100% net revenue retention does not mean every customer stayed

A business can report 100 percent net revenue retention after losing a customer. The measure follows recurring revenue from an existing group, including expansion within that group. It does not say that everyone stayed, that every account kept its old spending or that the company collected the money.

Keep the original five customers

Consider an invented subscription service. At the opening measurement date, five distinct customers each have a $200 monthly subscription. Together they represent $1,000 a month. At the closing date one month later, their subscriptions have changed as follows:

• Customer A leaves, reducing its monthly subscription from $200 to zero.
• Customer B reduces its subscription from $200 to $100.
• Customer C expands from $200 to $500.
• Customers D and E remain at $200 each.

There are no reactivations, one-time charges, currency changes, credits or other adjustments. These amounts are the complete agreed monthly subscription rates at each measurement date, not invoices issued during the month, earned accounting revenue or cash received. All rates use the same currency and monthly basis.

The original group now contains four active customers and still supplies $1,000 of monthly recurring value: zero, $100, $500, $200 and $200. The $300 expansion offsets the $200 departure and $100 reduction exactly.

Three retention questions

How many original customers remain? Four of the original five, or 80 percent. That is customer retention for this fixed group and interval. Remaining a customer and keeping the same subscription are separate observations.

How much of the original recurring value remains before expansion? Subtract the $200 lost with A and the $100 reduction at B from the original $1,000. The result is $700, or 70 percent. In this example, gross revenue retention excludes expansion and includes losses from both departures and reduced subscriptions.

How much recurring value does the original group supply after expansion? Add C's $300 expansion to the $700. The result is $1,000, giving net revenue retention of 100 percent.

Stripe's explanation of revenue retention sets out this distinction between a calculation excluding expansion and one including it. Here, 70 percent gross and 100 percent net describe the same original group. Neither should be reported as keeping every customer.

New customers answer a different question

Now suppose two entirely new customers join before the closing measurement date, with monthly subscriptions of $300 and $400. Neither belonged to the opening five. Their $700 raises the closing total across all customers to $1,700.

The total monthly recurring value has grown 70 percent from $1,000. There are six active customers rather than five, a net gain of one. But original-customer retention remains 80 percent, gross revenue retention remains 70 percent under the definition above, and net revenue retention remains 100 percent. The new customers do not enter those retention numerators.

Putting all $1,700 over the opening $1,000 would produce 170 percent. That is the closing total as a percentage of the opening total, including new business. It is not the net retention of the original five.

From a recurring rate to work you can rely on

A recurring rate is a useful starting point for a reliability review. Follow it through the evidence for the next service period, the collection date and the operation required to deliver. An attractive snapshot cannot supply a missing customer decision.

Take a separate, entirely fictional product-catalog maintenance studio. On November 30, 2026, its rate list shows three customers at $300 a month: $900 in total. Each monthly package covers checking and updating twelve product records and returning a checked proof. Each customer's supplied agreement requires a new accepted order for the next month; there is no automatic renewal. The November rate list therefore does not itself authorize a December invoice.

By December 1, two customers have accepted December packages at $300 each. The third has made no December decision. The studio issues two invoices totaling $600, due December 10. It has no December order, invoice or delivery commitment for the third customer. These are facts supplied for this case, not assumptions about cancellation rights or customer intent.

Evidence for this studio's December plan

Scroll across to follow each question and its remaining limit. Use the arrow keys when the table area is focused.

The delivery plan allows four hours per package, including communication, updates and checking. All twelve available hours can be scheduled between December 1 and December 8; there are no other commitments on that capacity. All twelve available hours are paid at a fully loaded $30 an hour, so the month's payroll is $360 even with only eight hours assigned. Materials and software usage cost $60 per accepted package, or $120. Workspace and administration cost another $120. The studio starts December with $800, pays the $120 usage cost on December 2 and the $480 payroll and overhead on December 5, before collecting either invoice. No unpaid owner work fills the plan.

Those are all costs and cash movements through December 31 in this limited model, before income tax, with no financing, equipment purchase or other work. The $600 of payments leave $200 before collections. The two actual $300 receipts restore cash to $800 by December 20. The studio can fund the late receipt here because it started with the money to pay for delivery. The recurring-rate total did not provide that money.

Keep the third customer's next-period choice unresolved. Its historical rate does not explain whether it will return, when it would pay or what another order would cost. The existing guide to revenue, profit and cash flow follows the accounting distinctions. This reliability review asks for the underlying orders, receipts and delivery resources before treating a recurring figure as support for the next commitment.

Read the definition beside the label

A report still needs its own calculation read back. In its March 2026 operational-measure definitions, CrowdStrike describes a twelve-month comparison and excludes new subscription customers from net retention. Its gross-retention definition deducts recurring value lost with customers who left; that differs from this example's definition, which also deducts contractions at customers who remain.

The same label can therefore cover different calculations. Before comparing two reported percentages, check the customer group, interval, recurring-value basis and treatment of expansion, contraction and returning customers. A one-month illustration is not a twelve-month company result.

The numbers also leave the reasons open. C may need more service, have accepted a price increase or have moved from another arrangement. A's departure needs its own explanation. A 100 percent net figure cannot establish satisfaction, profitability, future renewals or the ability to deliver the larger subscription.

Keep the original group visible alongside the whole business. The reading on new customers and net growth follows the setup work a rising total can conceal. This comparison asks what happened inside the customers already there.