# Customer Churn

Customer churn measures customers or accounts lost from an eligible base during a defined period, usually as a count or percentage.

- Type: metric
- Domain: Data, experimentation & growth
- Level: Intermediate
- Last reviewed: 2026-09-21
- Original: [Soroush’s Product Wiki](<https://wiki.soroush.uk/topic/customer-churn/>)

## What it means

Customer churn is the loss of customers, subscribers, or accounts from a starting population over time. A common logo-churn rate divides customers lost during a period by customers present at its start. Revenue churn instead tracks recurring revenue lost through cancellation or contraction, and net revenue retention additionally reflects expansion. These measures should not be blended because they describe different business dynamics.

## Why it matters

Churn determines how much acquisition must replace before a product can grow. Examining who leaves, when, and after which experiences can expose poor fit, unreliable service, pricing friction, failed onboarding, or unmet needs. Separating voluntary cancellation from failed payments or administrative closure also points to different remedies.

## When to use it

Use customer churn for subscriptions, memberships, marketplaces with persistent accounts, and other relationships where an end state can be defined. Select a period aligned with billing and decision cycles. For noncontractual products, behavioral inactivity or survival analysis may be more informative than declaring a customer permanently lost.

## How it works

1. Define the customer entity, eligible starting base, churn event, period, grace rules, and treatment of pauses, migrations, and reactivations.
2. Separate voluntary, involuntary, preventable, and structural churn where the data supports those distinctions.
3. Calculate lost starting customers divided by customers at the period start; report revenue-based measures separately.
4. Analyze churn by acquisition cohort, tenure, segment, plan, activation status, and preceding product or support events.
5. Use interviews, cancellation feedback, and controlled changes to test causes rather than assuming the final recorded event explains departure.

## A practical example

A project-management service sees stable account churn but rising revenue churn. Segmentation shows that a few larger customers are downgrading after reporting permission-management problems. The team improves role setup and migration support, then evaluates renewal and contraction among comparable upcoming cohorts instead of using save offers as the only response.

## Where it breaks

Churn is a lagging outcome and may arrive long after the experience that caused it. Annual contracts delay observation, while failed-payment recovery can reverse apparent losses. Small high-value segments make logo and revenue churn diverge sharply. Cancellation surveys suffer from nonresponse and socially convenient answers, so they should be triangulated with behavior and interviews.

## Common mistakes

- Using end-of-period customers as the denominator while comparing with a start-of-period churn definition.
- Combining customer churn, gross revenue churn, and net revenue retention as if they were interchangeable.
- Treating discounts that postpone cancellation as evidence that the underlying product problem is solved.

## Visualization

[View the visualization on the website](<https://wiki.soroush.uk/topic/customer-churn/#visualization-customer-churn>)

## Sources

1. [Churn rate 101: Customer churn explained](<https://stripe.com/resources/more/customer-churn-rates-101>) — Stripe · Payments industry guidance
2. [How to calculate monthly churn and what it can tell you](<https://stripe.com/resources/more/monthly-churn-101>) — Stripe · Payments industry guidance

## Related topics

- [Retention Rate](<https://wiki.soroush.uk/topic/retention-rate/>)
- [Cohort Analysis](<https://wiki.soroush.uk/topic/cohort-analysis/>)
- [Customer Effort Score \(CES\)](<https://wiki.soroush.uk/topic/customer-effort-score/>)
