Efficient scale
Passport
- Rock
- Mathematics
- Depth
- 4 · Mine
- Time to dig
- 10+ years
- Capital
- ◐ · medium
- Solo
- ~ partly
- AI
- → eroding
- Rent
- ✗ not for sale
Essence
A market that feeds one player, or two: demand is so small, or fixed costs so high, that a second entrant is guaranteed not to pay back — so rational competitors simply never come. Morningstar's fifth moat, and the only one of their five not covered by the first batches. What protects you is not superiority but the arithmetic of entry.
How it is built
Moody's and S&P — a market that needs everyone but no newcomer
An issuer needs a rating that every investor in the world recognises; nobody recognises a newcomer's rating, so nobody needs it. Demand is closed on recognition and only incumbents have it: the big three have held the market for decades not on the quality of their forecasts — 2008 showed what those were worth — but because the position of third or fourth does not economically exist.
Pipes, towers, airports — infrastructure with no room for a second
A second pipe alongside the first, a second mast next to an American Tower mast, is pure destruction of capital: one structure serves all the local demand, and its owner leases it to every operator at once. An investor in the duplicate will never get the money back, and everyone knows it in advance — so the duplicate never gets built.
Veeva — a micro-niche as a fortress
CRM for pharma: a market of a few hundred companies, extremely deep specifics — regulation, interactions with physicians — and all of it already with Veeva, to the point that Salesforce, on whose platform Veeva was built, contractually agreed not to compete in life sciences. There is nowhere for a second entrant: the cost of catching up cannot be recovered from what is left of the market. This is the general pattern of vertical SaaS — a niche too small for two is a moat for the first.
How it is bypassed
Wait for the market to outgrow the moat
Efficient scale protects while the market is small: growth in demand changes the arithmetic and invites a second player. "Software for podcasters" fed one company while podcasting was a hobby; once it became an industry it fed a dozen. The attacker watches monopoly niches on rising curves and enters when the addressable market crosses the threshold at which a second player pays back.
The regulator breaks up the comfort
Closed markets attract regulators: after 2008, Dodd-Frank cut mandatory references to the big three's ratings out of the rules and the EU tried to grow alternatives. The results were modest — which honestly shows the strength of the moat — but the direction is stable: where a market feeds one, political demand for a second eventually appears.
Eat the niche as a feature — and the AI shift in the threshold
A horizontal platform adds the niche monopolist's function as a module and its market dissolves into somebody else's bundle. More important for the canivibecodeit thesis: AI drops the cost of development and with it the minimum market size that can feed a player. Niches where the arithmetic forbade a second entrant with a team of twenty now feed a solo founder with agents — the incumbents' efficient scale is melting not from competition but from a change in the denominator.
Verdict
A moat made of the arithmetic of entry: impenetrable while the arithmetic holds, and gone when the market grows, the regulator intervenes, or the cost of entry falls. AI erodes this one directly.