The Cluster Playbook

Four of the last six companies we tore down come from one country. Three from a single city. That’s not a coincidence — it’s a map. In the first half of 2025, German startups pulled in roughly 90% of all European defense-tech funding, and the companies clustered around one southern German city raised about $1.7 billion in a year — up eighteenfold since 2020. This issue isn't a company. It's the engine room itself: why one place produces winner after winner, and what the pattern teaches a founder anywhere.

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

Clusters aren't an accident of geography, and they aren't about one lucky breakout. They're a machine with parts you can name — and, mostly, copy. Here are the six dynamics that turned one region into the engine room, and the operator lesson inside each.

Clusters form where the specialized talent already lives, not where someone declares a hub. This region had a decades-deep bench of aerospace, automotive, and mechanical engineers, anchored by a world-class technical university whose alumni keep showing up on founder lists across the sector. The companies didn't import that talent — they tapped a reservoir that was already there.

That's the precondition nobody can shortcut: a hard-tech cluster needs a pre-existing concentration of exactly the engineers the work demands. The defense-tech wave didn't create the talent pool; it drew from one that generations of industry had already filled.

Takeaway: Build where your specific talent is already concentrated, not where rent is cheap or a grant is offered. A cluster is a talent reservoir first; everything else is downstream.

The clearest sign of a real cluster is that its second generation of founders comes out of its first. We saw it directly in this series: one founder built the patient company and then the fast one; alumni of the early breakouts leave to start the next companies, carrying the playbook with them. Winners don't just succeed — they spin off founders, operators, and entire teams who start again next door.

This is the flywheel that makes a cluster compound. One success becomes five new companies within a few years, each staffed by people who learned how on someone else's dime. The region's advantage isn't any single firm; it's the rate at which firms reproduce.

Takeaway: The second wave of founders comes from the first wave's payroll. If you're building a hub, optimize for breakouts that will spawn spin-outs — the flywheel matters more than any one win.

Money is a trailing indicator of clusters, and a brutal one. Once a region proved it could produce real defense-tech companies, capital piled in — that southern German hub went from a rounding error to roughly $1.7 billion a year in about five years, and the country as a whole captured the overwhelming majority of the continent's defense-tech investment. Investors don't spread evenly; they crowd the place that's already working.

For a founder, the implication is uncomfortable but real: being inside the proven cluster is itself a fundraising advantage. The same pitch raises faster surrounded by funded peers than it does in isolation, because proximity to other wins is read by investors as lower risk.

Takeaway: Capital clusters harder than talent does. Being in the proven hub isn't vanity — it measurably shortens and de-risks your raise.

Defense hardware has to be manufactured, and this region sat on one of the world's deepest industrial supply chains — precision manufacturing, components, and a dense web of mid-sized suppliers built over a century of auto and machine-tool industry. A hardware-defense startup there inherits a manufacturing ecosystem it would take a decade to assemble anywhere else.

That's the part of a cluster that's hardest for a rival region to copy and easiest for a local founder to underrate. Software clusters need talent and capital; hardware clusters need those plus a physical supply chain. The regions that win defense hardware are the ones that were already good at building things.

Takeaway: For hardware, the cluster's existing supply chain is the moat you didn't have to build. Locate where the industrial base already exists — you can't bootstrap a century of suppliers.

In a real cluster, everything is closer: the engineer you want to hire is across town, the supplier is an hour away, the investor has already backed your neighbor, and the customer's procurement office is in the same time zone and culture. Each of those is a loop — hiring, sourcing, raising, selling — and density shortens all of them at once. A company in the hub iterates faster not because its people are better but because its loops are shorter.

Compounded over years, shorter loops are a decisive speed advantage. The isolated company spends its scarcest resource — time — on logistics and travel that the clustered company spends on building.

Takeaway: A cluster's real product is shorter loops. If you're inside one, exploit the proximity ruthlessly; if you're outside, you're paying a time tax on every cycle.

Once a region is known as the place for a category, the belief does its own work. The best people move there because the best companies are there; the best companies form there because the best people are there. Attention, press, and ambition concentrate, and the hub becomes the default answer to "where should I build this." That gravity is the last dynamic to switch on and the hardest to dislodge — it's why established clusters persist long after the original advantages have eroded.

For an operator, the lesson cuts both ways. Inside the cluster, the reputation is a recruiting and fundraising tailwind you should ride openly. Outside it, you need a reason for the best people to ignore the gravity — and "we're remote and cheaper" is rarely it.

Takeaway: A cluster's reputation eventually recruits for it. Use the gravity if you're inside; if you're outside, you need a genuinely compelling counter-reason, not just lower cost.

Issue 09. We've spent eight issues mostly inside the engine room — the cluster, its winners, its gravity. So next we go looking for the exception that tests the rule: a company building far from any hub, refusing the gravity, and winning anyway. If clusters are this powerful, what does it take to beat one from the outside — and is "outside" sometimes the advantage? We'll name it next issue: the case against the cluster.

Commander is the business behind the hardware. Not what they build — how they built it.

New transmission every two weeks. If it earned its place in your inbox, forward it to one operator who’s building something — that’s how Commander spreads.

COMMANDER · ISSUE 08 · commander.media · Istanbul