Day-1 Retention
IntermediateDay-1 retention is the share of a cohort that comes back and is active on the day after their first day with the product.
The first day is counted as day zero, so day 1 is the following calendar day and not the day of the first visit; a tool that labels the acquisition day as day 1 shifts the whole curve by one step and its numbers cannot be compared. The result also depends on the retention definition in use: N-day retention requires activity on that exact day, while unbounded or rolling retention counts anyone active on that day or later, and the second always reports a higher figure. Day-1 retention is the steepest point of the retention curve and reacts to onboarding quality more than any other point. It says nothing about long-term habit, and a product can hold a strong day-1 figure while flattening near zero by day 30.
How it is calculated
Day-1 retention = cohort users active on the next day ÷ all users in that cohort
The cohort is fixed by the day of first activity and that population stays constant for the whole calculation. Being active must be defined by a named event and the output is a percentage; the day boundary should follow the project's own time zone, because cutting days in UTC pushes late-night users into the next day's cohort.