From Prototype to Launch in Four Weeks

Four weeks is not a magic number. It is the longest a first no-code game can stay exciting to its own maker without the protection of a deadline, and the shortest in which a real launch, with its checklists, pages and fixes, fits without panic. LaunchLab 3D lays every new project onto the same four week frame. The calibration story is in this directory's own data: Color Switch, a number one arcade hit, was built with Buildbox in one week. Your game does not need to be a hit. It needs to be finished, and finished has a schedule.

WeekGoalExit test
1Stack and prototype: engine picked, formats checked, core loop playableA stranger can play the loop and explain it back
2Build: levels, art and sound in, ugly but completeThe game can be played from start to finish screen
3Harden: performance, tutorials, edge cases, project page liveThree playtesters finish without your help
4Launch: checklist for your platform, publish, be presentThe game is public and you are online when it lands

Week one is where this plan earns its keep. The format check and the stack decision happen while changing your mind is free, and the exit test, a stranger explaining your game back to you, kills more bad designs in an afternoon than a month of solo polishing. Week two forbids new features: anything that occurs to you goes on the sequel list, a document whose real purpose is to let the current game be small on purpose.

The frame bends, of course. A bigger game takes eight weeks, by running the frame twice with a hardening week between. A weekend jam game takes the frame in miniature. What does not bend is the order: stack before build, strangers before polish, checklist before launch. Open LaunchLab 3D, start a project page, and let the calendar be the strict friend every finished game in the showcase once had.

Color Switch

The hit reflex arcade game created in one week with Buildbox

arcadeMade with Buildbox

by David Reichelt

What to cut when the frame squeezes

Every four week plan meets a week that goes wrong, and the frame survives by cutting in the right order. Cut content before polish: three good levels beat six rough ones, because players forgive small and never forgive broken. Cut features before the loop: the loop is the game, features are its wardrobe. Cut platforms before scope: launch on one store, well, and let the second platform be the sequel's first update. And never cut the playtest or the checklist to save time; those are the two steps whose absence is invisible on launch day and obvious by day three.

The sequel list deserves the final word. Every idea cut from this game goes there, dated and unjudged. It turns restraint from a loss into a deposit, and when launch week arrives, it is the first page of the next project, already written by the person who learned the most from this one: you, mid build, seeing clearly.

Keep reading