Churn Rate Calculator

What this churn calculator gives you

Most churn calculators stop at one number. This one gives you the three a RevOps or finance lead reports:

• Logo churn: the share of customers who left, per period, per month and per year.
• Revenue churn: gross (losses only) and net (after expansion). When net goes below zero you have negative churn, and the calculator shows it.
• Annual GRR and NRR: the period rates compounded to a year, so a monthly or quarterly number reads like a board metric.

It also shows the average customer lifetime the logo churn implies, and what the multiply-by-12 shortcut would have told you. Under the cards, one line reads the result back and tells you whether the customers you lost were bigger or smaller than your average account.

Churn rate formulas

Every number in the calculator comes from one of these. The Excel column assumes customers at start in B2, customers lost in B3, revenue at start in C2, revenue lost from customers who left in C3, downgrades in C4, expansion in C5, and months in the period in D1. The examples use the calculator's default numbers, which are illustrative.

Revenue can be MRR or ARR. The rates are ratios, so the unit cancels out, as long as every revenue field uses the same one. For what the metric means, see churn rate. For the definitions behind each line, see the churn rate formula entry.

How to calculate churn rate, step by step

1. Pick the period and freeze the base. Count customers and recurring revenue on day one. Do not add customers you sign during the period. The churn rate denominator entry explains why.
2. Count the customers you lost from that base. A customer is lost when the contract ends or is canceled, not when usage drops.
3. Divide lost by start. That is logo churn for the period.
4. Split the revenue you lost. Revenue from customers who left goes in one bucket. Downgrades from customers who stayed go in another. Together they are gross revenue churn.
5. Subtract expansion for net revenue churn. Use only expansion from customers in the starting base. New customer revenue never goes in churn or NRR.
6. Convert with compounding. Turn a quarterly or yearly rate into a monthly one, or a monthly rate into a yearly one, with the exponent formulas above. Never multiply or divide by the number of months.

Worked example: one quarter (illustrative)

A B2B software company starts the quarter with 200 customers and $4,000,000 in ARR. During the quarter:

• 9 of those 200 customers leave. They were paying $140,000 in ARR.
• Customers who stayed downgrade by $60,000.
• Customers who stayed expand by $260,000.
• Sales also signs 15 new customers. They stay out of every number below.

Here is the math, checked twice:

Read it the way a CFO would. Net revenue churn is negative, so the base grew. But gross revenue churn of 5% a quarter compounds to a GRR of 81.5%, well under the 88% Benchmarkit reports. Expansion is covering a loss problem. The lost accounts were a little smaller than average, so this is a broad retention issue, not one big logo.

Logo churn vs revenue churn: what the gap tells you

Logo churn counts every customer the same. Revenue churn weights them by what they pay. When the two disagree, account size is the reason.

The calculator compares the average revenue of a lost customer with your average customer. Below 1x, smaller accounts are leaving, so logo churn runs higher than revenue churn. Above 1x, bigger accounts are leaving, and revenue churn is the number to worry about. For the full method, see how to calculate logo churn rate and how to calculate revenue churn rate.

Monthly to annual churn: why x12 is wrong

Each month can only lose the customers who are still there. Multiplying by 12 applies the rate to the full starting base twelve times, so it always overstates annual churn. The gap grows fast as churn rises:

The same rule works backward. A 15% annual churn assumption is 1.35% a month, not 1.25%. The monthly to annual conversion guide covers segments and annual contracts. If most of your customers sign annual contracts, monthly churn mostly reflects when renewal dates fall. See churn rate for annual contracts.

Customer lifetime from churn rate

Average customer lifetime is 1 divided by monthly logo churn. At 1.2% a month, the average customer stays about 83 months, or about 7 years. Half of today's customers are gone in about 57 months.

Treat it as a rough guide. The formula assumes churn stays flat for a customer's whole life, and it rarely does. Churn in the first year often looks different from later years. Use cohorts when you can. The method is in how to calculate customer lifetime from churn rate, and lifetime feeds the LTV formula.

Churn and retention benchmarks

The benchmarks we could confirm at the source measure revenue retention, not logo churn. So compare your gross revenue churn and NRR with these. The third column turns each one into churn, using the compounding rule.

We do not publish a logo churn benchmark. We could not confirm one at a primary source. For how churn shifts by ACV band, read what is a good churn rate and SaaS churn rate benchmarks by segment.

Common churn rate mistakes

• Counting new customers in the base. It makes churn look better in a growth quarter and worse in a flat one.
• Multiplying by 12. It overstates annual churn, and above about 8.3% a month it gives a number over 100%.
• Blending logo and revenue churn. They answer different questions. Report both, side by side.
• Leaving out downgrades. A customer who keeps one seat of ten is still a loss. Gross revenue churn has to include it.
• Calling net churn "churn". Negative churn is good news, but it can hide a gross loss problem. Show gross next to net every time.
• Converting a blended rate. A monthly self-serve base and an annual enterprise base do not churn the same way. Convert each segment, then report them apart.

What shows up before a customer churns

The calculator measures churn after it happens. Pete Furseth, ORM's COO, spends his time on the months before. Here is what ORM sees across its customers:

• Risk shows about six months out. ORM starts predicting a renewal when the renewal opportunity is created, usually 12 months ahead. Risk comes into focus about 6 months before the renewal date, and gets louder as the date gets closer.
• Six months is time to act. It gives you room to work on the customer's business case. Sending an invoice 30 days before renewal with no conversation puts the renewal at more risk.
• Support cases tell a strange story. A customer with 0 cases in a year is a churn risk. So is one with 7 or more. A customer with 3 to 5 resolved cases is usually your happiest. See how to use support tickets to predict churn.
• Contract terms matter. The biggest split is between customers on auto-renew and those who are not. Price increases of 10% a year or more put pressure on the renewal.
• Who owns renewals matters. Pete has seen GRR run lower when renewals are a finance function that sends invoices than when a dedicated customer success team owns them.

No single signal predicts churn. It is the mix of firmographics, product usage, buying history, support cases and contract terms that tells the story.

The one retention chart Pete would show every CFO

Pete's answer is a monthly retention waterfall. Each month starts with beginning ARR, which is last month's ending ARR. Then it shows the losses and gains in order:

• Losses: ARR from customers who left, ARR from products customers dropped, and ARR from product decreases.
• Gains: ARR from new customers, from new products, and from product increases.
• Then ending ARR, with GRR and NRR on the same chart.

The calculator's fields map to it. Revenue lost from customers who left is the first loss line. Downgrades are dropped products plus product decreases. Expansion is new products plus product increases. New customer ARR stays on the chart but out of churn and NRR. Because each month reconciles to the next, nothing hides in a blended number. The monthly ARR retention waterfall entry walks through it.

ORM's take: churn is a forecast input, not just a report

A churn rate tells you what you lost last period. It does not tell you which renewals are at risk this quarter, or how much of next year's ARR is already leaking.

ORM models each renewal on its own. The models are built to find the customers most likely to leave, early enough to act. That forecast then rolls into the revenue plan, so retention and new business are read on the same page.

Common questions

How do you calculate customer churn?

Divide the customers you lost during the period by the customers you had at the start, then multiply by 100. Start the month with 500 customers, lose 6, and customer churn is 1.2%. Count only losses from the starting base. A customer who signed and canceled inside the same period belongs in a new customer review, not in this rate. Run revenue churn next to it, because the two often disagree.

What is the best churn rate?

Lower is better, but compare like with like. Benchmarkit puts 2025 gross revenue retention at 88%, which is about 12% gross revenue churn a year. High Alpha says keeping 9 of 10 customers is the norm at every ARR band. The stronger goal is net revenue churn below zero, which means NRR above 100%. ChartMogul finds retention is far lower when the average account pays under $50 a month.

How do you convert monthly churn to annual churn?

Compound it: annual churn = 1 - (1 - monthly churn)^12. At 2% a month, 0.98 to the 12th power is 0.785, so annual churn is 21.5%. Multiplying by 12 gives 24%, which is too high, because each month can only lose the customers who are still there. To go the other way, monthly churn = 1 - (1 - annual churn)^(1/12).

What is the difference between gross and net revenue churn?

Gross revenue churn counts only losses: revenue from customers who left plus downgrades, divided by revenue at the start. Net revenue churn subtracts expansion from customers you kept. When expansion is bigger than the losses, net churn drops below zero. That is negative churn, and it means NRR is above 100%. Report both. A healthy net number can hide a loss problem that expansion is covering for now.

How do you calculate employee churn rate?

Use the same math with headcount. Divide the employees who left during the period by the headcount at the start, then multiply by 100. Twelve leavers from a starting headcount of 150 is 8% for the period. The annual rule is the same too: compound, do not multiply. The customer fields in this calculator work for it. Ignore the revenue results.

What is a churn model and how does it work?

A churn model scores each customer on how likely they are to leave. It uses signals such as product usage, support cases, contract terms and buying history. In ORM's experience no single signal predicts churn. It is the mix. Judge a model on the at-risk accounts it finds, not on overall accuracy. With a 90% renewal rate, a model that says everyone renews is 90% accurate and useless.

More free tools

ARR CalculatorTurn MRR or contracts into ARR, then run an ARR bridge with GRR and NRR.Open the tool →CAC CalculatorWork out customer acquisition cost and how long it takes to pay back.Open the tool →Growth Rate CalculatorMeasure growth from one period to the next, with CAGR.Open the tool →