9 Dec 2025 · Journal
The quiet drop between day 2 and day 5
Day 7 is a round number. Product reviews like round numbers. Users do not live on round numbers. In consumer apps we keep seeing the steepest loss of intent between the second and fifth calendar day, after the novelty of install and before anyone bothers to open a weekly dashboard.
That window is quiet because it is nobody’s meeting. Growth owns day 0. Lifecycle owns “month 1”. The middle is a corridor where push experiments and empty states fight without a named owner. Cohort Cartography forces a labelled window here even if the company currently has no owner. The emptiness is the finding.
A useful split: users who did the core action on day 1 versus users who only completed the tour. Their day-2-to-day-5 curves diverge. Publishing a blended day-7 number then asks the wrong team to “fix retention”.
We do not claim every product has a crisis on day 3. Utility apps used on payday have different clocks. The rule is to inspect the corridor, not to install a campaign there by default. If you cannot describe what a good day 4 looks like in a sentence, you do not yet have a lifecycle; you have a calendar.
Alumni sometimes return to clinic with a day-3 email they are proud of. We ask whether the email defines the window or merely annotates it. If removing the email would make the cohort vanish, you have built a dependency, not a habit. That is allowed. It should be written on the map.