How to calculate customer retention rate (CRR) correctly, the difference between customer retention and revenue retention, the six retention KPIs worth tracking and the formula for each, what a healthy SaaS retention curve looks like by cohort, and the two mistakes (new customers in the numerator, trials in the cohort) that inflate the number.
Short answer
How do you calculate customer retention rate?
CRR = (customers at end − customers acquired during the period) ÷ customers at start × 100. Subtracting the new customers is the step that matters; skip it and a growing company reports retention above 100%. Measure it on one cohort over one period, exclude trials that never paid, and read it next to revenue retention, because the customers who leave are rarely average-sized.
Customer retention rate is the metric most SaaS teams compute first and most get subtly wrong, usually by leaving new customers in the count. This is the formula, a worked example, the six KPIs that belong next to it, and what the retention curve should look like once you plot it by cohort.
Start of period: 400 paying customers. End of period: 430. Acquired during the period: 60. The tempting calculation is 430 ÷ 400 = 107.5%, which is impossible. The right one removes the newcomers: (430 − 60) ÷ 400 = 92.5%. Thirty of the original 400 left.
Two more rules. Count only customers who had paid at least once; trials that expired were never retained or lost. And fix the cohort on day one: a customer acquired in month two who cancels in month nine is not in this period's calculation at all. They belong to the next cohort's.
On the same cohort and period, customer retention rate and logo churn rate sum to 100%. 92.5% retention is 7.5% churn. They are one fact with two names, and the only reason to prefer one is emphasis: retention for a board deck, churn for a team that has to act on it. The pairs that are not interchangeable are covered in churn rate vs retention rate: monthly versus annual, and customers versus revenue.
Customer retention counts accounts. Revenue retention counts money. On a book where every customer pays the same, they are equal. On every real book they diverge, because the customers who leave are not average: in self-serve SaaS the small ones churn most, so revenue retention runs above customer retention; in enterprise SaaS one large loss can push revenue retention below customer retention for a quarter. Track both. The revenue versions are GRR and NRR.
| KPI | Formula | Answers |
|---|---|---|
| Customer retention rate | (end − new) ÷ start | Did the customers stay? |
| Gross revenue retention | (start MRR − churned − downgraded) ÷ start MRR | Did the money stay? |
| Net revenue retention | (start MRR − churned − downgraded + expansion) ÷ start MRR | Did the base grow on its own? |
| Involuntary churn share | MRR lost to failed payments ÷ total churned MRR | How much of the loss was never a decision? |
| Time to first value | Median days from signup to the first outcome the customer bought for | How long before a customer has a reason to stay? |
| Cohort retention at month N | Cohort still active at month N ÷ cohort size | When in the lifecycle do customers leave? |
The fourth row is the one most scorecards lack, and it changes what the others mean. If a third of churned revenue was cards that failed and were never fixed, then a third of the retention problem is a billing-operations problem with a known fix, and the retention rate overstates how many customers actually chose to leave.
Plot each acquisition month as a line: the share of that cohort still paying at month 1, 2, 3 and onward. Three shapes appear.
Building the table that produces the curve is a spreadsheet exercise; the cohort analysis guide walks through the columns, the SQL, and how to read it.
Customer retention rate (CRR) = (customers at end of period − customers acquired during the period) ÷ customers at start of period × 100. Subtracting the new customers is the step people skip; without it, a growing company reports a retention rate above 100%. If you started with 400 customers, ended with 430, and acquired 60 during the period, CRR is (430 − 60) ÷ 400 = 92.5%.
In Excel, put customers at start in A2, customers at end in B2, and customers acquired during the period in C2, then CRR is =(B2-C2)/A2. Format the cell as a percentage. For a cohort table, put one acquisition month per row and months-since-acquisition across the columns, with each cell holding the share of that month's cohort still active; the retention curve is the average down each column.
A good customer retention rate for SaaS depends on customer size: around 90-95% annually for enterprise, 85-90% for mid-market, and 60-75% annually for SMB and self-serve, where monthly retention of 95-97% is the equivalent figure. Revenue retention is usually higher than customer retention in the same business, because the customers who leave are smaller than average.
Customer retention counts accounts: the share of customers still active. Revenue retention counts money: the share of recurring revenue still being paid. They differ whenever customers differ in size. Losing ten $50 accounts and keeping one $5,000 account is 90% customer retention and 99% revenue retention. Track both; a board wants revenue retention, a product team needs customer retention.
Alongside customer retention rate, track gross revenue retention, net revenue retention, logo churn rate, involuntary churn share (the part caused by failed payments), time to first value, and the retention curve by cohort. Together they answer the four questions CRR alone cannot: how much money left, whether expansion covered it, whether the losses were decisions, and when in the customer lifecycle the losses happen.
A healthy SaaS retention curve drops in the first one to three months, then flattens into a near-horizontal line: early-life churn removes the customers who never activated, and the survivors stay. A curve that keeps sloping downward at the same rate every month means the product is not retaining anyone durably. A curve that flattens above 70-80% for B2B is strong; one that flattens below 50% means the onboarding is losing half the customers who paid.
The benchmark ranges are industry ranges, not a measurement Exeechain took. The KPI table reflects what Exeechain computes from a connected billing account: customer and revenue retention with the involuntary share separated, because that split decides whether the fix is a save conversation or a payment recovery sequence.
Keep reading
Retention metrics
9 min read · Sep 18, 2026
Retention metrics
9 min read · Sep 18, 2026
Customer success
10 min read · Sep 18, 2026
First scores in 15 minutes. Full accuracy in 24 hours. From $299/mo. Customer and revenue retention from billing, involuntary churn separated, saves drafted for approval.
Not ready to switch? Size your leak from your MRR and churn rate, with no billing access at all.