Stop building an MVP. Build version one.

In 2020 an MVP was the smallest thing you could sell. In 2026 it is the thing Lovable makes on a Sunday. The word hasn't changed; what it buys you has.

The 80 percent trap

The prototype proves the idea. Then the last 20 percent, the part that makes it safe to charge money, turns out to be most of the work: security rules, payment failure handling, email delivery, error tracking, backups, a handover somebody else can read. Founders stall here, often for months, because each fix through the AI undoes another.

What version one includes

Version one is the first release a customer can pay for without you holding your breath. Concretely:

• One core workflow, done properly, not six done halfway.
• Sign-up, login, roles and permissions that hold up to a curious user.
• Payments with the failure paths handled, not just checkout.
• An admin view so you can run the business without a developer.
• Security policies on the database, backups, monitoring and error tracking.
• Deployed on accounts in your name, with documentation.

What it doesn't include

Every feature on the roadmap. The point of version one is to be small in scope and complete in quality, which is the opposite of most MVPs: broad in scope and incomplete everywhere.

Using the prototype as the spec

If you already built a prototype, don't throw it away. It is the best specification you will ever write. A senior developer can read it, keep what works, and rebuild what doesn't in less time than the first attempt took, because the decisions are already made.

At JetBuild a web version one starts at a fixed $6,000 and ships in four to eight weeks, with a staging link every week so you never wonder where it is.