment.
While the first mile of innovation can mostly be bolstered by government process reform, private industry must also meet the DoD where it currently is. In many cases, the DoD doesn’t have the centralized muscle to figure out where a part of a capability should be inserted and therefore shies away from stand-alone software capabilities. Venture capital and consulting organizations who work with the DoD have recognized a proclivity to purchase a vertically integrated hardware and software product, and tech disruptors should integrate this insight into their go-to-market strategies.14
The Defense Innovation Board has already recognized that the so-called “middle of the valley” needs an “oasis” of funding to bridge DoD’s yearly portfolio of R&D investments into its two-often-three-year program-specific budget cycles.15 Absent a specific PEO for innovation, the government needs to implement an industrial policy-as-a-service model wherein it identifies key technologies it wishes to invest in, and handholds their development, maturation, and commercialization [at least within the defense ecosystem] throughout the complex acquisition process right up to and perhaps even including full scale procurement. However, several changes will need to be implemented in the existing acquisition process to make this possible.
The first, which is being debated by Congress and spearheaded by DIU is the concept of ‘colorless’ money. This provides PEOs with flexibility of timeline and allows them to select a wider variety of vendors in their contracting schemes such that these vendors do not force themselves into RDT&E buckets, or O&M buckets. The Department needs to incentivize broader interpretations of the colors of money even if they don’t create large pools of colorless money. Since the DoD is so afraid to run afoul of the Anti-Deficiency Act, it is disabused of the notion that adding some R&D activities to O&M programs will incentivize improvement. For this, Congress is essential and should create a loophole within this act for innovation-specific R&D activities.
The next change that needs to be made is for a DoD organization, likely DIU as this is partially within their mandate, to take up the mantle of shepherding promising SBIR Phase III companies, or DIU funded programs, and connect them to both operators that can test their specific technologies and PEOs that can fund them. If the DoD is to meet its 2027 pacing challenges, the time for passive defense technology diffusion must be over…the DoD needs to take an active role in pulling promising technology through the bureaucracy.
Lastly, existing Program Managers must begin to and continue to exhibit programmatic excellence in their approach to contracting within their PEOs. This involves adopting a “seeing sooner, faster” mentality, wherein program managers must be ruthless in experimentation, and even more ruthless in dumping failed experiments. Programmatic excellence also involves incentivizing creativity at the program manager level, which in turn, involves accepting greater risk in leveraging emerging technology in traditional programs. This can be accomplished by working closely with research labs in order to manage that risk and inform go/no-go decisions.16
In order for startups to be better designed for defense acquisition, and specifically for fast tracking through the acquisition process, founders and venture firms should move toward architecture-based solution design. For better or worse, if defense customers can’t immediately see how a solution fits into their big picture, they will have trouble allocating funding to it. According to In-Q-Tel, a venture firm dedicated to bringing promising technology with defense/intel applications to bear on behalf of these customers, ‘these architectures spark conversations that help work out the technological gaps that government partners have and where startups could potentially help address them.’17 Venture capital is key to that translation, and their most important function beyond providing funding runways is acting as the private sector counterpart to programs like DIU that sherpa startups through the acquisition landscape.
While having the equivalent of Seed and Series A investors that build portfolios of small investments is critical and needed in DoD and provides a constant inflow of early-stage tech bets, having Series D like investors that place big bets for crossing the valley to the POM is needed to finish the process. Such investors must also be POM sherpas.
Established corporations have already recognized this trend, and “while there were only one or two defense corporate venture capital arms five years ago, now nearly every major defense company has a venture arm. SAIC launched one last week, and Booz Allen Hamilton’s was worth $100 million at the end of last year.”18 The DoD must harness Silicon Valley’s rapid innovation both from a financing and from a capability development perspective to create a more responsive and adaptable acquisition ecosystem for defense technologies. The tools are there, the Department simply needs to take advantage.
Thus far, this paper has discussed ways to instantiate and cultivate a culture of entrepreneurship in government acquisition, but it is equally important to highlight existing examples of that culture. With its unique mission, U.S. Special Operations Command (USSOCOM) was given direct acquisition authority for unique Special Operational Forces equipment. Of all DoD entities, USSOCOM embraced an important reform of user input and user experience/interface (UX/UI) considerations by connecting operators directly with developers as a core means for accomplishing missions. While these operators face challenges at times learning this technology, that risk is baked into the vertically integrated acquisition process that views rapid iteration and risk-taking as instruments of learning. USSOCOM made daily decisions that would take the DoD bureaucracy years. An agile organization was the result, one capable of moving at commercial speeds.19 The downside of this, is that scale becomes much more challenging to other programs with more ingrained, and institutionalized processes and requirements.
To ensure this innovation does not live and die within one enclave, this initiative must be replicated in other fast-moving pockets within the DoD. Existing large-scale programs that are close to the ‘tip-of-the-spear’ should leverage this operating model in order to accelerate delivery of warfighter effect and incentivize faster cycles of experimentation. Once this culture reproduces itself across the Department and its services, program managers can take advantage of existing DoD incentives to reduce risk through consolidation and leverage an entire swath of vertically integrated programs housed under one authority that directly connects operators to engineers to program managers.
Overcoming the systemic challenges within the defense acquisition ecosystem to field emerging technology is crucial for maintaining strategic superiority in an era of great power competition. While the U.S. faces bureaucratic hurdles and a decentralized system of innovation, it can harness the power of its private sector to outpace adversaries. By streamlining acquisition processes and better leveraging the innovation of Silicon Valley, the U.S. can close the gap between research and deployment, ensuring that the defense ecosystem remains agile and responsive to emerging threats in a world moving towards multipolarity and instability.
Varshavsky, Steven . “From Silicon Valley to the Battlefield: Rewiring Defense for the Next Fight.” Belfer Center for Science and International Affairs, June 17, 2025
