SaaS churn rate benchmarks by price point and segment

A private business-to-business (B2B) software as a service (SaaS) company that loses about 9% to 10% of its recurring revenue a year, before expansion, sits at the median. About 5% a year is the mark usually called good. SaaS Capital's 2025 survey put median gross revenue retention at 90% to 91% for contracts up to $250k a year and 95% above that. Check whether a benchmark is monthly or annual before comparing.

What is a good churn rate for SaaS

The short answer: a private business-to-business (B2B) SaaS company that loses about 9% to 10% of its recurring revenue a year, before counting expansion, is at the median. About 5% a year is the mark usually called good. These figures come from different sources with different definitions, so check which one you are holding your number against.

Median gross revenue retention stays between 90% and 91% for contracts up to $250k a year and reaches 95% above that.

Median gross revenue retention by annual contract value, private B2B SaaSPer cent

Median gross revenue retention by annual contract value, private B2B SaaS

Item	Value, %
Under $12k	90%
$12k to $25k	91%
$25k to $50k	91%
$50k to $100k	90%
$100k to $250k	91%
Over $250k	95%

The most specific figure comes from SaaS Capital's 2025 survey. There, median gross revenue retention was 90% to 91% at every contract size up to $250k a year, which is about 9% to 10% annual revenue churn. Above $250k it was 95%, or about 5% churn. Retention is measured from December 2023 to December 2024, and the sample covers companies with more than $1M in annual recurring revenue (ARR).

The other figures you will see quoted are broader. Revenera, drawing on aggregated data from industry reports, puts the average churn rate for SaaS is typically 10-14% annually and treats under 5% as the benchmark for a good annual rate. In our reading, use those as rough markers, not as a peer comparison.

Benchmark	Figure	Source	Sample and period
Median gross revenue retention (GRR), contracts up to $250k	90% to 91%	SaaS Capital, 2025	Private B2B, more than $1M ARR, December 2023 to December 2024
Median GRR, contracts over $250k	95%	SaaS Capital, 2025	Same
Average annual churn	10% to 14%	Revenera, 2025	Aggregated industry reports
Good annual churn	Under 5%	Revenera, 2025	Aggregated industry reports

Get the definition right before you compare

GRR is the share of starting revenue you keep, excluding expansion. In the SaaS Capital data, churn is 100% minus gross retention. Churnfree gives a worked case: if your recurring revenue was $1,000,000 at the start of the year and you lost $100,000 of it, your annual gross revenue churn rate is 10%, which is the same as 90% gross revenue retention.

Then check the period. 5% annual churn is about 0.43% a month, 10% is about 0.87%. Mixing monthly and annual figures makes your number look far better or worse than it is.

Price point changes what counts as normal too. Churnfree notes that a 5% monthly churn rate would be a serious problem for an enterprise tool but can be normal for a low-priced tool sold to freelancers.

What to do this week

1. Pull revenue from customers who were active 12 months ago, then subtract cancellations, downgrades and failed payments. Leave upgrades out.
2. Divide what remains by the starting figure. That is your GRR.
3. If you track monthly, convert to annual before comparing.
4. Compare against the contract-size band closest to your own annual contract value (ACV), not the overall average.

Run the same check by signup cohort to see whether retention is improving.

How churn changes with price point and contract size

In our reading, contract size changes expansion more than it changes losses. The previous section showed gross retention sitting in a narrow band below the top contract tier. Net revenue retention (NRR), which adds upgrades, add-ons and price increases back in, separates the price points more clearly.

Median net revenue retention ranged from 98% in the under-$12k band to 106% above $250k.

Median net revenue retention by annual contract valuePer cent

Median net revenue retention by annual contract value

Item	Value, %
Under $12k	98%
$12k to $25k	103%
$25k to $50k	102%
$50k to $100k	104%
$100k to $250k	102%
Over $250k	106%

Segment	Source and measure	Figure
ACV under $12k	SaaS Capital 2025, median NRR	98%
ACV $12k to $25k	SaaS Capital 2025, median NRR	103%
ACV $25k to $50k	SaaS Capital 2025, median NRR	102%
ACV $50k to $100k	SaaS Capital 2025, median NRR	104%
ACV $100k to $250k	SaaS Capital 2025, median NRR	102%
ACV over $250k	SaaS Capital 2025, median NRR	106%
ARPA under $10 a month	ChartMogul 2023, share of companies with NRR over 100%	2.7%
ARPA over $500 a month	ChartMogul 2023, share of companies with NRR over 100%	41.1%

The two halves of the table measure different things. The SaaS Capital rows are medians by annual contract value (ACV), from the same survey sample and period described above. The ChartMogul rows come from data on over 2,100 SaaS businesses and give the share of companies clearing 100% NRR, grouped by average revenue per account (ARPA), a monthly figure. Do not read a ChartMogul share as a median.

In this table, the under-$12k band is the only ACV group with a median NRR below 100%. At the low end of ARPA, very few companies get existing customers to grow their spend. We think that is the main difference between cheap and expensive products: losses look similar, expansion does not.

ChartMogul also tied NRR to growth. Businesses above 100% NRR grow 43.6% per annum, while those below 60% grow at just 13.1% per annum.

A worked example

Take a cohort paying $100 in recurring revenue. This is an example, not a benchmark. At the under-$12k median NRR of 98%, that cohort pays $98 a year later. At the over-$250k median of 106%, it pays $106. The gap between $106 and $98 is $8 per $100 of starting revenue, with no new sales in either case.

What to do this week

1. Convert your ARPA to ACV by multiplying monthly ARPA by 12, then find your row.
2. Calculate NRR for the customers you had 12 months ago. Leave out anyone who joined since.
3. Compare against your own row only. If you sit below 100% at a low price point, we would test an expansion path (seats, usage tiers or add-ons) before adding acquisition spend. Track it alongside retention so you can see whether the change came from fewer losses or more expansion.

How to convert monthly churn to annual churn

Multiplying a monthly churn rate by 12 is the quick way to estimate annual churn, and it overstates the loss. It assumes every month's churn comes out of your starting customer base. In practice each month's rate applies to a smaller base, because customers who left in January cannot leave again in February. Losses compound on a shrinking base, so the real annual figure sits below the simple multiple.

Starting from 100 customers, monthly churn of 1% to 7% leaves between 88.6 and 41.9 customers after 12 months, an annual churn of 11.4% to 58.1%.

Annual customer churn implied by a constant monthly churn rate% lost in 12 months

Annual customer churn implied by a constant monthly churn rate

Item	Value, % lost in 12 months
1% a month	11.4 % lost in 12 months
2% a month	21.5 % lost in 12 months
3% a month	30.6 % lost in 12 months
5% a month	46 % lost in 12 months
7% a month	58.1 % lost in 12 months

The formula

Annual churn = 1 - (1 - monthly churn)^12

Read it in three parts. (1 - monthly churn) is the share of customers you keep each month. Raising it to the 12th power gives the share you keep after a year. Subtracting that from 1 gives the share you lost.

def annual_churn(monthly_rate):
return 1 - (1 - monthly_rate) ** 12

annual_churn(0.03) # 0.306, or 30.6%

In a spreadsheet, with the monthly rate as a decimal in cell B2: =1-(1-B2)^12.

Worked example: 3% a month

Take 100 customers as a round example, not a benchmark. At 3% monthly churn you keep 97% each month. After 12 months, 69.4 customers remain, so annual churn is 30.6%. The simple method, 3% times 12, gives 36%, which overstates the loss by 5.4 points (36 minus 30.6).

Monthly churn	Customers left after 12 months (of 100)	Annual churn
1%	88.6	11.4%
2%	78.5	21.5%
3%	69.4	30.6%
5%	54.0	46.0%
7%	41.9	58.1%

The gap grows as churn rises. At 7% a month, 7% times 12 gives 84%, against 58.1% compounded.

Why the period label matters

Benchmarks rarely make this easy. Paddle notes that some companies report revenue churn, others report customer churn, and often there's no distinction between annual or monthly churn. Comparing your monthly rate to someone's annual figure, or the reverse, produces a wrong read.

Contract length changes the math too. Paddle points out that with an annual contract there's only one opportunity for a customer to churn out, as opposed to 12 with a monthly contract. The formula fits monthly billing. For annual plans, measure renewal loss directly.

What to do this week

1. Label every churn figure in your dashboard as monthly or annual, and customer or revenue.
2. Convert monthly rates with the formula above before comparing them to an annual benchmark.
3. Report both numbers side by side so nobody multiplies by 12 in a board deck.
4. Check the assumption. The formula uses one constant rate for all 12 months. If your early months lose more customers than later ones, track a real cohort for a year and read annual retention from it.

Logo churn and revenue churn measure different losses

Customer churn, also called logo churn, counts accounts. Revenue churn counts dollars. The two can move apart, and a team that tracks only one will misread where it is losing ground.

Paddle gives the simplest version of the count: a business with 100 customers in January that keeps 90 in February has a customer churn rate of 10%. Use that as an example, not a benchmark, and extend it to revenue.

The three measures on one example

Say each of the 100 customers paid $100 a month. Starting monthly recurring revenue (MRR) was 100 times $100, or $10,000. Losing 10 of them at $100 each removes 10 times $100, or $1,000.

Measure	What it counts	Example result
Logo churn	Customers lost ÷ customers at start	10 of 100, so 10%
Gross revenue churn	MRR lost to cancellations and downgrades ÷ starting MRR	$1,000 of $10,000, so 10%
Net revenue churn	Gross MRR lost minus expansion MRR from remaining customers, ÷ starting MRR	Below 10% once upgrades come in; negative if expansion exceeds losses

In this example the first two match because every plan costs the same. Real customer bases are uneven. If the 10 who left were on your cheapest plan, revenue churn lands under logo churn. If one large account left, the reverse happens.

The gap shows up in survey data. A KBCM Technology Group survey reported a median annual revenue churn rate of 13.2% and a logo churn rate of 21% among 162 respondents, and that sample excluded companies with less than $5m in revenue. In our reading, a logo figure well above the revenue figure usually means smaller accounts are leaving faster than larger ones.

Count subscribers, not subscriptions

The denominator matters as much as the definition. Recurly defines churn rate by dividing the number of subscribers you lost during a period by the total number of subscribers you had at the start of that period. It counts people, not plan records, because subscription-level churn can be inflated when subscriptions automatically expire and roll into new subscriptions.

Here is how that inflation happens. A customer whose annual plan expires and renews as a new plan record looks like one loss and one new sale at the subscription level. At the subscriber level, nothing changed.

Two more rules from the same method are worth copying into your own definition:

• Paying customers only. Only subscribers with at least one paid subscription invoice are included, so trial users stay out of the denominator.
• Ignore ghost churn. Churn that occurs and resolves within the same month ("ghost churn") isn't included. A failed card that recovers before month end is not a lost customer.

This week, write all three formulas into your metrics doc with the counting unit stated. Then split each by cohort and plan so you can see which segment drives the gap between logo and revenue churn.

What is a good net revenue retention

Net revenue retention (NRR) starts with gross revenue retention (GRR), then adds back what existing customers spend on upgrades, extra seats and add-ons. GRR can never pass 100%. NRR can, and that gap is where expansion revenue shows up.

The benchmark to hold yourself to

For business-to-business software as a service (SaaS), Benchmarkit's 2025 benchmarks put NRR at 101%.

Use 100% as the line to clear. Below it, your base shrinks before sales adds anything.

How expansion lifts NRR above GRR

Here is a worked example, not a benchmark. Take a cohort paying $100 of ARR at the start of the year. Apply the two Benchmarkit figures above as if one company posted both.

Line	Example amount
Starting ARR from the cohort	$100
ARR left after churn and downgrades (88% GRR)	$88
ARR at year end (101% NRR)	$101
Expansion needed to close the gap	$13

The $101 ending ARR minus the $88 left after losses is $13 of expansion. With these inputs, most of the expansion only refills what churn took out. Above 100%, each extra point of expansion is growth.

Adjust for your customer count

ChartMogul found that NRR also depends on how many subscribers you have. In the first half of 2024, top-quartile companies with up to 1.5K subscribers reached 100% NRR, while companies with over 12K subscribers typically sat at 76%. If you sell high volume at low prices, compare yourself with companies of a similar subscriber count before you treat a figure under 100% as a failure.

What to do this week

1. Pull last year's starting cohort and calculate GRR and NRR on that same group.
2. Subtract GRR from NRR. The result is your expansion contribution in points.
3. If NRR sits below 100%, work on retention first. If GRR is healthy and NRR is flat, look for upsell paths in pricing and packaging.

How churn differs for SMB, enterprise, B2C and AI-native products

Who buys your product shapes retention as much as what you charge. Earlier sections showed net revenue retention (NRR) rising with contract size and with average revenue per account (ARPA). This section splits the same question by company scale and customer type.

Scale lifts the top of the range

ChartMogul's 2023 report found that top quartile NRR rose with annual recurring revenue (ARR): 94% for businesses at $1-3m, 99% in the $3-15m segment and over 105% at $15-30m. Even in that top band, only 35.1% of businesses had NRR over 100%. Benchmark against your own ARR band. A $2m company measured against a $25m peer will draw the wrong conclusion.

B2B, B2C and AI-native products

ChartMogul's 2025 retention report covered companies with at least $250k ARR, drawn from roughly 2,700 B2B SaaS, 600 B2C SaaS and 200 AI-native companies. Median NRR was 82% for B2B SaaS, with an upper quartile of 97%, and 49% for B2C. AI-native companies posted 40% gross revenue retention (GRR) and 48% NRR. Their median GRR did improve during 2025, from 27% in January to 40% in September.

As an example, not a benchmark: a cohort starting at $100 of monthly revenue keeps $82 a year later at 82% NRR and $48 at 48% NRR, a $34 gap on every $100 you sold.

In our reading, within AI-native products, price band carried most of the difference:

AI-native monthly price	GRR	NRR	Our reading
Over $250	70%	85%	Behaves like B2B SaaS
$50 to $249	45%	61%	Mixed buyers
Under $50	23%	32%	Mostly trial behavior

The report's author proposed treating serious spend as revenue from customers who have paid at least $250. A step to take this week: tag every account by cumulative spend above or below $250 and report GRR for each group separately. That keeps experimental buyers from hiding, or exaggerating, the churn of committed ones.

Who answers the private-company surveys

Survey benchmarks reflect their respondents. High Alpha's 2025 report drew 800+ respondents, 69% from the US and 17% from Europe. 37% targeted the enterprise, while 22% targeted SMB (small and midsize business), 7% micro-business and 3% consumer. In our reading, a consumer or self-serve team should give such survey medians less weight than segment-specific data like the table above.

Track failed-payment churn separately from cancellations

A customer who clicks cancel and a customer whose card declined both show up as one lost account in a single churn number. They are different problems with different owners. The first is a product or value problem. The second is a billing problem, and it can often be fixed without a single conversation.

Recurly's benchmark report separates the two: it covers voluntary and involuntary churn rates, decline rates, and recovery rates by decline reason, compiled from more than 2000 businesses. Your own reporting should make the same split.

Tag every churn event with a reason

Add one field to every churned subscription and fill it at the moment the subscription ends. Four values cover most cases.

Churn type	What triggered it	First response (our reading)
Voluntary cancel	Customer ended the plan	Exit survey, then a save offer or pause
Involuntary, declined card	Payment failed and was not recovered	Retry schedule and card update emails
Involuntary, expired card	Card on file reached its end date	Pre-expiry reminder before the renewal date
Paused	Customer stopped billing but kept the account	Reminder before the pause ends

As an example, not a benchmark: if 100 accounts churned last quarter, sort all 100 by this field before anyone writes a retention plan. The cancels go to product and customer success. The declines go to whoever owns billing. Each group gets its own monthly rate, tracked by cohort.

Offer a pause and count it on its own

Recurly reports that over 50% of consumers would consider pausing a subscription instead of canceling entirely if given the option. That finding is about consumers, so B2B (business-to-business) teams should test it before assuming it holds. Put the pause option inside the cancellation flow. Record a pause as its own state, separate from both active and churned, so a wave of pauses does not hide inside your retention number.

Watch usage, not only satisfaction scores

Customer Satisfaction (CSAT) and Net Promoter Score (NPS) surveys measure sentiment, not actual usage patterns. Revenera argues that churn risk is often more accurately predicted by low product engagement or reduced feature adoption, and notes that surveys are usually conducted periodically (quarterly or annually).

This week, pick the one action that defines activation for your product and set an alert when an account's weekly count of that action drops for two weeks running. Route the alert to a named person with a set response. Keep the survey, but treat usage as the earlier signal and the survey as context for why it moved.

How to benchmark your own churn this week

Benchmarking your churn takes one month of billing data, one formula and one peer row. The method below is ours. The example company's numbers are an example, not a benchmark.

Worked example: a $60 product

The company earns $60 average revenue per account (ARPA) a month. It loses 2.5% of its customers a month (logo churn). It also loses 1.8% of its monthly recurring revenue (MRR) a month to cancellations and downgrades.

1. Pick the matching contract-size row. $60 a month times 12 months is $720 in annual contract value (ACV). That puts the company in the under $12k band, where SaaS Capital's median GRR was 90% and median net revenue retention (NRR) was 98% (Churnfree). That sample covers companies above $1M in annual recurring revenue (ARR), with retention measured from December 2023 to December 2024 (Churnfree), so treat it as a reference point.
2. Compare GRR, then NRR. A median of 90% minus 80.4% leaves a 9.6-point GRR gap. Next, add expansion MRR back in, recompute the annual figure as NRR and set it against the 98% median. Logo churn running above revenue churn suggests that smaller accounts leave first.
3. Queue three retention experiments. Pick the three below and score them in the growth experiment backlog template with an ICE score.

Copyable template

Step	What to enter	Example company
Monthly logo churn	Customers lost ÷ customers at start	2.5%
Monthly revenue churn	MRR lost ÷ MRR at start	1.8%
Annual GRR	100% minus annual revenue churn	80.4%
ACV band	ARPA × 12	$720, under $12k
Gap to median GRR (our calculation)	Median minus your GRR	9.6 points

Three experiments to queue this week

• Shorten time to value. Find the activation step that customers who stay complete in week one, then move it earlier in onboarding.
• Offer a pause in the cancel flow. The pause interest covered earlier makes this a cheap test against outright cancellation.
• Split churn by plan and signup cohort. A spreadsheet works too. Run the experiment on the cohort with the steepest early drop.

Rerun the template monthly so you measure each experiment against your own baseline.

What this page does not cover

This page does not break churn down by industry vertical or region, and it does not cover how to design pricing, packaging or upsell paths in detail. The benchmarks come from surveys with their own samples and periods, so treat them as reference points. Nothing here is financial or legal advice.

Related on Race to Scale: What is a good activation rate? Benchmarks by product type and how to define yours; Free-to-paid conversion rate benchmarks for freemium, opt-in trials and card-required trials; A/B test sample size and duration calculator, with the formula.