The Hadrian Playbook

COMMANDER · ISSUE 15 · September 28, 2026 · Istanbul

In 2018, a software operator named Chris Power spent weeks cold-calling plant managers across Dallas and found an industry quietly dying of old age. The machine shops that make America's aerospace and defense parts were fragmented into thousands of tiny firms, running processes from the 1960s, staffed by operators whose average age was sixty-three. He first tried to fix it the obvious way — a private-equity roll-up to buy the shops — then concluded the shops themselves were the problem. So in November 2020 he founded Hadrian to build new ones: highly automated, software-run factories that make precision parts faster and cheaper, and can turn a former barista into a machinist in thirty days. By August 2026 Hadrian had raised a $1.37 billion round at a valuation near $8 billion.

The last two transmissions were about companies that built the ship and collapsed the cost of the weapon. Hadrian sells to both — and to their competitors — because it decided the real product isn't any part at all. It's the factory. This is the picks-and-shovels archetype: while everyone else races to build the hardware, Hadrian is building and renting out the means of production, and getting paid whoever wins.

The business behind the hardware. Not what they build, but how they built it.

Hadrian makes precision components, but that framing undersells the move. What it actually built is an automated, software-defined production system — the design-for-manufacturing review, the automated toolpaths, the custom ERP, the robotics, the quality machines — and the parts are the output of that system. The company sells capacity and speed, not a catalog. That is the difference between a machine shop and an infrastructure company.

Moving up the stack from output to the means of production changes what you are worth. A parts vendor competes on each quote; a company that owns the factory competes on the whole industry's throughput. Hadrian positioned itself as the layer every hardware company needs and none want to build, which is a far larger and more defensible place to stand than being one more supplier.

Takeaway: Move up the stack from selling output to selling the means of production. The company that owns the factory is worth more than the one that fills the order — reframe yourself as infrastructure, not vendor.

The market Hadrian chose is enormous and structurally weak: roughly thirteen thousand machine shops in the US, only about three thousand capable of aerospace-grade precision, most with fewer than twenty employees, and an operator base whose average age was sixty-three. Tens of billions in annual defense supply-chain spend runs through firms using processes older than the people retiring out of them.

These are the best markets to attack, not the worst. Large, fragmented, and demographically dying means incumbency is a liability rather than a defense — the existing players can't recapitalize, can't hire the next generation, and can't modernize fast enough to respond. A well-funded entrant with a modern system isn't fighting strong incumbents; it's inheriting a market its predecessors can no longer hold.

Takeaway: Hunt for markets that are large, fragmented, and aging out. Where incumbents can't recapitalize or rehire, their incumbency is weakness — you're not displacing them, you're inheriting what they can't keep.

Power's first instinct was the standard one: he launched a private-equity vehicle to roll up existing machine shops. Then he abandoned it. Buying old factories means inheriting their equipment, their processes, and their ceilings — you can only optimize a plant so far before its bones limit you. So Hadrian built greenfield: new factories designed from scratch around automation, from a 20,000-square-foot proof-of-concept in Hawthorne to a 100,000-square-foot plant in Torrance to a 270,000-square-foot, $200 million facility in Mesa.

The judgment call is knowing when the asset base is the problem. Roll-ups work when the assets are fine and only ownership is fragmented. When the assets themselves are the constraint — old, slow, unautomatable — buying them just buys you the ceiling you're trying to escape. Building new costs more up front and is the only way to reach a step-change instead of an incremental one.

Takeaway: When the existing asset base is the constraint, don't buy it — build past it. Roll-ups inherit the ceiling; greenfield is the only path to a step-change when the old assets are the problem.

The industry's binding constraint is skilled machinists: they take years to train and they're retiring faster than they're replaced. Hadrian's answer was to stop depending on them. By putting the expertise into software and automation, it retrains people from outside manufacturing entirely — baristas, nurses, drivers — into productive operators in about thirty days, and runs something closer to ten machines per operator where the industry norm is two operators per machine.

This is the move that unlocks everything else. If the scarce input is skilled labor, you can either compete for it or engineer around it — and Hadrian engineered the skill into the system so the labor no longer has to be scarce. That converts the industry's tightest bottleneck into Hadrian's cost advantage, and it's why the automation isn't a feature but the whole thesis.

Takeaway: If your industry's bottleneck is scarce skilled labor, engineer the skill into the system. Put the expertise in software so the worker doesn't have to be the expert — the constraint becomes your advantage.

Hadrian's edge shows up in unglamorous numbers: equipment uptime of 75–80% against an aerospace-industry norm near 30%, lead times of one to three weeks against a typical four to sixteen, and the labor ratio above. None of these is a breakthrough invention. Together they compound into unit economics competitors using 1960s processes cannot approach.

In manufacturing, the moat is not a single clever thing — it's a stack of operating ratios each a little better than the incumbent's, multiplied together. Uptime times labor ratio times cycle time is a number that quietly becomes unbeatable, and it's durable precisely because there's no one feature a rival can copy to close it. Win the boring metrics and price takes care of itself.

Takeaway: In operations-heavy businesses, the moat is a stack of boring ratios, not one breakthrough. Uptime, labor ratio, cycle time — compound them and the unit economics become impossible to copy feature-by-feature.

Once the factory worked, Hadrian didn't only make parts with it — it began to sell the system itself. Its Factories-as-a-Service model deploys Hadrian's automation inside the primes' own facilities, and it has spun up new divisions, including a maritime arm, to point the same production engine at new categories. The company that started by competing with the primes' supply chains started renting them the supply chain.

This is the highest-leverage move a platform can make. When your system is genuinely better, licensing it to the incumbents you were fighting turns competitors into customers and multiplies your reach without multiplying your capex. You stop being one more factory and become the standard everyone else runs on — which is a larger business than any single factory could ever be.

Takeaway: Once your system beats the incumbents', rent it to them. Licensing the platform turns rivals into customers and scales reach without scaling capex — become the standard, not just a competitor.

Factories are capital-heavy, and Hadrian raised accordingly: from an early stage where 2023 revenue was around $3 million, through roughly $30 million in 2024 and 10x year-over-year growth, to a $1.37 billion round at a valuation near $8 billion in August 2026. That is infrastructure-scale capital raised well ahead of infrastructure-scale revenue, on purpose, because you cannot build the next 270,000-square-foot factory out of last year's parts sales.

The discipline is matching your financing to your physics. A software company can grow on its own cash flow; a company pouring concrete and buying robots cannot. Raising like infrastructure — large rounds against future capacity rather than trailing revenue — is the correct move when the business is physical, provided the operating ratios prove the capacity will pay. Hadrian's boring numbers are what make the un-boring raise rational.

Takeaway: Match your financing to your physics. A capex-heavy business funds capacity ahead of revenue — raise like infrastructure, not like software, and let your operating ratios justify the size of the round.

Sell the factory, not the part, and become infrastructure instead of a vendor. Attack a market that's large, fragmented, and aging out, where incumbency is weakness. Build new rather than rolling up, because old assets are a ceiling. Engineer the scarce skill into the system so the labor stops being the bottleneck. Win the boring operating ratios — uptime, labor ratio, cycle time — and let them compound into a moat no feature can close. Once the system works, rent it to the rivals you were beating. And fund the whole thing like infrastructure, ahead of revenue, because concrete and robots don't come out of cash flow. The last archetype attacked the price of the product; this one sold everyone the means to make it.

Three issues, one thread: the ship, the weapon, and the factory that makes the parts — each a company that won by building. The next archetype breaks that pattern. It's a company that barely built anything at all, and won on something the first three treated as an afterthought. No company names yet. Collect the set: each issue is one operating archetype, and this one — sell the means of production — sets up its opposite.

— COMMANDER

———————————————————————————

A new Playbook every Monday, the Weekly Transmission every Friday. If it earned its place in your inbox, forward it to one operator who's building something — that's how Commander spreads.