Scope
What NOT to Build First (The Cheapest Advice We Give)
The six features founders insist on first that can almost always wait — and the two boring ones that cannot. One client said this alone saved six months.
A client once said the most valuable thing we did was tell them what not to build. It saved them six months. Here is that list, in the order it usually comes up.
Can wait
- An admin panel. For the first months you are the admin. A database client and two scripts beat a month of building screens for yourself.
- Roles and permissions. Two roles instead of five removes more complexity than any other single decision.
- Notifications everywhere. One email that matters beats twelve that get muted.
- A mobile app, when a responsive web app would do. Ship the web version, learn, then decide.
- Analytics dashboards. At a hundred users you can count. Build the dashboard when counting stops working.
- Multi-language and multi-currency, before anyone has asked in a second language or currency.
Cannot wait
- Correct invoices. Wrong GST from order one is painful to unwind and lands on your CA.
- Backups you have restored. Everything else can be rebuilt; data cannot.
Why this is hard
Because the features you want to build first are the ones that make the product feel finished to you. The features that matter are the ones that make it useful to someone else. Those two lists overlap much less than anyone expects.
The same discipline applies to a seven-day build — see what fits in a week.
Questions
How do we decide between two features?
Ask which one a current user has complained about. If neither, neither is next — go and watch someone use what exists.
Should we build an admin panel eventually?
Yes, once someone who is not an engineer needs to do the job daily. That is the trigger — not the plan.
