12 Aug 2026
How to scope an MVP without throwing it away later
A first release should be useful in the market and still be the product you keep. That is a scope problem, not a slogan.
Most first releases fail for a quiet reason. The team shipped a demo that looked finished, then discovered the next ten customers needed a different shape. The rewrite was sold as version two. It was the real version one.
An honest MVP is the smallest product a real user can run on. It is not a slide. It is not a clickable prototype with no data model. It is the cut list you can defend in a room.
Write the cut list before the stack
Start with the job the user must complete, the constraint around it, and the date it has to be live. Everything else is a candidate for later. If a feature does not change that job, it does not belong in the first release.
Put the cut list in writing. What ships. What waits. What you will not build. Founders remember the conversation. Engineers need the document when the next idea arrives on a Thursday.
Keep the bones you will still want in a year
Throwaway code is expensive even when it is cheap to write. Accounts, permissions, audit trails, and a store that can take a second product line are boring. They are also the parts you cannot bolt on after the fact without pain.
You can ship a thin interface. You should not ship a thin foundation. The first release can look small and still be the system you keep.
Measure the release like an operator
If you cannot tell whether the product is working by Friday, you did not ship a product. Decide the two or three numbers that prove the job is done. Instrument them before launch. Then listen to the first users without rewriting the brief every week.
Alvion scopes first releases this way. Senior engineers in the room. A codebase you own. A foundation that can take the next year of use.
