The MVP skill that matters most: cutting scope

The hardest MVP skill is not building — it is deciding what not to build. Start from the single sentence a user would say if the product worked, and treat everything else as optional:

1. 1Write the one-sentence promise your product must keep, e.g. "I can send an invoice and get paid" — that promise is your entire scope.
2. 2For every feature, ask "does the promise break without this?" If the answer is no, cut it, however much you like it.
3. 3Fake the expensive parts before building them — do the work manually, stub the admin screens, skip the dashboards — so you learn without over-engineering.
4. 4Narrow the surface: one type of user, one platform, one happy path, and resist every "just in case" edge case.
5. 5Ship embarrassingly early and watch what real people actually do, not what they (or you) say they want.
6. 6Let real usage — not your wishlist — decide which feature earns build time next.

A good MVP is embarrassing in its simplicity and undeniable in its value. If shipping it does not make you a little uncomfortable, you probably waited too long.