Back to course overview
Module 2Idea to spec 12 min

Scoping: the app you can actually ship

Version one is embarrassingly small on purpose: the walking skeleton, the cut list, and the scoping instincts that separate shippers from starters.

Apps die of ambition far more often than difficulty. The graveyard is full of 90%-done projects whose builders added accounts, teams, notifications, and dark mode before the core loop worked. This course's scope discipline — the same one the iOS course drills — is your inheritance from everyone who didn't ship:

  • Find the core loop — the one repeated action that IS the app. Replyable's: a message comes in → you read it → you send a good reply, fast. Everything that serves the loop is candidate scope; everything else is not. If your app has no single loop, it's two apps; pick one.
  • Build the walking skeleton first — the thinnest end-to-end version where every piece exists but nothing is fancy: messages appear in a list (typed in by hand at first!), clicking shows detail, replies get saved. Skeleton-first beats feature-first because integration problems — the ones that actually block beginners — surface in week two instead of week seven.
  • Keep a cut list with pride. Everything tempting goes on it: user accounts (v2 — you're the only user of v1), email integration (v2 — paste messages in for now), tags, search, teams, mobile app. The cut list isn't failure; it's your roadmap, pre-sorted. Shipping v1 earns you the right to reconsider it.

Your variant (make it yours without changing the skeleton)

Everyone builds Replyable's shape — inbox of things, detail view, AI-assisted response — but you're encouraged to reskin the domain to something you care about: customer feedback (the default), Airbnb-host guest messages, a freelancer's inquiry inbox, a community-manager's mod queue. Same tables, same screens, same AI feature, different words. Caring about the domain is worth a month of motivation, and the constraint (same skeleton) is what keeps caring from becoming scope creep. Declare your variant in this module's lab and hold it.

The three scope-creep lies (you will hear all of them, from yourself)

'It's basically the same feature' (it never is — accounts touch every table and every screen). 'It'll be harder to add later' (occasionally true, usually false, and later-you will have skills current-you lacks). 'Nobody will use it without X' (nobody's using it now — shipping is how you find out what X actually is). The counter-move is always the same: write it on the cut list, out loud, and return to the loop.