What is Retention Rate?
Retention Rate is the percentage of customers who were active at the start of a period and are still active at the end, after excluding anyone acquired during that period. Analyze it as a cohort curve: group users by signup month and track the share that returns in each later period. A curve that flattens above zero can be one signal to consider with other evidence of product-market fit.
Formula
Retention Rate = (Users Active at End / Users Active at Start) x 100%Count only users who were already in the cohort at the start; do not add anyone acquired mid-period. Example: a January cohort has 1,000 signups. By the end of March, 420 are still active. Three-month retention = 420 / 1,000 = 42%. Retention and churn are complements only when they use the same population, definition, and period. Plot retention for month 1, 2, 3 and onward to see the shape of the curve, not just the endpoint.
Benchmarks and interpretation
- A cohort curve that flattens above zero can indicate continued value for some users; interpret it with other evidence of product-market fit
- Choose daily, weekly, or monthly retention intervals that fit the expected use of the product
- Compare cohorts at the same age with the same activity definition and customer segment
- Net revenue retention above 100% means expansion exceeds revenue lost to contraction and churn in the existing customer base
When to use Retention Rate
- Tracking whether product changes improve the shape of the retention curve over successive cohorts
- Diagnosing where in the lifecycle users leak (day one, week one, or after the first billing cycle)
- Comparing retention across acquisition channels to find which sources bring users who actually stay
- Building the case for product-market fit using a flattening cohort curve rather than topline growth
- Reading a single blended retention number instead of cohort curves, which hides a recent cohort collapsing under healthy averages
- Counting any login as "retained" rather than a meaningful action, which inflates the curve without reflecting real value
- Judging retention by day-one height instead of where the curve flattens, which helps assess whether users continue to find value
- Investigate a continuing decline after onboarding with user research; the curve alone cannot tell you why users leave
- Segment retention by acquisition channel and persona to find your highest-retaining customer profile, then point acquisition at it
- Pair the quantitative curve with interviews of users who churned right after activation to learn what value they expected but never reached
Common questions
Retention rate is the percentage of customers still active at the end of a period who were active at the start, after excluding new acquisitions. Group customers by acquisition month into cohorts and track what share returns in each later period. Retention and churn are complements: they sum to 100% only for the same population, definition, and period.
A useful retention curve is interpreted in the context of its product, customer segment, and cohort age. A curve that flattens above zero can indicate that a group continues to receive value, while a continuing decline warrants investigation. Consider the curve with other evidence of product-market fit.
They describe the same movement from opposite sides: retention counts who stayed, while churn counts who left. They sum to 100% only when both use the same population, definition, and period. Retention is commonly shown in cohort curves, while churn is often used to analyze revenue impact.
A blended retention number averages new and old customers together, which can obscure differences between cohorts. Cohort analysis groups users by acquisition month so you can see whether newer customers retain differently and investigate possible changes in product or customer fit.
Go deeper: Retention Analytics: Where Users Leak
Read the full guide on Retention Rate.
Related terms
Free Retention Rate Calculator
Enter your inputs in the free Retention Analytics and review the result alongside its guidance.