Journal
From install to habit: three weeks of watching
Habit, in vendor language, is a streak with a flame. In the studio it is a window: a start condition, an end condition, and a behaviour you can actually observe without inventing loyalty.
Week one after install is noisy. People tap everything. Week two is where many products tell a story of “drop-off” that is really a story of finishing a job. A parking app does not need a habit; it needs to be findable in a car park. Calling that audience disengaged is a category error.
Three weeks of watching, as we run it, means the same named slice, the same screen, notes on the same weekday if the product has a weekday clock. You write what repeated, what happened once, and what you hoped would repeat. The hoped column is usually the longest on day one and the shortest by day twenty-one, if the work is honest.
End conditions matter. A window that never ends is a brand speech. “Until the shared list is empty” is an end condition. “Until they love us” is not. Claim court is unkind to the second sentence, as it should be.
If your product is too new for repetition, say so. The method thins out in empty rooms. We would rather send you to a shorter clinic on entry traces than pretend a habit window exists because a roadmap needs one.