{"id":2803,"date":"2026-08-20T12:43:16","date_gmt":"2026-08-20T10:43:16","guid":{"rendered":"https:\/\/mairateam.com\/?post_type=glossary&p=2803"},"modified":"2026-08-20T15:56:06","modified_gmt":"2026-08-20T13:56:06","slug":"churn","status":"publish","type":"glossary","link":"https:\/\/mairateam.com\/en\/glossary\/churn\/","title":{"rendered":"Churn"},"content":{"rendered":"

Churn<\/strong> is the share of customers who left during a period: cancelled the subscription, stopped buying, let the contract lapse. It is the negative side of retention, and in subscription models the single most important number: it sets how long the average customer keeps paying, and with it the entire value of acquisition.<\/p>\n

In short<\/h2>\n\n\n\n\n\n\n
Formula<\/strong><\/td>\nChurn = customers lost in the period \/ customers at the start of the period<\/td>\n<\/tr>\n
Consequence for LTV<\/strong><\/td>\nAverage relationship length is roughly 1 \/ churn. A 5% monthly churn means about 20 months<\/td>\n<\/tr>\n
Two forms<\/strong><\/td>\nCustomer churn counts people, revenue churn counts money. They can differ substantially<\/td>\n<\/tr>\n
Measuring it without subscriptions<\/strong><\/td>\nIn e-commerce, as the share of customers who did not buy within their usual cycle<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

How it is calculated<\/h2>\n

Customer churn and revenue churn<\/h3>\n

The basic formula uses two numbers for one period: the number of customers at the start and the number of those who left during the period. Customers acquired during the period do not belong in the calculation: they are not in the base, and even if they leave within the same period, they are not counted among that period's departures. Otherwise new acquisition dilutes churn and it comes out lower than it is.<\/p>\n

Worked example: a service has 2,000 paying customers at the start of the month. During the month, 100 of them cancel their subscription and 300 new ones join. Customer churn = 100 \/ 2,000 = 5%. The fact that the month ends with 2,200 customers (2,000 minus 100 plus 300) changes nothing about churn: the base is growing, and yet one in twenty leaves every month.<\/p>\n

Revenue churn calculates the same thing in revenue terms: monthly recurring revenue (MRR) lost during the period divided by MRR at the start. If the 2,000 customers together paid \u20ac120,000 a month and the 100 who left carried \u20ac9,000, revenue churn = 9,000 \/ 120,000 = 7.5%. So the departing customers are larger than average, and revenue churn is higher than customer churn. Gross revenue churn also includes downgrades; net revenue churn subtracts upsell to existing customers from it.<\/p>\n

Converting between month and year<\/h3>\n

Monthly churn is not multiplied by twelve to get annual churn, because the base shrinks every month. Annual churn = 1 minus (1 minus monthly churn)^12. At 5% a month: 0.95^12 = 0.54, so annual churn is 1 minus 0.54 = 46%, not 60%. In the other direction: monthly churn = 1 minus (1 minus annual churn)^(1\/12).<\/p>\n\n\n\n\n\n\n\n
Monthly churn<\/strong><\/td>\nAnnual churn<\/strong><\/td>\n<\/tr>\n
2%<\/td>\n21.5%<\/td>\n<\/tr>\n
3%<\/td>\n30.6%<\/td>\n<\/tr>\n
5%<\/td>\n46.0%<\/td>\n<\/tr>\n
10%<\/td>\n71.8%<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Cohorts and e-commerce without subscriptions<\/h3>\n

A single aggregate number mixes customers one month old with customers three years old. That is why churn is calculated by cohort: the group of customers acquired in the same month is tracked separately, and churn is reported for month 1, 2, 3 and onward from acquisition. Only a comparison of cohorts shows whether departures concentrate at the start of the relationship and whether newer cohorts behave differently from older ones.<\/p>\n

In e-commerce without subscriptions there is no \"cancellation\", so leaving is defined by inactivity: a customer who has not bought within a window derived from the usual purchase cycle counts as churned. The calculation: out of a cohort of 1,000 customers who first bought in January, 380 bought again within 12 months. Cohort churn = (1,000 minus 380) \/ 1,000 = 62%, retention 38%. The window must be the same for all cohorts; otherwise the numbers cannot be compared.<\/p>\n

Why a small churn difference changes the whole business<\/h2>\n

A worked example: a subscription of 50 a month. At 5 percent monthly churn the average customer pays for about 20 months and brings 1,000. At 3 percent they pay for about 33 months and bring 1,650. Two percentage points of difference raise customer value by two thirds, and with it the ceiling of what acquisition may cost. No campaign optimization can do as much for the economics as shaving a few points off churn.<\/p>\n

From our own practice<\/h2>\n

Churn is the reason our audits of subscription and service models start with the retention curve, not with campaigns: when customers leave after the first month, acquisition just fills a leaking bucket, and the right investment is product and onboarding, not media. The opposite direction holds too: low churn is the argument for more aggressive acquisition, because every customer gained keeps paying for a long time. We are our own example: 90 percent of our clients stay with us five years or more, and that is exactly what lets us invest more into the relationship than one-off agencies can.<\/p>\n

Common mistakes<\/h2>\n