I agree, but this assumes you actually know what you are building. This isn't usually the case for a startup.

I'm sure whatsapp wanted a million users, but they didn't originally know that would happen. If whatsapp later turned out to be about serving fewer users with more features, this becomes a story about premature optimization.