Profit attracts competition. That is the whole background assumption of economics and it is usually right: a business making unusual returns is a signal, competitors respond to signals, and the returns erode. The interesting question is not how to become profitable. It is why anyone stays profitable after the signal has been received.

The word for the answer is a moat — Warren Buffett's term, and useful precisely because it points at a property of the position rather than of the people holding it. A moat is something a competitor would have to overcome, not something they would have to match.

The test

Most things called moats fail one question: would a well-funded, competent rival doing everything right still find this hard?

Good people fail it — they can be hired. Good service fails it — it can be delivered. A strong brand mostly fails it, because brand in the sense usually meant is recognition, and recognition can be bought with enough money and time.

What passes are structural facts about the world that are true whether or not the rival is good. There are not many, and they come in four shapes.

The four that actually work

The product gets better as more people use it. Each new user makes the thing more valuable to every existing user, so a competitor starting from zero is offering a worse product even if their software is better. This is network effects, and it is the strongest of the four because it makes the advantage grow rather than merely persist.

Leaving costs the customer something. Accumulated data, integrations built, staff trained, workflow shaped around the thing. The rival has to be better by more than the cost of moving, and that gap is switching costs. It is the most available moat to a small business, because it is built by doing ordinary work deeply rather than by being large.

Being larger makes each unit cheaper. Economies of scale mean the incumbent can price where the entrant loses money. Real, and mostly unavailable below a certain size — which is why it is the wrong moat for most businesses to aim at.

Something is genuinely hard to get. A license, a location, a long-term contract, an ingredient with one supplier. Barriers to entry of this kind are the oldest form and the least fashionable, and they are often the most durable, because they do not depend on the incumbent continuing to be good.

What is not a moat, in the specific sense

Being first. Being early buys time to build one of the four. It is not itself one of the four, which is the argument on first-mover advantage.

A better product. The gap closes. It closes faster now than it used to.

Price. Anyone can lower a price. A business whose position rests on being cheapest has a cost advantage or it has nothing, and cost advantage is economies of scale under a different name.

Relationships. This is the one small business owners defend hardest, and it is worth being precise rather than dismissive. Relationships are a real barrier where they are attached to the business — a decade of institutional knowledge, integration into a client's process. Where they are attached to a person, they are that person's moat, and they walk out with them. Which is the same fact that shows up as owner dependency when someone tries to value the business.

Why it matters below enterprise scale

The usual objection is that moats are a public-company concept, irrelevant to a business with eleven clients. That has it backward. A large business can survive without a moat for years on sheer momentum. A small one cannot: its profitable niche is small enough for one competent competitor to take, and there is no buffer.

The practical form of the question is not "what is our moat" — which produces a list of adjectives — but "if a well-funded competitor decided to take our best client next quarter, what actually stops them?" If the answer is that the client likes us, that is not a moat. It might still be enough this year. It is not a reason to expect it to be enough in three.