Every founder wants to move fast. The trap is moving fast on the wrong things: gold-plating features nobody asked for while the core loop stays untested.
We start every MVP by cutting scope to a single, sharp question the product needs to answer. Everything that does not help answer it gets parked.
From there it is weekly shipping on a stack you can grow on, with real analytics from day one. The goal is not a demo, it is something you can learn from and, if it works, keep building on.
Eight weeks is enough when the scope is honest. It is never enough when the scope keeps growing.