How to scope a first release you can actually launch

Define one job, one user

Name a single user type and a single job they need to complete end to end. Anything that does not serve that path is a candidate for a later phase, however obvious it feels.

The three-question test

For each proposed feature, ask:

• Can the user finish the job without it?
• Would we delay launch to build it?
• Can it be handled manually for the first few months?

Plan the second release before you launch

A visible phase two is what makes cutting scope acceptable to everyone involved. It moves features into a queue rather than a bin, and it keeps the launch date honest.

Build the technical foundations for growth — accounts, permissions, data structure — even when the first release uses only part of them. Retrofitting those is expensive; leaving room for them is not.