Rachel Kleinfeld, president of the Truman National Security Foundation.

Rachel Kleinfeld, president of the Truman National Security Foundation.


As Congress awaits word from the Pentagon as to just how it will manage sequestration (the plan is due to Sen. Carl Levin‘s Senate Armed Services’s Committee today), we’ve got this interesting piece from Rachel Kleinfeld, a member of our board of contributors and president of the progressive Truman National Security Project. Kleinfeld argues for broader adoption of an innovative acquisition approach called FIST. No, this isn’t yet another version of acquisition reform, which has done so little to improve acquisition. Breaking D readers will recognize some of the arguments. Think of Bob Gates’ frequent mention of 80 percent solutions. Read on. — The Editor.

Clayton Christen, a Harvard business school professor, wrote The Innovators’ Dilemma to describe how disruptive innovation can undermine successful companies. First, market leaders dismiss disruptors, because the little guy can’t possibly threaten them. Then, they pay some attention, but they can’t change old habits and standard operating procedures fast enough. Then, the disruptive start-ups put them out of business.

America’s military is facing disruptive challenges ranging from improvised explosives to cyber attack. We still need our advanced stealth bombers and missile systems – but they can no longer ensure security against all threats. To avoid being toppled like other dinosaur companies, we need to make FIST – Fast, Inexpensive, Simple, and Tiny – standard operating procedure for our military.

FIST is an idea that Lt. Col Dan Ward and a growing group of procurement leaders in the DoD and DHS have been applying for the past years. It’s not a right or left idea – it’s simply better, and it turns long-standing defense procurement ideas upside down.

For instance: standard procurement practice is to ensure maximum capabilities for our warfighters. That means requirements grow and grow. Since large, expensive projects create room for corruption, contracts become more and more baroque to ensure accountability. When requirements change five years into a contract, it takes renegotiation. The contractor throws more people at the problem to try to end on time, the time lines increase, the budget increases, and delivery slows.

Now, we might have been able to put up with slow, expensive projects during the long, costly Cold War, when change occurred at a slower pace. Today, expense matters. We are in a short term state of sequester, and long term, our debt itself is a national security problem. Meanwhile, we are fighting multiple hot wars in which late delivery can lead to death on the battlefield. An 80 percent solution now beats a 100 percent solution delivered later.

But FIST isn’t just about a “good enough” system. The biggest challenge to conventional wisdom is recognizing that reducing requirements and increasing speed don’t worsen outcomes, because of the rapidity of technological change. Nowadays, a two-year-old phone is outdated. A ten-year-old computer is antique. A contract set for five years, much less 20, will either lock in long-obsolete technology – or inevitably change so many times that it is an exercise in futility. Imagine getting delivery today of cell phones requisitioned in 1993, when they were the size of a briefcase!

FIST tells procurement officers: keep a fixed schedule and delivery date – but keep requirements open, so that technological change can be built into the product. To handle the uncertainty, have a small, tight team able to respond rapidly to change during development, deployment, and post-deployment. This has worked well for rapid acquisition projects that were “spiral developed” (constantly prototyped and tested by users) during our recent wars – such as the V design on the bottom of an IED-resistant MRAP. Even when making a large, complex fighter jet, project pieces can be broken down into small parts, which develop quickly, use heavy prototyping, are tested by end users, and improve over multiple iterations.

FIST is not just a clever idea – it works. Lt. Col Ward, FIST’s champion, has demonstrated numerous examples of success  – such as the Marine Corps’ Harvest Hawk gunship that in just a year-and-a-half went from idea to battlefield deployment in Afghanistan, or the Air Force Project Liberty ISR aircraft that took just nine months. The DoD’s fastest supercomputer was created using FIST principles, and it was delivered for one tenth the price of a typical supercomputer and uses less than one one hundredth the energy. It works in industry, too: Rally Software found an average 37 percent decrease in time to market and a 16 percent increase in productivity using FIST processes– findings that hold through multiple studies.

So who is against saving the taxpayer money, delivering better products, and serving the warfighter? Well, some of the more hidebound parts of the defense industry, which make their money on large, slow, expensive projects where costs can be added in to each stage of the process. The larger and more complex the contract, the greater the barriers to entry to competition from smaller firms, who must instead subcontract bits and pieces of the work from the few big defense contractors.

We need a world-class U.S. defense industry. That means we need the industry to be profitable. And industry can be highly profitable using FIST methods – it just requires changing traditional billing and contracting structures. This will threaten some industry leaders by reducing barriers to entry, but it needn’t reduce profitability.

Beyond business models, FIST also requires cultural shifts. For many, complexity equals sophistication. Managing high cost programs bring prestige. As Ward has written, shifting to a FIST system means getting militaries to switch from valuing the most advanced, complex, expensive system to valuing a system that is affordable, effective, and available when we need it – not during the next war.

It’s time to make FIST the norm, and acknowledge that often, adding more requirements and more cost simply leads to over-priced, over-engineered goods that have more points of failure – and therefore, fail when you most need them. We owe warfighters more.

Rachel Kleinfeld, a member of the Breaking Defense Board of Contributors, is president of the Truman National Security Project.


  • Don Bacon

    “using FIST methods – it just requires changing traditional billing and contracting structures”

    That’s an important “just” when it comes to major development programs, since we’re given only two sketchy examples of how FIST has worked on two small programs, and now it’s being advocated across-the-board.

    we need to make FIST – Fast, Inexpensive, Simple, and Tiny – standard operating procedure for our military.

    Okay — but the devil is in the details. Military contracting is a complicated art & science which has developed in time, so what “contracting structure” changes are recommended in real-world development programs?

  • Don Bacon

    “the progressive Truman National Security Project”

    Not really. The Truman Project was founded in 2005 by Rachel Kleinfeld and Matthew Spence as a bridge between the armed forces and Democrats, bringing together military types and liberal arts types. It seemed necessary at the time, so they have national security bootcamps, etc. and advisers like Madeleine Albright.

    The primary underlying goal seems to be a promotion of neoliberal (= neocon) ideas of an American world order backed up by US military force. The perception was that Democrats were weak on defense, and offense, and anything military, therefore the Truman National Security Project (and Obama).

  • Weaponhead

    I haven’t heard of Project Liberty as a success. I heard we divered something that was unwanted and of marginal utillity because is was politically driven. As Don points out the two examples were mod’ programs not clean sheets. I also can’t see how open ended requirements will work on fixed price and PoP? Who decisdes when the requirements change and if the new requirements are riskier and more expensive to comply with does the contractor get compensated? Or is that where the fist comes into play?
    This all souunds oversimplified and overhyped to me.

  • Don Bacon

    rapid acquisition projects that were “spiral developed” (constantly prototyped and tested by users)

    I’m confused. First I hear acquisition, meaning production, and then prototyped, meaning pre-production. Should it be “rapid development projects” instead? And this “spiral” thing has a hint of F-35 about it. Not good.

    So what we need (to understand FIST) is to get to the real world of development, test (engineering and user) and production, and also to talk about contract types (cost plus, fixed price, etc.) with real-world examples.

    Incidentally, I just read that the F-35 program office has ‘revolutionized’ their project management and gone to what in the sixties we called PERT charts — putting all sub-projects on a time line to the end project completion. Amazing. They weren’t doing it before, really?

    I’d like to hear Dan Ward on that subject, for sure.

  • Illini Rob

    I enjoyed the article. It isn’t a new problem however. I recently read the autobiographies of both Ed Heinemann and Kelly Johnson. They had to resist military bureaucracy in the 40’s, 50’s and 60’s.

    Johnson advocated lean government oversight offices with the capability and authority to respond within 24hrs to questions regarding mission requirements.

    Heinemann’s account of the transitional period between the Dauntless dive bomber and the Skyraider is very telling. The Navy kept trying to attach unnecessary requirements to attack aircraft which was increasing their cost, weight, and complexity — ultimately degrading capability rather than amplifying it.

    Those who fail to learn from history are …

  • Sean Wagner

    Interesting comments above, serving up a dose of historical – and hence well established – realism. I have learned a lot from partially inadequate, but working prototypes. The more, the by far better. It’s a simple maxim that requires everyone to understand in advance there will be [ sometimes arbitrarily triggered ] times when you must deliver your best working implementation before being allowed to proceed – or rethink your next short hops.

    “An 80 percent solution now beats a 100 percent solution delivered later.”

    Too often at the current pace of progress, the 100 percent effort at introduction is relegated to the museum of asymmetry by a readily conceived, oblique and cheap 200 percent solution, leaving an open feedback loop with all its attendant loss of lessons. Gulliver’s unwieldy team never gets to stand up before falling flat on its face, while the successful experimenters [!] are honing all the paths to dominance, from technology selection to hardware [ and hard software ] creation, from good failures to good implementation and yes, currently useful doctrine enabling the flow of knowledge up and down the grapevine.

  • Clarence D’Amato

    FIST is a great idea when presented in
    a 700 word op-ed, but in reality, is difficult to find tangible elements in the
    current acquisition process that can be changed. How do you pursue an 80%
    percent solution on something that needs to fly, fire, or protect 100% of the

    the DoD’s process for qualifying parts and testing systems is too slow, and the
    DoD could undoubtedly streamline the process by which it collects ‘information content’ on systems
    to certify them. Currently, The DoD approaches technology development as the
    FDA approaches the approval of a new drug, by gathering as much data as
    possible on the system to determine its effectiveness and interoperability with
    legacy systems, and to minimize errors. As Ms. Kleinfeld argues, this
    technology development pipeline needs to be streamlined for the DoD match the
    breakneck speed of technology in other sectors.

    comments below ask essential question: Where should changes in the acquisition
    process be made, and what does this “streamlining” look like? The
    answer is that the DoD will have to pick its spots. Harvest Hawk, highlighted
    by Ms. Kleinfeld, is a fantastic example of a need that was ripe for the
    application of the FIST method. The system is a modular surveillance and
    bombing kit–think a pod that can hold cameras and weapons–that can be rapidly
    mounted onto existing KC-130 refueling aircraft . A KC-130, which previously
    did nothing but refuel other aircraft, equipped with Harvest Hawk kits immediately
    becomes an intelligence, surveillance, and reconnaissance asset. The rapid
    development of Harvest Hawk, however, was only possible because it does not
    necessitate any re-engineering of the KC-130 structure and avionics systems.
    Its data remains separate from the mission system, and Harvest Hawk only relies
    on the aircraft for power, communications, and GPS data.

    Harvest Hawk example shows how the DoD identified an area where FIST methods
    can be applied and then followed through. On a broader scale, DoD program
    offices should examine their R&D portfolios and identify technologies that
    can be rapidly brought to bear, paying particular focus to solutions that are
    modular and have a small footprint (i.e. don’t necessitate massive
    re-engineering of complementary or legacy systems).

    is another example. The current model used by military space agencies, is to
    purchase large, expensive satellites that are costly to build, inevitably
    suffer from delays, and are increasingly costly to launch. In commercial space,
    a revolution is occurring with small satellites that are classic 80% solutions;
    maybe they won’t last as long in orbit, but they offer similar capabilities to
    large, expensive satellites and can be launched more cheaply and flexibly.
    “Constellations” of small satellites can replace a single
    “exquisite solution” with significant reductions of risk. When one of
    the satellites in a constellation fails, another one can easily be launched
    into orbit to replace it. If small satellites aren’t the definition of Fast,
    Inexpensive, Simple, and Tiny, I don’t know what is.

    surveillance kits and satellites may have very little in common, they both
    represent technologies where FIST can apply and where acquisition reforms are
    tangibly apparent. Harvest Hawk works because it can be easily popped onto
    existing aircraft and immediately provides them with addition capability. Small
    satellites work because they exploit the massive gains in satellite technology
    to provide 80%+ replacements of existing solutions that are prohibitively
    expensive. These are scenarios that exist in countless corners of the defense
    technology universe – and it is up to the DoD to find and exploit them.

    • Sean Wagner

      “How do you pursue an 80%
      percent solution on something that needs to fly, fire, or protect 100% of the

      Thank you for providing practical examples,
      like Harvest Hawk and small satellites. And as you note,
      getting to a useful 80% involves
      A] Not reinventing the big wheel – which requires great perseverance.
      B] Revolutionary capabilities in tiny envelopes
      C] Modularization via proliferation in numbers

      D] Working the man-machine interface. This is where the next revolution will happen, and its physical envelope will be tiny to nil.

      Time-to-field must be short to enable the entire team’s feedback loop. The learning, adaptable team is the true sine que non. Some steps may resemble a stumble – small money drain, big experience gain.

      Finally, the 80% solution is so vital because it will never last very long.
      Part of the perfect 100% reside in the knowledge and proficiency of the team developing / producing / fielding a sub-system, while enabling the fighters to adapt it optimally to the battlefield. A kind of lenghtened kill-chain, where excellent feedback is essential.

    • Sean Wagner

      Come to think of it, how would you exercise the Kill-Chain / Long ?

  • Don Bacon

    In any case, nice photos of Ms. Kleinfeld. Sure beats ugly ships and planes.

    • Sean Wagner

      Well, ships used to have figureheads…

  • Daniel Goure

    The idea of building the next fighter, tank, destroyer, submarine or even an artillery piece according to the FIST paradigm is simply absurd. Unless that is we agree that we will replace the F-35 and legacy fighters with a version of the Light Attack Support aircraft being built for the Afghan Air Force. The military examples cited, Harvest Hawk and MC-12 Liberty are, respectively, a variant of an airplane in service for decades and a commercial non-developmental platform. That is relatively easy.

    There is nothing new in the FIST concept. The Army’s Rapid Equipping Force (REF) has successfully demonstrated a FIST-like approach for narrowly defined urgent operation needs The REF also has the advantage of being able to work around the acquisition system.

    The NASA model does not apply to defense procurements. The NASA problem is far simpler than that confronting defense, even with the challenges of going into space. No one is shooting back. Tasks can be broken down into simple parts. Try and do manned space flight according t FIST. Don’t think so.

    But how do you apply FIST or the REF model to a major next generation platform? Fast relates to two things: the pace of the acquisition system and the complexity of the system being built. Since we have examples of brand new systems incorporating advanced, even experimental technologies, being designed and built rapidly (U-2, SR-71, Polaris) the problem is mostly with the burdens that government now places on industry. We can have systems that are fielded fast, inexpensive to procure and simple in form so long as we don’t have them do much or don’t ask it to go in harm’s way. Then they can all be built to commercial standards. We then can buy lots of them and drive the price down even further.

    Tiny, that is the best of all. How about a tiny aircraft carrier, destroyer or nuclear submarine? Why not just a Boston whaler with an outboard motor? We have had arguments about small is beautiful in defense systems forever. Remember Admiral Czebrowski? Not easy in practice.

    FIST may apply in the electronics world but not clear it works anywhere else. Also, the development costs and learning curve to get to the modern, throwaway cell phone has been in the tens and tens of billions. It also helps to make it in China.

    By the way, the F-35 was being built according to a FIST-like approach. It is called concurrency. Build what you can now and update as you go along. In fact, one could use the initial production lots of the F-35 that lack the sophisticated features in a FIST-like mode: just fly them until they wear out and throw away. But, now everyone thinks concurrency is a bad idea. Under Secretary Kendell called it “acquisition malpractice.”

  • john martin

    What a joke, about the spiral development of the v design on the MRAP. Just about every designer in the world has known about a v hull dispersing mine blasts since the Second World War. This exemplifies how young editors like Rachel think everything they read about was invented after the Internet came on line. It is sickening to me that people with virtually no engineering experience get passed along as experts. Another issue she is totally lost on is giving a contractor an open ended contract: where do you start on this one, how about constructive change to contracts, or scope creep. Has this woman ever managed a time and materials contract? I think not or she wouldn’t be spewing this BS. Rachel please do a case study on the Boeing Lead System Integrator contract for the now defunct FCS program and you’ll see the biggest cost driver they had was giving Boeing open ended system requirements. Sort of like, give me a cost plus contract to build you a house, and turn me loose, wow, your house never gets finished, my work goes on to infinity, life is good. The Army finally shut their FCS program down after spending $20 billionusing your suggested approach, and they got nothing, zero, nada!