Capital intensity is a filter, not a flaw.
The money and years hard technology demands are usually framed as its weakness. They're also the reason the winners face so few challengers.
By Owen E. H. MeyerJune 25, 20264 min read
Somewhere in almost every hard-technology memo is a paragraph that apologizes for the capital plan. The factory is expensive, the certification takes years, the company will need several rounds before revenue — unlike software, the memo concedes, where the product ships in months and the margins arrive early. The apology treats capital intensity as a defect to be discounted. It's a cost, certainly, and it should be priced. But the discount usually ignores what the cost buys.
The same spending that makes the business expensive to build makes it expensive to challenge. A commissioned factory, a qualified process, an approved safety case, a decade of flight heritage — each one is a barrier a competitor must pay for in the same currency: years and capital, spent in sequence, with no shortcut for either. A software feature can be replicated in a quarter by a good team. Nobody replicates a qualified production line in a quarter. The capital plan the memo apologizes for is the moat, being poured — in the most literal sense, sometimes as concrete.
The filter works on capital too
Intensity screens investors as thoroughly as it screens competitors. Capital that arrived for a quick markup leaves when the second facility needs funding, and each successive round is priced by a smaller group — the investors who underwrote the whole arc from the start. Scarcity of bidders is usually read as a warning sign. Here it's structural: the difficulty is the discount, and the investors still at the table when the capital gets heavy are buying with less competition precisely because most capital was filtered out earlier.
A business anyone can afford to build is a business anyone can build.
None of this makes burn a virtue. Capital intensity pays only when the spending buys assets that compound — capacity, certification, data, heritage — rather than covering costs that simply recur. The test is worth writing into the memo: after this money is spent, is the next competitor's path harder, or is only this company's path shorter? Spending that fails the test is just expensive. Spending that passes it is the barrier being built, on the balance sheet, where a competitor can read exactly how much crossing it will cost.
Software taught a generation of investors that the best businesses are asset-light, and inside software the lesson holds. Hard technology is re-teaching an older one: some of the most durable businesses ever built were the ones expensive enough that almost nobody else was willing to build them.