How to Calculate SaaS Churn Rate
Logo churn, gross revenue churn and net churn with formulas, the denominator choice that changes the answer, and how to annualise a monthly rate correctly.
On this page 8 sections
- The four metrics, with formulas
- The denominator problem, shown on one set of numbers
- Annualising a monthly rate without inflating it
- Downgrades, pauses, non-renewals and reactivations
- Benchmark ranges, and why most of them do not apply to you
- Building the calculation so it does not drift
- How churn connects to everything else you report
- What to do next
- Frequently asked questions
The short answer
SaaS churn is four separate metrics. Logo churn counts customers lost divided by customers at period start. Gross revenue churn counts lost recurring revenue including downgrades, divided by starting recurring revenue. Net revenue churn subtracts expansion from that loss and can be negative. Customer churn and logo churn are often used interchangeably. The denominator changes the answer substantially, so publish which base you used, and never annualise a monthly rate by multiplying by twelve.
Key points before you start
Two companies report 4 percent churn. One means customers, monthly, on a start-of-period base. The other means revenue, annually, on a cohort base, net of expansion. These numbers have almost nothing to do with each other, and they end up in the same benchmark table. Churn is four metrics and a denominator argument, and most reported rates are not comparable to anything.
The four metrics, with formulas
Get these straight before touching a benchmark.
Logo churn = customers lost during the period / customers at the start of the period.
Customer churn is usually the same thing. Some teams use it to mean churn weighted by account, some by billing entity. Pick one definition and write it down, because the two terms get swapped in board decks constantly.
Gross revenue churn = (recurring revenue lost from cancellations + recurring revenue lost from downgrades) / recurring revenue at the start of the period.
Net revenue churn = (that same loss, minus expansion revenue from existing customers) / recurring revenue at the start of the period. This one can be negative, which means your existing base grew despite losses. Negative net revenue churn is the same fact as net revenue retention above 100 percent, expressed with the opposite sign.
The one that hides the most
Logo churn ignores downgrades entirely. A company where every customer cuts their seat count in half while nobody cancels will report zero logo churn and roughly 50 percent gross revenue churn. If you report only logo churn, you can lose most of your revenue without the metric moving at all.
The denominator problem, shown on one set of numbers
Here is a single month. Start with 500 customers and 250,000 dollars MRR. During the month you add 60 new customers and 28,000 dollars of new MRR. You lose 20 customers and 9,000 dollars of MRR, plus 3,000 dollars of downgrades. Expansion adds 7,000 dollars.
| Denominator choice | Gross revenue churn | What it answers |
|---|---|---|
| Start of period MRR (250,000) | 12,000 / 250,000 = 4.8% | How much of what I started with did I lose |
| Average of start and end MRR (~273,000) | 12,000 / 273,000 = 4.4% | A smoothed rate for a fast-growing month |
| Cohort base, only customers present at start | 12,000 / 250,000 = 4.8% here, but diverges once new customers churn within the month | How does an existing customer behave over time |
The three converge in a flat month and diverge sharply in a growing one, because new customers dilute the base. In a month where you grow MRR 20 percent, the average-base method can report a churn rate roughly a fifth lower than the start-of-period method on identical data. Neither is wrong. Quoting one without saying which is.
4.8% vs 4.4%
Gross revenue churn on identical data, start of period base versus average base, in a single growing month
saas-marketing.net model, method shown on the page
My position: use cohort-based revenue churn as the operating metric, because it answers the question you actually care about, which is whether customers acquired under a given set of conditions stay. Report start-of-period gross and net revenue churn to the board, because it is the convention investors expect. Publish the denominator alongside both. That is three extra words in a slide and it prevents a year of confused comparisons.
Editable CSV worksheet
SaaS benchmark evaluation worksheet
Record the source, date, cohort and metric definition before comparing your numbers with a benchmark.
Annualising a monthly rate without inflating it
Multiplying monthly churn by twelve is wrong and it is wrong in a consistent direction: it overstates. Each month’s churn applies to a base already reduced by the previous months.
The correct conversion: annual churn = 1 minus (1 minus monthly churn) to the power of 12.
| Monthly churn | Multiplied by 12 (wrong) | Compounded (correct) | Overstatement |
|---|---|---|---|
| 1% | 12.0% | 11.4% | 0.6 pts |
| 2% | 24.0% | 21.5% | 2.5 pts |
| 3% | 36.0% | 30.6% | 5.4 pts |
| 5% | 60.0% | 46.0% | 14.0 pts |
| 8% | 96.0% | 63.3% | 32.7 pts |
At low rates the error is trivial. At high rates it is enormous, and high-churn self-serve businesses are exactly the ones most likely to reach for the shortcut. An 8 percent monthly churn business does not lose 96 percent of its customers in a year. It loses about 63 percent, which is still bad and is a meaningfully different number to plan against.
Going the other way, from an annual rate to a monthly one, use the twelfth root: monthly churn = 1 minus (1 minus annual churn) to the power of one twelfth. The same compounding logic that governs CAGR applies here, just pointed downward.
Where compounding breaks down
This maths assumes churn is evenly distributed across months, which is false for annual contracts. If 70 percent of your contracts renew in January, your January churn is not comparable to your April churn and no compounding formula fixes that. For annual-contract businesses, calculate churn on the renewal cohort: contracts up for renewal in the period, and what share of them renewed.
Downgrades, pauses, non-renewals and reactivations
The edge cases are where churn numbers get quietly manufactured. Decide each of these in writing, once.
Downgrades. Count the lost recurring revenue in gross revenue churn. Do not count the customer in logo churn. Both statements are true simultaneously and that is fine.
Pauses. Pick a window, commonly 90 days, after which a paused account counts as churned. Without a rule, a growing pause population sits in limbo and flatters every metric you have.
Annual non-renewals. The customer churns in the month the contract ends, not the month they told you. Recognising it early smooths the number and misstates the period.
Reactivations. A returning customer inside your pause window reduces net churn for that period. Outside the window, they are new. Never retroactively restate a prior period’s churn because someone came back.
Failed payments. Involuntary churn belongs in the number, but track it separately. A business with 3 percent churn where a third is card failures has a billing problem, not a product problem, and dunning fixes it far more cheaply than a retention programme would.
Editable CSV worksheet
Save your marketing measurement plan
Keep a worksheet for your inputs, assumptions and next actions. You can also print the calculation directly from your browser.
Benchmark ranges, and why most of them do not apply to you
Published SaaS churn benchmarks are mostly unusable because they do not state segment, contract length or denominator. Here is a rough orientation, and treat it as orientation rather than a target.
| Segment | Typical contract | Commonly cited range | Metric usually meant |
|---|---|---|---|
| Self-serve SMB | Monthly | 3% to 6% monthly logo churn | Logo, start of period |
| Mid-market, sales-assisted | Annual | 10% to 18% annual gross revenue churn | Gross revenue, annual |
| Enterprise | Multi-year | Under 10% annual gross revenue churn | Gross revenue, annual |
| Product-led with strong expansion | Mixed | Net revenue churn near zero or negative | Net revenue, annual |
Stop comparing your monthly SMB churn to an enterprise annual figure you read in a blog post. That comparison has made more founders panic unnecessarily than almost any other number in SaaS. If you need something better calibrated, the segmented view in NRR and churn benchmarks is more useful than a single headline figure.
Building the calculation so it does not drift
Setting up churn reporting once, properly
- Write the definitions
Four metrics, one denominator per metric, the pause window, and the downgrade rule. One page. Circulate it to finance and sales.
- Pick the period that matches your contracts
Monthly for monthly billing, renewal cohort for annual contracts. Reporting monthly churn on annual contracts produces noise, not information.
- Separate voluntary from involuntary
Card failures get their own line. Mixing them hides a cheap fix behind an expensive problem.
- Build the cohort view
Customers grouped by signup month, retention tracked across subsequent months. This is the only view that shows whether churn is improving or just being masked by growth.
- Reconcile with finance monthly
Marketing MRR and finance recognised revenue will not match. Agree the bridge once rather than arguing about it every board meeting.
- Publish the denominator on every chart
In the axis label or the footnote. Non-negotiable if you want anyone outside the team to use the number correctly.
Run the numbers with the NRR and churn calculator if you want to see how expansion and contraction interact before you commit to a reporting format.
How churn connects to everything else you report
Churn drives lifetime value directly, which is why an overstated churn rate makes your unit economics look worse than they are and an understated one makes payback look better. If you use the naive twelve-times annualisation and then feed that into lifetime value, you will materially understate LTV and possibly cut acquisition spend that was working.
It also interacts with payback. A business with a 14 month CAC payback and 30 percent annual churn is in trouble in a way that a business with the same payback and 10 percent churn is not, because the second one is still collecting when the first one has lost the account. And when you look at efficiency measures such as revenue per employee, a churn number calculated on a shifting denominator will make quarter-over-quarter comparisons meaningless.
What to do next
Pull last month’s numbers and calculate gross revenue churn three ways: start of period, average base, and cohort base. If the three answers differ by more than a few tenths of a point, you now know why your churn number never matches finance’s. Pick the operating metric, write the definitions page, and put the denominator on every chart from here on. The rest of your metrics stack gets more trustworthy the moment this one stops being ambiguous.
Editable CSV worksheet
SaaS Metrics and Analytics planning worksheet
A practical metrics planning worksheet: decisions, owners, evidence and next actions.
Frequently asked questions
What is the formula for SaaS churn rate?
Logo churn equals customers lost in the period divided by customers at the start of the period. Gross revenue churn equals recurring revenue lost from cancellations and downgrades divided by recurring revenue at period start. Net revenue churn equals that same loss minus expansion revenue from existing customers, over the same starting base. All three need the denominator stated.
How do you convert monthly churn to annual churn?
Use one minus the retention rate compounded: annual churn equals 1 minus (1 minus monthly churn) raised to the twelfth power. At 3 percent monthly, that gives roughly 30.6 percent annually, not 36 percent. Multiplying by twelve overstates the loss because the customer base shrinks each month, so each month's churn applies to a smaller number.
What is a good churn rate for SaaS?
It depends entirely on segment. Self-serve SMB products commonly run 3 to 6 percent monthly logo churn and are not necessarily unhealthy. Mid-market annual contracts typically target under 15 percent annual gross revenue churn. Enterprise targets are lower still. Comparing a monthly SMB figure to an annual enterprise figure is the most common benchmarking error in the category.
What is the difference between logo churn and revenue churn?
Logo churn counts customers, revenue churn counts money. They diverge whenever your churned customers are not average-sized. Losing ten small accounts and keeping two large ones can show high logo churn and low revenue churn, or the reverse. Track both. A company with acceptable revenue churn and terrible logo churn has a retention problem it has not felt yet.
Should downgrades count as churn?
Yes, in gross revenue churn. A customer who drops from 40 seats to 10 has not churned as a logo but you have lost 75 percent of that revenue. Excluding downgrades makes gross revenue churn look artificially close to logo churn and hides contraction, which is often the earliest visible signal of a product or pricing problem.
How do you handle pauses and reactivations in churn calculations?
Define a fixed window. A common convention is that an account paused or cancelled for more than 90 days counts as churned, and a return after that counts as new rather than reactivated. Whatever you choose, apply it consistently and state it. Teams that reclassify reactivations retroactively can make almost any churn number they want.
What denominator should you use for churn rate?
Cohort base for the operating metric, start of period for board reporting, and state which one you used. Average of start and end works for high-growth months but muddies comparison across periods. The choice matters: on the same data, start of period, average base and cohort base commonly produce rates that differ by a third or more.
The saas-marketing.net editorial team Research and editorial
We research, write and maintain every page on this site. The library explains marketing decisions through practical frameworks, explicit assumptions and references. Corrections can be requested through the contact page.
Published September 11, 2026. Last updated .