Retention rate
Retention rate is the percentage of customers or users still with you at the end of a period. It mirrors churn rate and is best tracked by cohort.
Retention rate is the share of customers you kept. If you start a period with 200 customers and 190 are still paying at the end, retention is 95%. It carries the same information as churn rate (retention is 100% minus churn), but people reach for it when they want to talk about how long customers stay and how good the product is, not about the loss.
How to calculate retention rate
You start a quarter with 200 customers, sign 40 new ones and end with 230. Retained customers are 230 - 40 = 190, so retention is 190 / 200 = 95%, and quarterly churn is 5%. The subtraction of new customers matters. Without it you would report 115%.
Customer retention vs revenue retention
Retention can be counted in accounts or in dollars. Dollar versions are gross revenue retention, which counts only losses, and net revenue retention, which adds expansion and can pass 100%. Plain retention rate usually means the account version.
User retention and customer retention
For products with free users, retention often refers to active users coming back, not paying customers staying subscribed. Lenny Rachitsky's study of what good retention looks like measures it as the share of users still active after 6 months, and gives these targets as good and great:
- Consumer SaaS: 40% and 70%
- SMB and mid-market SaaS: 60% and 80%
- Enterprise SaaS: 75% and 90%
The lower the price and the smaller the customer, the lower the retention you should expect. The study also notes that weaker retention can be acceptable if acquisition is cheap or you are not chasing venture-scale growth.
Why a single number is not enough
A blended retention rate mixes customers who joined last week with customers who joined three years ago, and those two groups behave very differently. Sequoia's data science team, in its retention essay, looks at retention by cohort and by curve. See retention curve and cohort analysis.
Lost customers who come back
Retention rate treats a cancellation as final, but a share of churned customers return. Kyle Poyar's September 2026 analysis of ChartMogul data (10.5 million purchases across 4,029 software companies, mostly self-serve plans under $100 a month) found that 2.67% of churned customers reactivated in the first month after cancelling, 1.24% in month two and 1.14% in month three. That is more than 5% within three months. AI-native products saw nearly 10% of churners return, which Poyar suspects comes from AI FOMO and project-based use.
Churn isn’t necessarily a goodbye.
Kyle Poyar / The 30-day upsell window (September 30, 2026)
Two practical consequences:
- Decide how a returning customer counts. Count them as retained in their original cohort, or as a new customer, and stick to the rule. The revenue side has its own metric, reactivation MRR.
- Make coming back easy. Poyar suggests keeping a cancelled customer's workspace, data, settings and integrations so returning takes one click, and running a win-back sequence at days 1, 30 and 60 after cancellation.
Common mistakes
- Leaving new customers in the numerator.
- Comparing a monthly rate with an annual one. 97% monthly retention is about 69% annual, since 0.97 to the 12th power is 0.694.
- Counting a paused or downgraded customer as lost, or as fully retained, without deciding on a rule first.
For small SaaS
Pick a single definition, such as paying customers at day 90, and report it for each monthly signup group. With a few hundred customers a month-to-month change of a point or two is noise, so judge by trend across several cohorts.
Listen, watch and read next
- Watch: SaaS Product Design: How a B2C App Got to 4% Churn, Jacob Wright at MicroConf Europe 2026 (published October 26, 2026). The founder of Dabble, novel-writing software used mostly by hobbyists, shows the product decisions behind a B2C app with nearly $2M ARR.
- Read: The 30-day upsell window, Kyle Poyar (September 30, 2026). Reactivation and upgrade rates by month, with concrete win-back plays.
- Listen: Key Learnings from MicroConf Europe 2026, Startups for the Rest of Us, episode 856 (October 27, 2026). See "Jacob Wright on UX as a moat" (19:44) for Rob Walling and Derrick Reimer's take on the Dabble talk.
- Watch: $1M ARR, how to expand into a new market, Jason Cohen with Rob Walling (August 15, 2026). A founder with 2,000 happy customers, 2% monthly churn and a plateaued growth curve, and what to do next.
Related terms
- Churn rate
- Retention curve
- Cohort analysis
- NRR (Net revenue retention)
- GRR (Gross revenue retention)
- Customer churn
Sources
- What is good retention?, Lenny Rachitsky, Lenny's Newsletter
- Retention, Sequoia Capital Data Science Team
- Cohort analysis, ChartMogul Help Center
- The 30-day upsell window, Kyle Poyar, Growth Unhinged, September 30, 2026