Growth
Retention curve
A graph showing what percentage of a cohort of users remains active (or paying) at each interval since signup — week 1, week 2, week 4, week 12, etc.
In plain English
Of 100 people who signed up on the same day, how many are still using the product a week later? A month? Six months? The shape of that curve tells you everything about product-market fit.
Example
A SaaS sees 100 signups on March 1. By week 1, 70 are still active (70%). By week 4, 45 (45%). By week 12, 38 (38%). The curve flattens around 35-38%, suggesting ~35% of users find lasting value — a solid PMF signal for the right segment.
Why it matters
Retention curves that flatten = product-market fit. Curves that decay to zero = no PMF, no matter how many signups you have. The single most predictive long-term metric.
Common mistakes
- Looking at aggregate retention instead of cohorted — masks deteriorating new-user retention
- Measuring 'logged in' rather than 'completed core action' — flatters the numbers
- Ignoring the segment dimension — overall flat curve can hide a great segment + bad segments averaged together
- Celebrating week-1 retention as if it's the metric — week-12+ is where PMF shows