An MVP is not the smallest thing you can build—it is the smallest thing that proves a hypothesis. Teams often confuse minimal with incomplete: no analytics, no onboarding, no path to iterate.
Scope to learning
List assumptions in order of risk. If pricing is uncertain, test willingness to pay before perfecting the settings screen. If activation is the bottleneck, instrument the first session before adding secondary features.
Keep the foundation honest
You can defer polish, not data models or auth. Choose a stack your team can extend and document decisions so week six does not become a rewrite. The goal is momentum with optionality.
