Business
Moats: What Actually Makes a Business Defensible
Capitalism has an immune system: unusually high profits attract competitors until profits fall back to ordinary. A moat — Warren Buffett's term — is whatever disables that immune response for a particular company. Understanding moats is the closest thing business analysis has to a unified theory, and it applies whether you're evaluating a stock, a startup idea, or your own employer.
The four that matter most
- Network effects. The product improves as more people use it: each marketplace seller makes it better for buyers and vice versa; each contact on a messaging app raises the cost of leaving. This is the strongest moat in software because it compounds — the leader's advantage grows automatically with size. Its signature failure: networks can unravel as fast as they form when a cohort defects together (ask MySpace).
- Switching costs. Customers stay because leaving hurts: data migration, retraining, integration rework, contractual lock-in. Enterprise software, banking relationships, and medical-records systems live here. The subtle version is workflow entrenchment — when your product is the thing everyone's habits are built around, ripping it out costs more than any competitor can save the buyer.
- Cost advantages / scale economics. The biggest producer's fixed costs spread across the most units, letting it profitably charge prices that lose money for everyone smaller. Retail giants and cloud providers run this play; so does any software firm whose enormous development cost amortizes over millions of users. Scale moats are powerful but capital-hungry, and technology shifts can reset the game (mainframes → PCs → cloud each obsoleted an incumbent's scale).
- Intangibles: brand and regulation. A true brand moat isn't recognition — it's pricing power: the same physical good selling for more because of the mark on it (luxury goods, or the trust premium on baby products). Regulatory moats — licenses, patents, compliance regimes only incumbents can afford — are quietly among the most durable, since they're enforced by the state rather than by market fortune.
What isn't a moat
Great products, first-mover status, growth, hustle, and "our team" are all replicable — they're how you build a moat, not the moat itself. The test question: if a well-funded competitor cloned this product tomorrow and priced it 20% lower, what specifically stops customers from leaving? If the answer is "nothing, but we'd out-execute them," there is no moat yet.
Watching moats erode
Moats are stocks, not flows — they deplete without maintenance. The warning signs are measurable: shrinking pricing power (discounting to hold share), rising churn, falling returns on capital toward industry averages. And the great destroyer is platform shift: a new technology layer that makes the old advantage irrelevant rather than beaten. The most useful habit is asking not "does this company have a moat?" but "what is the moat's half-life, and what specific change would breach it?" That question separates durable franchises from profitable coincidences — and it works just as well on careers as on companies.