Customer churn vs revenue churn

Customer churn is the share of your customers who left in a period. Revenue churn is the share of your MRR that left, through cancellations and usually downgrades too. The first tells you whether people stay, and the second how much their leaving costs. With a small customer base and uneven prices, one cancellation can make the two disagree completely, which is why it helps to know what each one is for before you pick one to watch.

A quick example

You have 40 customers on a $20 plan and one on a custom $400 plan. That's 41 customers and $1,200 MRR.

The $400 customer cancels. Nobody else does.

Almost everyone stayed, and a third of the MRR still walked out with one customer. Either number alone only shows you half of that.

How do you calculate each one?

Customer churn counts people. Take the paying customers at the start of a period, count how many of them aren't paying at the end, and divide. Over a longer window you convert it to a monthly rate (the full method, with a script).

Gross revenue churn counts money lost from the customers you already had:

gross revenue churn = (MRR lost to cancellations + MRR lost to downgrades) / MRR at the start

Net revenue churn also subtracts what those same customers added through upgrades, extra seats or usage:

net revenue churn = (MRR lost - MRR gained from existing customers) / MRR at the start

New customers don't count in any of the three, since all three are about the customers you already had.

Gross and net, with numbers

Back to the 41 customers and $1,200. Same month: the $400 customer cancels, two customers upgrade from $20 to $50 (+$60), and one moves down from $20 to $10 (-$10).

• Customer churn: 1 of 41, 2.4%.
• Gross revenue churn: $410 of $1,200, 34.2%.
• Net revenue churn: $410 minus $60, so $350 of $1,200, 29.2%.

Net revenue churn can go below zero. Take a SaaS priced by seats with $5,000 MRR. In a month it loses $150 to cancellations and $50 to downgrades, and existing customers add $400 in seats. Gross revenue churn is 4.0%. Net is (200 - 400) / 5,000, or -4.0%. That's what investors call net revenue retention above 100%, here 104%. For a product priced by seats or usage it's a very good sign, though it also lets a few fast-growing accounts hide a long list of customers who left.

A price increase moves one and not the other

Say you have $1,000 MRR, lose 3.0% of it to cancellations in a month, and raise prices 10% for everyone who stays. The remaining $970 becomes about $1,067. Gross revenue churn is still 3.0%. Net revenue churn is (30 - 97) / 1,000, or -6.7%.

Nobody stayed longer, and the same customers left as before, yet net revenue churn improved by almost ten points because of a price change. If you know about the price change, fine. Put that month next to one without a price change, though, and the comparison misleads you. Customer churn doesn't move at all.

Which one should you watch?

Both, if you can, because they answer different questions.

• Customer churn tells you whether the product works for the people who buy it. It treats a $9 customer and a $900 one the same, which is what you want when the question is "do people stay".
• Revenue churn is what churn costs you. It matters for cash and forecasts, and your biggest accounts dominate it.

If your prices are close together (one or two plans, monthly and yearly), the two numbers move together and customer churn is enough. If you have a wide range, like a self-serve plan and a few custom deals, look at them separately. In the first example, 2.4% against 33.3% is one big account leaving, and it says little about whether the product works.

When revenue churn is much higher than customer churn, your bigger customers are the ones leaving. When it's much lower, the small ones are. Either way, look at who exactly left and why (asking customers why they cancel).

Where Stripe stands

Stripe's Billing analytics shows a subscriber churn rate, which is a customer churn number with its own definition, and churned revenue as an amount (cancellations plus downgrades), without turning it into a rate. Its cohort view shows revenue retention by starting month, and that one can go above 100%. If your numbers don't line up with Stripe's, this is usually why.

Why the TrustChurn ranking goes by customer churn

A ranking has to compare SaaS of very different sizes and prices. Revenue churn handles that badly. A $3K MRR SaaS with one big customer can swing from 0% to 30% in a month, and net revenue churn rewards price increases as much as retention. Customer churn means the same thing whatever a SaaS charges, so the ranking sorts by it.

Profiles can still show revenue churn if the founder turns it on. Ours is gross and only counts cancellations, not downgrades, because Stripe doesn't keep the history of a subscription's plans. The exact formula is on the methodology page.

FAQ

Can revenue churn be negative?

Net revenue churn can. When existing customers add more MRR through upgrades, seats or usage than you lose to cancellations and downgrades, the result goes below zero. Gross revenue churn can't, because it only counts what you lost.

Is logo churn the same as customer churn?

Yes. Logo churn is another name for customer churn, from the idea that each customer is a logo on your customer list. Both count customers, not money.

Do downgrades count as churn?

They count in revenue churn, both gross and net, because the customer pays less. They don't count in customer churn, because the customer is still paying.

Which churn number matters more for a small SaaS?

Usually customer churn, because a small SaaS rarely has the volume for revenue churn to be stable. One large account can swing revenue churn by tens of points. If a few customers pay much more than the rest, keep an eye on revenue churn too, and look at those accounts one by one.

Why is my net revenue churn so much lower than my customer churn?

Most likely existing customers are paying more over time, through upgrades, extra seats, usage or a price increase. That covers some or all of the money lost to cancellations, even when you're losing as many customers as before.