You pull up your analytics dashboard and look at the retention chart. A steep line drops from Day 0 to Day 1, then continues falling through Day 7, Day 14, and Day 30. It never flattens out. It just keeps declining. You are looking at a product that users try and abandon — and no amount of marketing will fix it.
Retention curves are the most honest chart in product analytics. They strip away vanity metrics and growth noise to show a simple truth: do users come back? A curve that flattens at any level — even a low one — suggests you have found something users value enough to return to. A curve that keeps declining suggests they did not. Learning to read these curves is one of the highest-leverage analytical skills a PM can develop.
The Core Idea
A retention curve plots the percentage of a user cohort that remains active over time. Day 0 is always 100% — every user in the cohort signed up. Day 1 typically shows a steep drop as users who were casually browsing or accidentally signed up leave. Day 7 shows who found enough value to come back. Day 30 shows who is forming a habit. The shape of this curve matters more than any individual number.
A healthy curve drops steeply in the first few days, then bends and flattens. The point where it flattens is called the retention floor. If the curve flattens at 20%, you have a product that retains one in five users long-term. If it flattens at 40%, you have something quite strong. If it never flattens — if it just keeps declining toward zero — you have a product that users try but do not find compelling enough to return to. This is the clearest signal of a product-market fit problem.