The Churn and Retention Formulas

5,000 customers at the start of the month, 300 gone by the end: churn is 6.0%, retention is 94.0%. From there, average customer lifespan is 1 ÷ churn — about 16.7 months — which is the number that turns a retention metric into a lifetime value.

Why 2% a Month Is Not 24% a Year

This is the single most common churn error, and it is in a lot of published calculators.

Churn does not add up across periods. It compounds downward. If 98% of your customers survive each month, the share surviving twelve months is 0.98 multiplied by itself twelve times — 78.5%. So annual churn is 21.5%, not 24%.

The gap widens sharply as the rate rises:

Any churn figure above 100% is a multiplication error. You cannot lose more customers than you have.

The Denominator Problem

Two teams can compute honest churn on identical data and disagree, because there are two conventions for what goes underneath.

• Customers at the start. The standard definition. New customers acquired during the period are excluded, on the reasoning that somebody who joined on the 28th barely had a chance to leave.
• Average of start and end. Fairer when you are growing fast, because it acknowledges that the base was larger than the opening figure for most of the period. It always gives a slightly lower number — 5.91% against 6.00% on the worked example.

Neither is wrong. Switching between them month to month is, because it manufactures a trend that is not there. The calculator makes it an explicit toggle so you can see the size of the difference on your own numbers before you pick one and stay with it.

Customer Churn vs Revenue Churn

Counting customers and counting revenue can point in opposite directions, and the gap is usually where the story is.

• Gross revenue churn — cancellations plus downgrades, over starting revenue. What you actually lost.
• Net revenue churn — the same, minus expansion from existing customers. Can go negative.
• Gross revenue retention (GRR) — 100% minus gross churn. Cannot exceed 100%.
• Net revenue retention (NRR) — includes expansion. Above 100% means the base grows on its own.

On $250,000 of starting revenue, $9,000 cancelled, $3,500 contracted and $14,000 expanded: gross churn 5.0%, net churn −0.6%, GRR 95.0%, NRR 100.6%.

That business has net negative churn and is compounding. It also still loses 5% of its revenue base every month, and only expansion is hiding it. Track both, because net retention tells you whether you are growing and gross retention tells you what you are papering over.

What Counts as a Good Churn Rate

Published benchmarks for churn are close to useless, because the number depends on contract length, price point, category and how you define an active customer. A consumables brand and an enterprise contract are not comparable in any useful way.

Use the structure instead:

At $42 a month, a 65% margin and 6% monthly churn, lifetime value is $455. At 4% churn it is $682. If acquisition costs $85, the ratio moves from 5.4x to 8.0x.

A churn rate is good if the LTV it produces clears your acquisition cost by enough margin to fund the business. A high-churn business acquiring at $12 can be far healthier than a low-churn one paying $400. Compare yourself to your own economics before comparing yourself to a benchmark.

What Churn Does to Your Acquisition Budget

Churn shows up in the ad account as a budget that never seems to buy growth.

At 5,000 customers and 6% monthly churn you lose 3,600 customers a year. Replacing them at $85 each is $306,000 of acquisition spend that produces zero net growth. Take churn to 4% and that bill falls to $204,000 — $102,000 a year released from standing still into actually growing.

Two consequences worth being deliberate about:

• Retention work and acquisition work compete for the same budget, and retention usually wins on price. A point of churn is normally cheaper to fix than the equivalent volume of new customers is to buy.
• Lower churn raises what you can afford to pay per customer. Longer lifespan means higher LTV, which means you can outbid competitors in the same auction and still make money. Retention is an acquisition advantage, not just a finance metric.

Run the downstream numbers with the LTV calculator, the LTV:CAC calculator and the payback period calculator — all use the same definitions as this page.

What This Calculator Cannot Tell You

• It assumes a constant churn rate. Real churn is front-loaded — most cancellations happen in the first few periods. The 1 ÷ churn lifespan is a reasonable planning figure and an unreliable prediction for any individual cohort.
• It does not do cohort analysis. A blended rate can hide a serious problem in recent cohorts behind a healthy older base. If your blended churn is stable but new-cohort churn is rising, you find out late.
• It cannot tell you why. Churn is an outcome. Price, onboarding, product fit, delivery times and involuntary churn from failed payments all land in the same number, and the fixes are completely different.
• Involuntary churn is included but not separated. Failed card payments are often 20–40% of consumer subscription churn and are a dunning problem, not a satisfaction problem. Worth splitting out before you go and rebuild the product.
• It does not model reactivation. Customers who come back are neither churned nor new in most reporting, and how you treat them changes the number.