Every founder learns the word "moat" before they've built anything worth defending. It gets reached for early — in the pitch deck, in the first hire's first week — as if naming a defensibility strategy were the same as having one.
The moats that aren't
Network effects, switching costs, economies of scale: these are real categories, but most early companies don't have enough of anything — users, data, scale — for any of them to bind yet. Calling a thin user base a "network effect" doesn't make it one.
What actually compounds early
In the first few years, the closest thing to a moat most companies have is speed of iteration compounding against a specific, narrow problem. That's not a moat in the textbook sense. It's closer to a head start that has to be continuously re-earned.
The honest version
Say "we don't have a moat yet, and here's the sequence of things that could become one" instead. It's a less comfortable sentence in a pitch, but it's the true one, and it tends to produce better decisions than pretending the moat already exists.