Cohort Retention Visualizer

The shape is the story

Retention curves almost always fall at first — what matters is whether they flatten. A curve that levels off at 30% means you have a durable core of users; one that decays toward zero means you are renting growth, not building it.

Compare cohorts over time

Stack cohorts as rows and read down a column: if later cohorts retain better at the same age, your product is improving. If they retain worse, something recent broke the experience.

Stop exporting spreadsheets

Pasting counts is fine for a one-off. Pug computes cohort retention continuously from your events — pick a start event and a return event and the curve updates itself. Open-source product analytics with retention built in.

Frequently asked questions

What is a retention cohort?

A cohort groups users by when they started (e.g. their signup week). A retention curve then tracks what fraction of each cohort comes back in period 1, 2, 3, and so on — revealing whether your product keeps users over time.

How do I read the heatmap?

Each row is one cohort; each column is a period after acquisition. Period 0 is always 100% (the full cohort). Darker cells mean higher retention. A healthy product shows the curve flattening rather than decaying to zero.

What data do I paste in?

One cohort per line, comma- or tab-separated counts starting with the cohort size. Set the toggle to "percent" if your values are already retention percentages. Rows can have different lengths (newer cohorts have fewer periods).

How do I get real cohort data?

Pug computes retention cohorts automatically from a start event and a return event — no manual export. This visualizer is for quick what-ifs; Pug is for the live picture.