The Need for Speed
In The Fast and the Furious, Ja Rule tells Paul Walker: “It’s not how you stand by your car. It’s how you race your car. You better learn that.”
We’d better learn it, too.
Today the Pentagon is standing by a very expensive car. To use the industry lingo, it’s an exquisite car. But our entire system is geared for standing still and looking impressive, when what really matters is whether we can race.
Do you feel the need for speed? Speed is the critical variable that separates winners from losers, in business and on the battlefield. The OODA loop applies to procurement as well as strategy. Move fast, you control events. Move slow, events control you. Right now, we’re moving painfully slow.
Take a look at this chart, from Bill Greenwalt and Dan Patt’s awesome paper, “Competing in Time.” It shows time-to-market for three groups of comparable goods: military aircraft (measured from first contract award to initial operational capability), commercial aircraft (from board approval to first revenue flight), and commercial automobiles (from board approval to first dealer sale):
This is a visualization, above all, of speed. What do you notice?
On the commercial side, time-to-market increased somewhat for aircraft and decreased for cars. But on the government side, time-to-market absolutely exploded, diverging from the commercial sector in the 1970s and never looking back. The F-35 took more than 20 years to achieve initial operational capability—and it wasn’t even the slowest of the bunch. The ill-fated Osprey took nearly a quarter-century to take flight.
In technological terms, 25 years might as well be an eternity. Twenty-five years ago, Windows 2000 was the state-of-the-art and the world was preparing for Y2K. If your procurement cycle takes a quarter-century to complete, that means by default you’re fighting today’s war with yesterday’s army. You’d better pray your enemy is as slow as you are.
So, to paraphrase a great website, WTF happened to cause this?
The bureaucratic “reforms” of that era are the main culprit, namely, Robert McNamara's PPBS (now PPBE) in the 1960s and David Packard's Directive 5000.1 in 1971. They attempted to give planners greater insight into, and control over, the defense acquisition process.
No doubt these reforms were well-intentioned. It’s only natural to want visibility into your organization, to control how it works, and to have assurances it will deliver—especially when billions of dollars and potentially millions of lives are riding on the outcome.
But what these reforms meant in practice is endless delay: mountains of programming requirements, competitions, compliance audits, and oversight requests. Today, the DoD takes a couple years, at least, to come up with the requirements for a new system, whether it’s an air-superiority fighter or a pistol. Then it has to incorporate the new system into its budget, another multi-year process. Then it has to persuade Congress to authorize and fund the system, which results in further delay since Congress can’t pass a budget on time and instead relies on Continuing Resolutions, which keep the government running in a state of stasis: no new projects allowed. Then it has to compete the contract, build the system, test it, etc. Then maybe, a decade or two down the road, the military can get its fighter or its pistol. Oh yeah, and along the way, DoD also has to respond to time-consuming oversight requests, while coloring within the lines of its original cost baseline and authorization parameters, lest it be accused of wasting taxpayer money or acting unaccountably.
A private company that operated in this manner would be dead before Step Two. But because it’s government, we treat endless delays as if they’re different, excusable—even necessary. This is a deep-rooted culture problem. Virtually everyone in government today grew up in a post-PPBE world. Delays are just the water they swim in. Slow is what they know.
I can’t tell you the number of times I’ve heard general officers say that red tape and crazy oversight demands from Congress are necessary because the Pentagon has proven it can’t be trusted to spend money without wasting it. These same general officers know the importance of speed on the battlefield—but they don’t apply the lessons closer to home. In both cases, gambling at the tactical level means winning at the strategic level. No risk, no reward.
Today the bureaucracy tries to minimize risk to programs, but paradoxically maximizes risk to mission. We sacrifice speed on the altar of control.
Paradoxically, we didn’t get any of the other things the reformers wanted, either. The main variables these reforms are designed to control are cost, schedule1, and performance. Famously, cost has ballooned. Schedules have been blown out. And quality has been undermined by delay in a world ruled by Moore’s law and instantaneous progress. Palantir ships 90,000 software updates a week, and we’re just one company. When you spend years gatekeeping a system, the system you ultimately field will be well behind the state-of-the-art, drifting hopelessly toward obsolescence.
It didn’t have to be like this. A different world was, and is, possible.
Look again at the chart. In the two decades after the Second World War, the United States moved fast. Very fast. Military aircraft projects took an average of five years from first contract to first flight. The trend was similar for other systems. It took five years for Bernard Schriever to build the Atlas, Minuteman, Thor, and Titan missiles. An upgraded version of the Minuteman is still in service more than a half-century later. If you need more examples, Patrick Collison has an entire article full of them, here.
There are many reasons we were able to move so quickly back then. A big reason was psychological and cultural. DoD had a vastly different culture than the one we have today. Back then, virtually everyone in government had the muscle-memory of wartime urgency to guide their actions. They remembered what it took to move quickly under emergency conditions—so they did. Another reason is that Congress and the DoD invested greater authority in specific, trusted leaders to complete projects. Instead of dictating how every penny of the budget would be used for every program, they empowered the Schrievers and the strivers in their midst to get the job done.
That isn’t to say the process of creation was conflict- or drama-free back then—far from it. There were plenty of disagreements about the best way to get things done, plenty of bare-knuckle bureaucratic brawls to settle matters. We needed heretics back then, too. But critically, the system tolerated and (perhaps unintentionally) encouraged disagreement and productive competition. The services engaged in ferocious turf battles and rivalries that pushed them to develop new, better systems faster than their “opponents.” Competing programs and personalities had champions in Congress and the civil service removing obstacles and securing funding. Today, many look on these kinds of rivalries and see unaccountable chaos that takes away from the “joint” mission of the military. I see intellectual ferment and innovation. The bureaucracy hadn’t yet grown so powerful and stifling that, like the Eye of Sauron, it could see all, control all, and stamp out all dissent.
People rationalize the ever-slowing pace of defense procurement by pointing to the increasing complexity of military technology in the modern age. In the 1950s, airplanes were aluminum tubes with wings and a pilot in them, the argument goes; today, they’re flying computers. No wonder they take longer to build!
This rationalization ignores that the commercial sector has experienced the same exponential increase in technology and complexity… but without the same increases in time-to-market. Not even close.
More fundamentally, this argument misunderstands the nature of technological change. Technology, at the most basic level, is a tool. It doesn’t slow us down—that would defeat the point. Technology enables ever-greater agility, iteration, and speed. The F-35 is a much more sophisticated machine than previous generations of fighters. But guess what? We have much more sophisticated tools to build them than we did a generation ago, too—if we choose to harness them.
The commercial sector has embraced speed, out of necessity. Just this century, software ate your granddads’ companies, chewed them up, and spit them out. It used to be that companies operated on two-year budgets. Today they live or die quarter-to-quarter, and have adopted flexible budgeting practices to move money almost in real time. AI promises even greater agility and speed for companies that treat it as more than a hype engine and shiny object. In other words, companies today are lightyears ahead of the companies just a few short decades ago.
Meanwhile, our government is stuck with an operating system that was a lemon when it was first installed, and that has only gotten slower and buggier with age.
To be sure, plenty of valiant attempts have been made to reboot the system. The Defense Innovation Unit, AFWERX, Replicator, and the proliferation of Other Transaction Authorities are just a few examples of attempts to sidestep a very broken system. (Congress has started scrutinizing OTAs, which just goes to show that no good deed goes unpunished.)
But even much of the conversation about reform focuses on the wrong variables. Would-be reformers and defense wonks obsess about dollars: getting more dollars, stretching existing dollars further, wasting fewer dollars. This wrongly assumes that money is the limiting principle, and that if we only had more of it, we could literally paper over the flaws in our procurement system.
The problem is deeper, with the system itself and the idle mindset it encourages. If you care about dollars, then what you should really focus on is speed. As Greenwalt and Patt put it, “Rather than pining for a larger budget and more capacity, the US should consider an agile budget.” Speed means fewer wasted manhours, less overhead, and longer service life for new systems. I’d gladly accept half the money for a contract if the government could deliver the money twice as fast. I know plenty of business leaders who feel the same way.
Embracing speed as the operating principle of defense procurement won’t be easy. The Pentagon is the ultimate slow-moving ship. Its turning radius is massive. Congress might be even worse.
But if we want to race our car, not just stand by it, then big changes are needed. The PPBE Reform Commission has plenty of technical suggestions on how to do that, starting with root-and-branch replacement of PPBE. That has to be a top priority.
Here’s a more radical suggestion. What if Congress acted more like a limited partner of a VC than a bureaucratic bean-counter? Instead of haggling over every detail in the portfolio, it could invest in offices and leaders (the Schrievers of our day) who have compelling theses about how to transform the military and defend the nation. These “VCs,” in turn, would make bets on a diverse pool of companies and capabilities. Some of these bets would fail. Enough would be smash hits to underwrite our national security and economic prosperity. And the whole process could move a heck of a lot faster.
Greenwalt and Patt call this approach “dynamic budgeting.” It would require risk and letting go of the wheel, which cut against the culture of control that has developed in Washington the past half-century or more. But there’s far greater risk in standing still.
A general told me recently that, at the pace things are going, “the military we have today is the one we’ll fight with in 2032.” If that’s the case, we’re in deep trouble.
The modern battlefield, like the Wild West, has two kinds of gunslinger: the quick and the dead. Better think fast.
“Schedule” may seem like a synonym for speed, but it isn’t. “Schedule” is the central planner’s prediction of how long it should take to field a particular system. At the end of the day, it’s an educated guess—or a wish. Speed is different. It’s immediate, visceral, real: How fast are you moving in reality? Can you push yourself to move faster?