MVP Scope Checklist

Quick answer

An MVP is the smallest product that can charge a customer or validate a hypothesis — not a slimmed-down version of your full roadmap. Scope it around one primary user journey, one monetization or validation path, and the minimum roles needed to support that journey. Everything else is a v2 decision.

Start with the primary journey, not the feature list

If you cannot describe the primary user journey in five steps or fewer, the product is not ready to scope. Write it out: who logs in, what job they hire the product for, how you know they succeeded, and how you get paid (or validate interest) in v1.

Every feature that does not serve that journey directly is a candidate for deferral, regardless of how important it feels in the pitch deck.

The MVP scope checklist

Use this before sending a brief to any developer.

• Primary user journey written in five steps or fewer
• v1 monetization or validation path named explicitly
• Roles limited to what the primary journey needs — not the eventual role matrix
• Must-have integrations listed (Stripe, email, auth provider) and nice-to-haves cut
• Admin/ops path defined only if a human must support users from day one
• Design polish deferred past the first working version of the core loop

What to cut first

Settings and account-management pages beyond the bare minimum. Every role except the one or two needed for the first users. Every integration that is not required for the core loop to work end to end. Visual polish beyond "doesn’t look broken" — this can follow once the loop is proven.

A brief template you can copy

Paste this into an email or message when requesting a quote:

• Product one-liner and who pays (or what you are validating)
• Primary user journey in five steps or fewer
• Must-have integrations only
• Roles needed for v1 only
• Deadline and USD budget band
• Links to competitors or wireframes, if any

Common mistakes

Designing every role before the first user exists. Building settings pages before the core loop works. Treating "AI-powered" or similar labels as the product instead of a feature with a scoped job. Deferring payments until "later" and discovering the data model cannot support them when the time comes.