Customer retention rate

Retention rate answers one question: of the customers we had at the start, how many are still with us at the end? It deliberately excludes new customers, because growth from acquisition can hide the fact that existing customers are leaving.

What counts as “retained” depends on the business model. For contracts and subscriptions it is simple — the account is still active. For businesses without contracts (stores, restaurants, clinics, hotels), retention is defined by behaviour: a customer who made at least one visit or purchase in the period, given they were active in the previous one. Agree on the definition once and keep it stable, or trends become meaningless.

Retention rate = ((E − N) ÷ S) × 100

• S — customers at the start of the period
• E — customers at the end of the period
• N — new customers acquired during the period

Example: a private bank segment starts the year with 5,000 clients (S). During the year it gains 800 new clients (N) and ends with 5,300 (E). Retention rate = (5,300 − 800) ÷ 5,000 × 100 = 90%. Churn for the same period is 10%: the 500 clients who left out of the original 5,000.

Input	Value
Customers at start (S)	5,000
New customers (N)	800
Customers at end (E)	5,300
Retained (E − N)	4,500
Retention rate	90%

Calculate it per segment and per location as well as overall. A 90% average can contain a branch at 96% and a branch at 78% — and the gap between them is where the management lesson is.

On the same base and period, retention rate and churn rate add up to 100%. They are two views of the same movement, but they are used differently: churn is the language of risk and loss analysis; retention is the language of loyalty and lifetime value. Retention is often preferred in management reporting because it rises when things get better.

Retention matters financially because keeping a customer is usually cheaper than winning a new one. Research by Frederick Reichheld with Bain & Company found that a 5-percentage-point increase in retention can raise profits by 25% to 95%, depending on the industry.

Retention responds to experience with a delay of months. To manage it, watch the indicators that move first:

• Share of detractors in NPS responses, by branch and segment.
• Complaint resolution speed — first response time and time to resolve.
• Reopened tickets — complaints that were “closed” but not actually fixed.
• Visit or purchase frequency of previously regular customers.
• Onboarding scores — early friction is a common predictor of early loss.

1. Find out why customers leave. Ask leavers directly and analyse comments from low scorers before they go.
2. Recover unhappy customers fast. Route negative feedback to the responsible manager as a ticket with a deadline; contact the customer the same day where possible.
3. Fix recurring causes, not only individual cases — the outer feedback loop.
4. Strengthen onboarding, the stage where relationships are most fragile.
5. Turn passives into promoters by finding what separates an 8 from a 9 — often speed or follow-through. See passive.

Qmeter supports this loop with automatic tickets from negative feedback, SLA tracking and escalation, a customer Health Score that flags at-risk customers, and customer segments (for example “low NPS and more than two open tickets in 30 days”) that can be targeted with follow-up campaigns.

An overall retention rate mixes long-standing customers, who rarely leave, with recent ones, who leave more often. Cohort analysis separates them: group customers by the month or quarter they joined and track each group’s retention over time. If every new cohort retains worse than the one before, something in onboarding or early experience has changed — even while the overall rate still looks stable because loyal veterans hold it up.

• Including new customers in the retained count.
• Changing the definition of “active” between periods.
• Reporting one company-wide number instead of segments and locations.
• Waiting for the annual figure instead of managing the leading indicators every week.

Our guide to churn rate and retention covers cohort analysis and a retention playbook in more depth.