How to Calculate Customer Retention Rate Without Counting New Customers
By Ulas ArslanPublished Updated
Customer retention rate measures the share of your starting customer group that is still a customer at the end of a period. Count the same customer units at both dates and keep new customers separate. For subscription teams, following customer IDs directly is often clearer than relying only on three headline totals.
Define the customer before you count
A CRM contact is not necessarily a paying customer. A company can have many contacts, several subscriptions or multiple legal entities. Decide which unit your retention metric follows: a paying company, billing account or subscription. Keep that choice consistent.
Also define when a customer becomes active and when it leaves. A cancellation request and the end of a paid subscription can be different dates. For a paid-customer measure, use the rule agreed with your revenue reporting owner and document how trials and overdue accounts are treated.
Calculate retention with a fixed cohort
Imagine a fictional business with 200 paying customer accounts on 1 August. At the end of August, 188 of those same accounts are still active. Customer retention is 188 ÷ 200 × 100 = 94%. The twelve accounts that left represent 6% customer churn for this cohort and period.
The business also acquired 30 customers during August, of which 27 remained at month end. Its ending total is therefore 188 + 27 = 215. New sales have increased the total customer count even though the original group has lost customers.
| Customer group | Count |
|---|---|
| Customers at the start | 200 |
| Starting customers still active at the end | 188 |
| New customers acquired during the month | 30 |
| New customers still active at the end | 27 |
| Total customers at the end | 215 |
| Retention of the starting group | 94% |
Why the shortcut formula needs care
HubSpot’s retention-rate guide presents the familiar ending-minus-new formula. It is a useful shortcut when the customer counts reconcile. In the example above, subtracting 27 new customers still present gives (215 − 27) ÷ 200 = 94%.
Subtracting all 30 customers acquired would produce 92.5%. That understates retention of the original group because three new customers have already disappeared from the ending count. The safest calculation matches the starting customer IDs directly to their status at the end.
This distinction matters when acquisition and early churn both occur during the same period. It also explains why two dashboards can report different retention rates from apparently similar customer totals.
Handle reactivations, mergers and duplicates explicitly
An end-of-period measure can count a starting customer that left and returned before the period ended as retained. A continuous-retention measure may treat the interruption differently. Choose the definition that answers your question and show the rule beside the report.
Account mergers and duplicate CRM records need similar care. If two company records become one, the apparent decrease should not automatically be treated as customer loss. Keep a stable reporting identity or a mapping from old IDs to the continuing account.
In HubSpot, start with the company or billing relationship that actually represents the customer. Reconcile it with your subscription or finance source. A lifecycle label can help organize records, but it is not a complete record of paid-customer status over time.
Read the result beside revenue and customer stage
Calculate the same measure for onboarding cohorts, mature accounts and relevant service segments. A blended figure can hide early customer loss or a problem limited to one plan. Use the same definition in every comparison.
Next, review GRR and NRR to see how much recurring revenue stayed. Do not insert a customer-count retention rate into a revenue model unless you explicitly assume comparable customer values. Losing five small accounts is not financially equivalent to losing five large ones.
For a forward-looking revenue scenario, CLE Index makes the model’s scope visible. For an operational response, a customer success playbook helps turn an identified issue into an owned next step. Neither replaces a correctly defined retention cohort.