It has become commonplace for leaders across the U.S. Defense ecosystem to bemoan the slow pace of DoD acquisitions. We covet a future military that is agile and mission-centric. The Department and Congress have attempted to address this lethargy with innovative, streamlined contracting and acquisition policies (e.g., OTs and “804 authority"), new organizations (e.g., DIU), and even new warfighting concepts (e.g., JADC2).
But these are not enough. Despite all these efforts and cause for optimism, a fundamental roadblock is the way in which funding is appropriated. Appropriations bills are passed by Congress on an annual basis, and the planning process begins two years prior to that. This lag lends itself to the Services making huge, “all in” bets on their programs. If you were a program manager and couldn’t get your funding for three years, wouldn’t you throw everything you could into that budget request? The outcome is that DoD capabilities cannot scale without pushing for very large multi-billion dollar commitments because it just takes too long to fund things in incremental chunks.
Too big to fail
But expectations grow with size. The government can be very agile when there is not a lot of money involved. An example is the so-called “technical operations” mission in the intelligence community and special operations world. While much of this work is shrouded in secrecy, it has been popularized in the entertainment world by James Bond’s “Q” and the old television series, MacGyver. These depictions have organizations crafting technical solutions tailored to immediate operational needs. Government organizations that do this for real avoid the oversight associated with major acquisition programs because they are small. A typical tech ops group may have a budget of "only" a few tens of millions of dollars per year. While this is a lot of money in the eyes of a typical American household, it pales in comparison to the billions spent on a major acquisition. A multi-billion dollar program becomes "too big to fail," and when the bets are a billion dollars, laborious oversight is justified.
An alternative model from the Startup world
Silicon Valley and the venture capital communities are known for their innovation and agility. While most of these success stories start small, the figurative "two guys in a garage," many have ended up becoming "unicorns" and are now some of the biggest, most powerful corporations in the world. These startups start small, accept risk and move fast. They can experiment, fail, learn, pivot, and repeat until they grow – or they die with relatively little loss or regret. This is the model posed by Eric Ries in his book, The Lean Startup. Ries posits that investors can balance risk with the reward that comes from scaling using speed. He uses the term “minimum viable product” or MVP (currently in vogue around the Pentagon but running a serious risk of being sloganized) as the simplest, fastest, cheapest test of the startup’s product or business model. If the MVP works, the investors give them the next round of funding to take the next baby step in scaling the business. If the MVP does not work, the founders can make a case for a pivot based upon what they learned, or the VC can stop funding and cut their losses.
Each step is small, so the risk to the investors is minimized. Their hope in growing into a unicorn comes from speed. The increment of each step is small in money but also in time. If they occur fast, the startup can use the law of compounded scaling to grow quickly. During that time, they have likely gone through many MVP cycles.
Appropriators as VCs?
Despite many current DoD prototyping efforts claiming to do MVPs, there are many factors that prohibit this model from working in the same way it does in the startup world. While many of the problems are tied to culture and bureaucratic inertia, the Budgeting and Appropriations process (hereafter just called “appropriations”) is a fundamental barrier to DoD achieving agility at scale.
The ideal solution is for appropriators to perform their duties like VCs and to compress the timelines between budget allocations to be true MPV cycles, say another funding round once a quarter. This requires a rigorous but fast feedback cycle and the ability to vote on a decision and move funds equally fast. Unfortunately, members of Congress cannot and should not spend all of their time and effort on approving MVP cycles. They have too many other responsibilities and lack the skills and means of evaluation even if they did have the time.
A more practical model might be for Congressional appropriators to act more like Limited Partners or LPs. In the private capital world, LPs are institutional investors, sovereign wealth funds, or even wealthy individuals who invest their money with a VC to make investments and manage on their behalf. So if Congress is an LP, OSD CAPE, or perhaps some construct that does not exist today, call them a Mission Manager, would serve as a VC to make rapid investment rounds in Department MVPs. Congress would appropriate the traditional large allocations to the Mission Manager VC on the traditional budget cycle but with discretion that does not exist today in a program office. Program Offices coming to the Mission Manager would be the startups with their MVPs. The Mission Manager would have the freedom to make rapid cycle, scalable budget allocations for these MVPs but would also be the ones accountable back to the appropriators to demonstrate the return on their investment.
Treat Agile Appropriations itself as a MVP
There is no presumption these initial recommendations are viable. What is known is that risk management and accountability must be balanced with speed. This same tenet applies to the implementation of these concepts themselves. Nothing proposed here should be leapt into as an all-in bet. Instead, new ideas should be piloted as MVPs. Moving in too abruptly risks breaking a system that is not great but functions—not moving at all risks our Nation's defense slipping continuously behind and eventually becoming irrelevant in a dangerous and complex world.
Dr. Timothy P. Grayson, Director, DARPA Strategic Technology Office
The ideas and opinions expressed in this article are the personal views of Dr. Timothy Grayson and do not necessarily represent the position of the Defense Advanced Research Projects Agency or the U.S. Department of Defense.