When the MVP becomes a full product too early
Many teams call a backlog of 30 features an MVP. In reality, an MVP should test a clear hypothesis. If you do not know what the first version is supposed to teach you, you are probably building too much.
Validation without measurement
Without analytics and carefully chosen events, the team ends up deciding on intuition alone. A good MVP has funnels, activation events, and feedback loops from the first release.