Churn rate is the percentage of the customers you started a period with who left during that period. The formula is simple: churn rate equals customers lost divided by customers at the start, times 100. If you started a month with 1,200 customers and lost 48, your monthly churn is 48 divided by 1,200, which is 4.0%. Retention is the mirror image — 100 minus churn — so 96% of customers stayed.
That one number carries more weight than most dashboards admit, because it decides how long customers stay, how much growth you need just to stand still, and whether your revenue base is compounding or leaking. The tool recommended here — Asad Webs' free Churn Rate Calculator — works it out for customers and for revenue, in your browser, with the annualised figure compounded properly instead of just multiplied by 12.
Try it now: Churn Rate Calculator — Customer Churn, Retention and Revenue Churn. Free, no sign-up: enter customers at the start and customers lost, and churn, retention, annualised churn and implied lifetime update live.
The formula, with nothing hidden
Customer churn rate equals customers lost in the period divided by customers at the start of the period, multiplied by 100. Retention rate equals 100 minus the churn rate. Average customer lifetime equals 1 divided by the churn rate expressed as a fraction, measured in periods — so a 4% monthly churn implies an average lifetime of 1 divided by 0.04, which is 25 months.
One rule in the formula does most of the honest work: customers gained during the period do not reduce churn. They belong to growth, not to the rate. A company that starts with 1,200 customers, loses 48 and gains 200 has still churned 4%, even though it ended the month bigger. Netting new sign-ups against losses produces a net growth figure, not a churn rate, and it can disguise a serious retention problem behind strong sales — the exact trap the calculator is built to prevent.
A worked customer example, step by step
Take the calculator's own worked example, because every number in it is checkable. A software business starts the month with 1,200 customers and loses 48. Churn is 48 divided by 1,200, which is 4.0%. Retention is 100 minus 4, which is 96%. The implied average customer lifetime is 1 divided by 0.04, which is 25 months. Customers remaining are 1,200 minus 48, which is 1,152.
A second example, computed the same way, shows the formula holds at any size. Start a month with 800 customers and lose 24: churn is 24 divided by 800, which is 3.0%; retention is 97%; implied lifetime is 1 divided by 0.03, which is about 33.3 months; customers remaining are 776. The arithmetic is deliberately boring — that is what makes it trustworthy. If a churn figure you are shown cannot be reproduced from a start count and a lost count, it is not a churn rate, whatever it is called.
Always ask for the two counts behind any churn percentage: customers at the start, and customers lost. A percentage without its base can hide almost anything, including a shrinking base that flatters the rate.
Worked example rates from the calculator's own example — hover or tap each bar for details.
Why annualised churn is not the monthly rate times 12
This is the mistake that most often inflates a churn story. Annualised churn compounds the period rate, because each period's losses shrink the base the next period's rate applies to. The formula for a monthly rate is 1 minus (1 minus the monthly rate) raised to the power of 12.
With the worked example: losing 4% of 1,200 customers leaves 1,152, and the next month's 4% applies to 1,152, not to 1,200. Over twelve months that gives 1 minus 0.96 to the power of 12, which is about 38.73% — noticeably less than the naive 48% you get from 4 times 12, and the gap grows with the rate. The same compounding applies quarterly, with the power of 4 instead of 12. The calculator's annualised figure uses this compounding, and its guidance calls out the 4% example explicitly: 38.73% over a year, not 48%.
For the second example above, the same maths gives 1 minus 0.97 to the power of 12, which is about 30.6% annualised from a steady 3% monthly churn — again well below 36%. When someone quotes you an annual churn figure, check whether it was compounded or just multiplied; the multiplied version overstates the damage, and an overstated problem leads to the wrong fix.
Customer churn vs revenue churn: count accounts and money separately
Customer churn counts accounts. Revenue churn counts money, and the two can tell opposite stories. Revenue churn starts from monthly recurring revenue (MRR) at the start of the period, not from a customer count. Gross revenue churn equals MRR lost to cancellations plus MRR lost to contraction — downgrades, removed seats and discounts from customers who stayed — divided by starting MRR, times 100. Net revenue churn subtracts expansion MRR first: upgrades, added seats and add-ons from existing customers. Net revenue retention equals 100 minus net revenue churn.
The calculator's revenue example makes the distinction concrete. Starting MRR is $95,000; $3,800 is lost to cancellations and $1,200 to contraction, so gross revenue churn is $5,000 divided by $95,000, which is 5.26%. But expansion adds $7,000, so net revenue churn is ($5,000 minus $7,000) divided by $95,000, which is minus 2.11% — net negative churn, with net revenue retention of 102.11%. The existing base grew revenue on its own, even while 5.26% of its revenue was leaking. That pattern — expansion beating everything lost — is what the calculator labels net negative churn, and it is the pattern behind the fastest-compounding subscription businesses.
Watch the gap between the two rates, because it tells you who you are losing. Losing a few large accounts hurts revenue more than losing many small ones, so revenue churn running well above customer churn means your big accounts are the problem, and customer churn running above revenue churn means the losses are concentrated in small accounts. Report whichever your audience expects, but never mix a customer retention figure with a revenue churn figure in the same comparison.
| Measure | Formula | Worked example result |
|---|---|---|
| Customer churn (monthly) | Lost ÷ start × 100 | 48 ÷ 1,200 = 4.0% |
| Retention | 100 − churn | 96% |
| Average customer lifetime | 1 ÷ churn (as a fraction) | 1 ÷ 0.04 = 25 months |
| Annualised churn (monthly rate) | 1 − (1 − rate)^12 | 1 − 0.96^12 = 38.73% |
| Gross revenue churn | (Churned + contraction) ÷ starting MRR × 100 | $5,000 ÷ $95,000 = 5.26% |
| Net revenue churn | (Churned + contraction − expansion) ÷ starting MRR × 100 | −$2,000 ÷ $95,000 = −2.11% |
| Net revenue retention | 100 − net revenue churn | 102.11% |
What counts as a good churn rate
There is no universal good number, and any article giving you one without a segment is guessing. The calculator's own guidance frames it by who you sell to: consumer subscriptions often churn several percent a month, small-business software commonly runs 3% to 7% monthly, and enterprise software can sit under 1% monthly because contracts are annual. Judge your rate against your segment and your own trend, and watch revenue churn alongside it, since keeping large accounts matters more than the raw count.
The trend matters more than any single month. A steady 4% monthly churn with a rising customer count is a business that must win 48 customers a month just to replace losses on a 1,200-customer base — and more as it grows, because the replacement number scales with the base. A falling churn rate on a flat base is worth more than a flashy acquisition month, because every point of churn saved compounds into lifetime: cutting monthly churn from 4% to 3% lifts implied lifetime from 25 months to about 33 months, which is eight extra months of revenue per customer without winning a single new one.
That lifetime link is also why churn belongs next to acquisition cost. If you know what it costs to win a customer — see How to Calculate Customer Acquisition Cost (CAC) — and how long customers stay, you can judge whether the acquisition spend pays back. A business with 4% monthly churn and a 25-month lifetime needs its payback well inside that window; the break-even guide shows the same payback thinking applied to units and revenue. Churn, acquisition cost and lifetime are one argument, not three separate metrics.
Five mistakes that make churn look better than it is
First, subtracting new customers from lost customers — the netting mistake covered above. Second, annualising by multiplying by 12 instead of compounding, which overstates the yearly figure. Third, mixing periods: comparing a monthly churn figure with a quarterly one as if they were the same unit. Fourth, mixing customer and revenue figures in one comparison, which hides whether accounts or money are leaking. Fifth, quoting churn on a base that excludes the customers most likely to leave — for example, counting only paying customers while trial cancellations quietly vanish from the denominator. The calculator avoids the fifth trap by asking for one clean base: customers at the start of the period, whatever your definition, applied consistently every period.
A zero churn rate gives an infinite implied lifetime, so the calculator shows a dash instead of a number — and the same dash appears when the starting count is zero, because a rate with no base is meaningless. If your dashboard shows a lifetime for a zero-churn month, it is inventing data.
Work out your own numbers: Churn Rate Calculator is free and browser-only. And if your own website needs metrics this honest for your customers, talk to Asad Webs about your project.
Frequently Asked Questions
What is churn rate in simple terms?
The share of customers you started a period with who left during it: customers lost divided by customers at the start, times 100. If you started with 1,200 and lost 48, churn is 4.0%.
What is a good churn rate?
It depends on who you sell to: consumer subscriptions often churn several percent a month, small-business software commonly runs 3% to 7% monthly, and enterprise software can be under 1% monthly because contracts are annual. Judge your rate against your segment and your own trend.
What is net negative churn?
When expansion revenue from existing customers — upgrades, extra seats, add-ons — exceeds everything lost to cancellations and downgrades, so net revenue churn drops below zero and net revenue retention rises above 100%.
How is churn different from retention?
They are two sides of the same period: retention is the share of starting customers still there at the end, and churn is the share that left, so the two always add to 100%.
Why is annualised churn not just the monthly rate times 12?
Because churn compounds on a shrinking base. Losing 4% of 1,200 leaves 1,152, and the next month's 4% applies to 1,152, not 1,200. Over twelve months that gives about 38.73%, noticeably less than the naive 48%.
Should new customers be subtracted from customers lost?
No. Churn is measured only against the customers you started the period with. Netting new sign-ups against losses produces a net growth figure, not a churn rate, and it can disguise a serious retention problem behind strong sales.
Sources: Asad Webs Churn Rate Calculator tool page (formulas, worked customer and revenue examples, annualised compounding, segment guidance — verified live 2026-11-03); Asad Webs CAC and break-even guides linked in the text. All example arithmetic in this article was recomputed from the tool page's own inputs.
Be the first to share a useful question or thought.