Every app project starts with a list of twenty features and a budget for six. The failure mode is not choosing badly; it is refusing to choose at all, building thin versions of all twenty, and shipping something that does nothing especially well.
The useful discipline is a cut list: a written document naming everything version one is explicitly not doing, signed by the same person who signs the scope.
One job, done unarguably well
The first release should replace one specific thing your users currently tolerate: a spreadsheet, a WhatsApp group, a phone call, a queue. If you can name what it replaces, you can tell whether it worked.
On a record label app we cut the feature list from twenty to three: releases, pre-orders, tickets. It shipped in eighteen weeks. The community features everyone wanted are now being built with revenue rather than with hope.
Architecture is where you buy the future
Cutting scope does not mean cutting foundations. Authentication, data modelling, environments and a deployment pipeline all belong in version one even though no user will ever see them, because retrofitting them is where rewrites come from.
The rule I use: cut features freely, never cut the things that determine how easy the next feature will be.
Budget for the second year
A live app costs roughly 15 to 20 per cent of its build cost every year just to stand still: OS updates, store policy changes, dependency patching, certificate renewals. That is not maintenance in the sense of fixing your mistakes; it is the platform moving underneath you.
Clients who plan for that keep their app. Clients who do not end up with something that stops working on a phone released after launch, and a quote for a rebuild.
If you take four things away
- Write the cut list down and get it signed alongside the scope
- Version one should replace one nameable thing
- Cut features, never foundations: auth, data model, pipeline
- Budget 15–20% of the build cost a year to keep a live app alive
Get the next one when it lands.
Questions about anything in this piece? Email me directly.
Next: Hosting decisions you make once and live with for years
