Why defense primes still can’t innovate

Professor Clayton M. Christensen was in his Harvard Business School office when his telephone rang. On the other end was William Cohen, Secretary of Defense, who asked: “Clay, I wonder if you could come to the Pentagon because we want to study what you have written.”

That meeting would change not only the professor’s life but how the US military operates.

The Defense Secretary had read Christensen’s book The Innovator’s Dilemma and wanted to know how his theory of disruption applied to the US military.

Christensen expected a small meeting. Instead General Hugh Shelton, Chairman of the Joint Chiefs of Staff, had cleared schedules and gathered the chiefs of the Army, Navy, Air Force , and Marine Corps, their undersecretaries, and their assistant secretaries.

Christensen got nervous and asked the Defense Secretary, “What do you want me to say?” Cohen answered calmly, “Just present your material.”

So Christensen told them about steel…

Historically, most steel has been made by integrated mills—massive, costly operations that handle everything from raw ore to finished product.

The alternative is the mini-mill, which melts scrap in electric furnaces and makes steel 20% cheaper.

Imagine you’re the CEO of an integrated steel company. Here’s a technology that cuts the full cost of making your commodity by 20%. Wouldn’t you adopt it?

Yet not a single integrated steel company anywhere in the world did it successfully. Why they failed is what Christensen calls the innovator’s dilemma:

The things that make perfect sense are the very things smart people find impossible to do.

Like any other market, the steel market is composed of tiers. At the bottom of the steel market is concrete reinforcing bar, or rebar—the steel inside concrete to make it stronger. At the high end of the market is sheet steel, used to make planes and cars, while angle iron and structural steel lie in between.

When mini-mills became viable in the late 1960s, their quality was uniformly bad. Their only opening was rebar—a crummy, low-margin market with almost no specifications.

When the mini-mills attacked the rebar market, integrated steel companies had no reason to protect a 7% profit margin when they could exit to earn 12% in angle iron. So the integrated mills ceded rebar and moved up-market, while mini-mills and their 20% cost advantage took over.

Both the integrated steel companies and the mini-mills increased profits, and everyone was happy—until the last integrated mill exited rebar. Once only mini-mills were left, they could only compete with each other. Prices collapsed 20%.

With rebar prices collapsing, the mini-mills needed a new market. So they looked at angle iron: raised steel quality enough to step up into that business, and they’d have a 20% cost advantage. Competing for crumbs in rebar no longer made sense.

When the mini-mills attacked angle iron, integrated mills again had no reason to protect this business. Focusing on structural steel would lift their margin from 12% to 18%.

By 1996, the mini-mills dominated sheet steel, the top tier, and the last integrated giant closed in 2003.

When Christensen described how disruption had killed the integrated steel companies, General Shelton raised his hand…

“You’re clueless about why we’re interested in your model, aren’t you?”

Shelton stood up, walked to the chalkboard, and pointed at sheet steel—the highest-tier, most demanding market segment.

“For us, that’s the Russian military threat. And what you call the integrated steel companies, we call the US Department of Defense.” He then pointed down to rebar. “That is terrorism. And what you call mini-mills, we call non-state actors like al-Qaeda. Nothing in our current structure is built to fight rebar entrants.”

Every hand in the room went up. The question was the same: had any industry leader ever survived being disrupted from below?

Few industry leaders survive disruption, Christensen said, and poor management isn’t the reason. A leader does exactly what it is trained to do: make its good product better for its highest-margin customers. Steel did. So did the defense industry, and the Pentagon itself.

The few that survived set up a separate business unit with its own incentives, built to meet the new competition.

Two months later, Secretary Cohen called Christensen again. “Clay, today we are announcing a different organization, separate from the Army, Navy, and Air Force. We are calling it the Special Forces Command.

For six years we have been arguing about how to organize ourselves,” Cohen said. “Your research gave us a common language.”

Christensen’s theory applies far beyond steel, but to see why, we need to understand one thing:

Not all innovation is created equal.

Disruptive innovation works its way from the bottom of the market to the top—as mini-mills did in steel, like Sony did with the Walkman, and Toyota did with the Corona. These are products that many people can afford.

Because many people can buy them, these innovations create jobs and capital: more people are required to make, market, sell, and service them. Christensen argues that nearly all historical job creation traces back to disruptive innovations.

Sustaining innovations make good products better. The disadvantage is that they replace products that already exist, so they don’t increase the number of jobs.

Efficiency innovations let you sell the same product to the same customer more cheaply. They eliminate jobs, but, Christensen argues, they also free capital.

When Taiichi Ohno introduced the Toyota Production System, vehicle assembly time dropped from 60 days to two, freeing capital sitting trapped in unfinished cars. So where does that capital go?

Businesses are judged on profitability ratios like RONA (Return on Net Assets), ROI (Return on Investment), and EPS (earnings per share). Disruptive bets can take 10 to 15 years to pay back and drag down those ratios in the meantime.

Investing in another round of efficiency innovations improves ratios in one to two years.

So the vicious cycle begins.

Internal processes are improved, labor is reduced, production is outsourced, and labor is reduced again. Not because of poor management, but because it was the right—and most efficient—thing to do.

Gradually the economy becomes financialized, people lose their jobs, capital becomes abundant, inequality increases, and discontent builds up, until political disruption resets the system.

This exact dynamic paralyzed the defense industrial base. 

Following the post-Cold War “Last Supper” consolidation, the Pentagon’s 51 prime contractors shrank to five. At the same time competition between branches, such as the Navy and the Air Force, was eliminated and procurement centralized.

Without competition in the low-tier segments, innovation stopped. Efficiency innovations took over, to the extent that China now makes 40% of microchips in advanced US military platforms.

At the core of Christensen’s teaching is competing against nonconsumption—entering where the customer’s alternative is nothing. For today’s defense disruptors, like Palmer Luckey of Anduril and Alex Karp of Palantir, that means building scalable, software-first products for capability gaps incumbents consider too small or unproven to touch.

Luckey’s first major win was not even a traditional military contract; it was a deal with Customs and Border Protection, a market the incumbents deemed unprofitable.

As Luckey puts it: “All the billionaires do the same thing. They all establish their own space company. I don’t want to do that…”

“I just want to create things no one else has thought of.”

That is how you enter a market with a single buyer, a monopsony. You do not attack sheet steel. You invent a market the primes have no incentive to see.

In commercial markets, millions of buyers decide winners on price and utility. In defense, a single buyer controls the market—and it’s bound by rigid acquisition codes and pressured by incumbent lobbying.

Cost-plus contracts pay primes more the longer and costlier a program runs, while acquisition officers are rewarded for minimizing risk, not for moving fast.

Modern disruptors bypass this monopsony lock by building software-defined hardware on private capital. By bringing field-ready assets to the Pentagon at a fraction of the cost, they force acquisition leaders to choose between a working capability today or a multi-billion-dollar promise a decade out.

Where necessary, they sue: Palantir took the Army to court in 2016 to force open competition. This approach dramatically improves military effectiveness, even if it complicates things for managers used to thinking in Soviet-style five-year plans.

For the Pentagon in 1999, the disruptive “rebar” threat was asymmetric terrorism. Today, for defense industry entrants like Palmer Luckey, it’s autonomous software, cheap drones, and distributed sensor networks.

Having won the bottom of the market, these software-first entrants are now climbing toward sheet steel—the primes’ jets, warships, and submarines.

The primes are defending the top tier, just as the integrated mills did. And when they fall, the top of the defense market will look very different.

—Andreas Svanlund

Andreas Svanlund is a former military officer turned business strategist. He was central to the work that opened the Norwegian Continental Shelf to deep-sea mining, connecting diplomats, politicians, and industrial leaders on both sides of the Atlantic.

He has since turned around companies across defense, law enforcement, and oil and gas.

His conviction is simple: a system that stops improving starts contracting. Rational optimism is the force that keeps progress happening.