Mi az a Retention?

Retention is the share of users who return after their first visit or signup: of everyone who signed up in a given week, how many were back in week 1, week 4, week 12. Where acquisition metrics measure how well you attract people, retention measures whether the product gives them a reason to come back, which is why it is the number investors and product teams fixate on.

Retention is almost always read through cohorts, grouping users by when they started so product changes show up as differences between groups. What counts as "retained" should match real usage: daily return rates make sense for a chat app and are absurd for tax software. Expect the curve to drop steeply early and then flatten; the height of the flat part is your product's floor, and a curve that never flattens means no durable audience yet.

For anonymous web visitors, long-window retention requires the persistent identifiers that privacy rules restrict, so cookieless tools measure it for identified users instead. That is the model behind Analyse's cohort analysis, which tracks retention on first-party data after users log in.