How to scope a six-week MVP

Six weeks is enough time to build a real product if, and only if, you are ruthless about the difference between the thing that proves the business and the things that make it comfortable.

Cut features, never foundations

Skip the admin panel, the second integration, the notification preferences. Do not skip auth done properly, a data model that can grow, and deployments you can roll back. Features are cheap to add later; foundations are expensive to retrofit.

One loop, end to end

• Pick the single loop that creates value and finish all of it
• Half of two loops proves nothing to a user or an investor
• Instrument the loop before launch, you are buying evidence, not code

An MVP is not a smaller product. It is a complete answer to a smaller question.