Subscriptions

Subscription churn rate: how to calculate it and read it properly.

Churn rate looks like the simplest number in your dashboard. Then three tools show you three different values. Here's how to calculate it yourself and know what you're looking at.

Your subscription churn rate is the share of subscribers you lose in a period, usually a month. The idea is simple. The details aren't, and that's why your subscription app, your spreadsheet and your agency can all show a different number for the same month. This guide covers the formula, the difference between monthly and cohort churn, why you should always split voluntary from involuntary churn, and the mistakes we see most when brands calculate it themselves.

The basic subscription churn rate formula

For a single month:

Monthly churn rate = subscriptions that ended during the month / active subscriptions at the start of the month

Multiply by 100 and you have a percentage. That's it for the formula. Most of the work is in deciding what counts.

A few definitions you should write down before you calculate anything:

  • What is "active"? Usually a subscription that has a next charge date scheduled. Paused subscriptions are a judgment call, more on that below.
  • What is "ended"? A cancellation by the subscriber, plus subscriptions ended by the system after failed payments.
  • Subscriptions or customers? One customer can have two subscriptions. Pick one unit and stick to it. We usually report on subscribers (customers with at least one active subscription) and look at subscriptions separately when needed.

Why the start of the month as the denominator? Because those are the people who could have churned during the whole month. If you use the end of the month, new signups dilute the number, and a growing brand looks better than it is.

A worked example

The numbers below are made up and round on purpose, so the math is easy to follow. They are not a benchmark.

Example monthCount
Active subscribers on day 11,000
Cancelled by the subscriber80
Ended after failed payments20
New subscribers during the month150
Active subscribers at month end1,050

Total churn: (80 + 20) / 1,000 = 10%.

Voluntary churn: 80 / 1,000 = 8%. Involuntary churn: 20 / 1,000 = 2%.

Now the common mistake. Divide the same 100 by the 1,050 at month end and you get about 9.5%. Same subscribers, same month, half a point less churn, just because you grew. That's why the denominator matters.

The example also assumes none of the 150 new subscribers cancelled in their first month. In real life some will. You have two options: count them in the numerator and accept a slightly inflated rate, or track first month cancellations as a separate number. We prefer the second, because first month cancels usually point to an acquisition or expectation problem, and they deserve their own line.

Voluntary vs involuntary churn

Always split these. One total number hides two very different problems.

Voluntary churn is a subscriber who clicked cancel. You fight it with better product fit, onboarding, renewal reminders and a cancellation flow that matches the offer to the reason.

Involuntary churn is a subscription that ended because a payment failed and nobody fixed it. You fight it with retries in the subscription app and a failed payment recovery sequence in Klaviyo.

If you only see the total, you can't tell whether a better month came from the new cancel flow or from fewer expired cards. And you'll put work in the wrong place.

Monthly churn vs cohort churn

Monthly churn tells you how your whole base behaved last month. It's the number for your dashboard and your MRR forecast.

Cohort churn follows one group of subscribers, say everyone who started in March, and shows how many are left after month one, two, three and so on. You get a retention curve for each starting month.

Why bother? Because subscribers don't churn evenly. In most subscription businesses, churn is highest in the first few orders and lower after that. A brand that's growing fast has lots of new subscribers in the base, so its monthly churn looks worse. A brand that stopped acquiring has mostly loyal long term subscribers left, so its monthly churn looks great while the business shrinks.

Cohorts also show you whether changes work. If you rebuilt onboarding in May, compare the May cohort's month two retention with April's. A blended monthly number would bury that difference.

Many subscription apps have a cohort or retention report built in. If yours doesn't, export subscriptions with a start date and an end date and build a simple table in a spreadsheet: start months down the side, months since start across the top.

Common mistakes

Dividing by the end of month count

Covered above. It makes a growing brand look better and a shrinking one look worse.

Changing the definition halfway through the year

If you count pauses as churn in January and stop in June, your churn drops in June and nothing actually happened. Write the rules down.

Ignoring pauses that never come back

A pause isn't churn on the day it starts. But a pause that runs out and gets cancelled, or a pause that never ends, is churn with a delay. Track how many pauses restart. If that share is low, your churn rate is flattering you.

Counting skips as churn

A skip is a subscriber staying. It lowers revenue for one cycle, which shows up in MRR, not in churn.

Treating prepaid plans like monthly ones

A subscriber on a six month prepaid plan can only really churn at renewal. If prepaid plans are a big share of your base, look at their renewal rate separately. Mixing them in makes monthly churn look calmer than it is between renewals and spikier in renewal months. Our guide to prepaid subscriptions covers this.

Not counting reactivations correctly

Someone who cancels in March and restarts in May is a churned subscriber in March and a reactivation in May. Don't quietly net them out, or you lose sight of how well your winback works.

Mixing subscriptions and customers

A customer with a coffee subscription and a filter subscription who cancels the filters is still a subscriber. If you count subscriptions, that's churn. If you count customers, it isn't. Both are valid, but a report that mixes the two can't be compared month to month. Pick one for the headline number and track the other on the side.

Looking at one month in isolation

Churn jumps around, especially with a small base. A single bad month after a price change or a shipping problem says less than the three month trend. Look at both.

Where to find the numbers

Your subscription app is the source of truth for active subscriptions, cancellations and failed payment endings. Recharge, Loop, Skio and the others all have some reporting on this, though the names and definitions differ. Check how your app defines churn before you compare it with anything else.

Klaviyo gets the subscription events through the integration, so you can count cancellations and failed payments there too, and see the cancellation reason if your app sends it. It's useful for checking your flows, but use the app for the official number.

For a full list of the numbers we track and where each one lives, see our guide to subscription retention metrics.

What to do with your churn rate

A churn rate on its own is just a number. Two things make it useful.

First, turn it into money. Lifetime value depends heavily on churn, and our guide to subscription LTV shows the math. Or put your numbers into the LTV calculator and see what a point less churn is worth.

Second, break it down until you know where to act: voluntary vs involuntary, then voluntary by cancel reason, then by cohort. When you know which bucket is biggest, our pillar guide on how to reduce subscription churn tells you which lever to pull.

Quick checkAsk your subscription app and your own spreadsheet for last month's churn. If the two numbers differ, find out why before you report either. It's almost always the denominator or the way pauses are counted.

Questions we get a lot

What is the formula for subscription churn rate?

Monthly churn rate is the number of subscriptions that ended during the month divided by the number of active subscriptions at the start of that month. Multiply by 100 for a percentage. Leave subscribers who signed up during the month out of the denominator, and decide up front how you treat those who join and cancel in the same month.

What is the difference between voluntary and involuntary churn?

Voluntary churn is a subscriber actively cancelling. Involuntary churn is a subscription ending because a payment failed and was never fixed. Track them separately, because they have completely different fixes: the cancel flow for one, failed payment recovery for the other.

What is cohort churn and why does it matter?

Cohort churn follows a group of subscribers who started in the same month and shows how many are still active after one, two, three months and so on. It shows early churn that a blended monthly number hides, especially when you are growing fast and new subscribers make up a big part of the base.

Do paused subscriptions count as churn?

Most brands do not count a pause as churn, but you should track pauses that never restart. Whatever you decide, write the rule down and keep it the same every month, otherwise your churn rate moves for reasons that have nothing to do with your subscribers.

Want to see where your extra revenue is?

Book a free 30 minute audit. We go through your Klaviyo together and show you what we would change first.

Get a free audit