A first version is not a smaller version of the finished app. It is the shortest path to learning whether anyone wants it at all.

The mistake is scoping the first release around every feature the roadmap promises. That build takes six months, ships late, and learns nothing until it is too expensive to change. We scope the opposite way, around the one thing a user must be able to do.

Everything else is a candidate, not a requirement. Sign-in with a phone number can wait. A settings screen can wait. The core loop cannot.

Ship that, watch real people use it, and the roadmap rewrites itself with evidence instead of opinion. The second version is always better for it.