Average SaaS Churn Rate: What the Benchmarks Hide

Ask five founders for the average SaaS churn rate and you can get five confident answers: 1%, 3%, 5%, 10%, or “it depends.” The last answer is the only honest one, but it is not useful until you define what it depends on.

A $29 monthly product losing 4% of its customers each month and a $30,000 annual contract business losing 4% of revenue each year do not have similar retention. They barely have the same metric. Yet both numbers get dropped into board decks under a line called churn.

That is how benchmarks become excuses. A founder finds an industry average near the current number, declares retention “normal,” and moves the meeting along. Six months later, Customer Acquisition Cost (CAC) is rising because the acquisition team must replace a third of the customer base before it can create net growth.

I use a stricter approach at Momentum Nexus. Before comparing a churn number, we normalize five variables through the SCOPE Framework: Same metric, Convert cadence, Organize peers, Pinpoint segments, and Establish the economic ceiling. Only then does a benchmark earn the right to influence a decision.

The average SaaS churn rate is not one number

ChartMogul’s analysis of more than 2,500 SaaS businesses shows why a universal average fails. Median monthly customer churn was 6.5% for companies below $300,000 in Annual Recurring Revenue (ARR), 3.7% for companies between $1 million and $3 million, and 3.1% for companies above $8 million.

Those figures are not targets. They describe populations at different stages of product maturity, customer fit, pricing, and operational discipline. A company below $300,000 ARR is still discovering who should buy. A company above $8 million has survived enough renewals to remove much of the obvious misfit.

Price changes the comparison again. In the same ChartMogul dataset, median monthly customer churn was 6.1% below $25 Average Revenue Per Account (ARPA), 3.1% from $100 to $250, and 1.8% above $1,000. Higher paying customers usually buy with more intent, receive more support, sign longer contracts, and face higher switching costs.

The raw benchmark tells you where a mixed population landed. I would never use it alone to call a retention result healthy.

Context	Median monthly customer churn	What the number may reflect
Below $300K ARR	6.5%	Early ICP discovery, weak activation, small base volatility
$1M to $3M ARR	3.7%	More stable fit, but retention systems still developing
Above $8M ARR	3.1%	Mature cohorts, stronger process, more customer diversity
Below $25 ARPA	6.1%	Low commitment, easy cancellation, limited expansion
Above $1,000 ARPA	1.8%	Higher intent, deeper workflow adoption, longer contracts

There is another trap. One benchmark source may publish monthly customer churn while another publishes annual Gross Revenue Retention (GRR). Both can be accurate. Putting them in the same comparison without conversion is nonsense.

The 2026 SaaS Capital survey of more than 1,000 private B2B SaaS companies found median GRR of 91% and median Net Revenue Retention (NRR) of 103% among bootstrapped companies between $3 million and $20 million ARR. That implies 9% annual gross revenue loss before expansion, not 9% monthly customer loss. The time period and denominator change the meaning completely.

The SCOPE framework for SaaS churn benchmarks

SCOPE is the checklist I use before I accept any retention comparison. The first pass takes about 30 minutes. I have seen teams spend an entire quarter arguing over a number they could not define, so those 30 minutes are cheap.

Step	Question	Output
Same metric	Are we comparing customers, gross revenue, or net revenue?	One defined numerator and denominator
Convert cadence	Are all figures monthly, quarterly, or annual?	One compounded time basis
Organize peers	Do the peers match our ARPA, ARR, contract, market, and model?	A narrow comparison group
Pinpoint segments	Which cohorts or customer groups create the result?	Segment level diagnosis
Establish ceiling	What retention does our unit economics require?	An internal target, not an external excuse

1. Same metric

Start by banning the word “churn” from the meeting unless the speaker adds a qualifier.

Customer churn counts logos. Gross revenue churn counts recurring revenue lost to cancellations and contractions. Net revenue churn gives credit for expansion and reactivation. Each answers a different operating question.

Metric	Basic calculation	Decision it supports
Customer churn rate	Customers lost divided by customers at period start	ICP fit, onboarding, logo retention
Gross revenue churn	Churned MRR plus contraction MRR, divided by starting MRR	Base revenue durability
GRR	Starting MRR minus churn and contraction, divided by starting MRR	Revenue retained before expansion
NRR	Starting MRR plus expansion minus churn and contraction, divided by starting MRR	Whether the installed base compounds

Suppose you start with 100 customers and $100,000 MRR. Five customers leave. If those five paid $500 each, customer churn is 5% while gross revenue churn is 2.5%. If one large account expands by $4,000, net revenue churn becomes negative even though five logos still left.

That business has a logo retention problem and a healthy expansion engine. Reporting only NRR hides the first issue. Reporting only customer churn hides the second.

I want at least three numbers on a founder dashboard: customer retention, GRR, and NRR. The gap between GRR and NRR is expansion contribution. The gap between customer retention and GRR tells you whether churn concentrates in large or small accounts.

This is also why I would not replace cohort retention analysis with a benchmark table. Benchmarks compare you with outsiders. Cohorts show whether your own business is getting better or worse.

2. Convert cadence

A monthly churn rate multiplied by 12 is a shortcut, not an annual churn calculation. Churn compounds because each month starts with fewer customers than the month before.

The correct conversion is:

Annual churn = 1 minus (1 minus monthly churn) to the power of 12

The difference becomes painful fast.

Monthly customer churn	Annual retention	Annual customer churn
1%	88.6%	11.4%
2%	78.5%	21.5%
3%	69.4%	30.6%
4%	61.3%	38.7%
5%	54.0%	46.0%
6.5%	44.6%	55.4%

Three percent monthly churn sounds small in a weekly meeting. It means losing about 31 of every 100 customers over a year if the rate holds. At 5%, nearly half the opening base disappears.

Annual contracts create the opposite illusion. A company can show little churn for eleven months and then take a large renewal hit in one quarter. Monthly averaging smooths the spike but can hide the renewal process failure. For annual contracts, I track renewal cohorts by contract end month and show trailing twelve month GRR and NRR beside them.

Never compare a monthly self service metric with an annual enterprise metric until both are expressed on the same basis. Even then, contract structure remains part of the peer filter.

3. Organize peers

“B2B SaaS” is not a useful peer group by itself. A developer tool with usage pricing, a seat based CRM, and a fixed price compliance platform can all sell to businesses while producing very different retention mechanics.

I narrow the peer set across five variables:

ARPA or Annual Contract Value: A $49 subscription and a $40,000 contract imply different intent, service, procurement, and switching costs.
ARR stage: Early companies carry more ICP mistakes. Mature companies have more renewal history and a larger base.
Contract duration: Monthly cancellation behavior cannot be compared directly with annual renewal behavior.
Customer segment: Small business, mid market, and enterprise buyers have different budget risk and expansion paths.
Pricing model: Seat, usage, fixed subscription, and hybrid pricing create different relationships between customer value and revenue.

Pricing model matters more than many founders expect. The 2025 Benchmarkit study of more than 500 B2B SaaS companies reported median NRR of 110% for hybrid subscription plus usage models, higher than subscription or usage alone. The report also found that both GRR and NRR generally rise with Annual Contract Value.

That does not mean every company should bolt usage pricing onto its product. It means a usage component can let revenue expand when customers receive more value. A pricing model that has no expansion mechanism should not copy the NRR target of one that expands automatically.

Public companies make the distortion visible:

Company	Public retention disclosure	Useful interpretation
Snowflake	126% NRR in Q1 fiscal 2027	Consumption expansion can outweigh churn and contraction
Datadog	NRR in the low 120s in Q1 2026	Product breadth creates several expansion paths inside one account
monday.com	110% overall NDR, 116% above $50K ARR in Q1 2026	Larger customer segments can retain and expand differently from the full base

Snowflake’s 126% disclosure, Datadog’s low 120s result, and monday.com’s segment results are impressive. None tells you the companies’ customer churn rates. NRR can remain high while smaller logos leave if larger customers expand enough.

Datadog makes the mechanism unusually clear. Its Q1 2026 disclosure showed 85% of customers using at least two products, 56% using four or more, and 35% using six or more. That product breadth helps explain the expansion capacity behind NRR. Copying the number without copying the expansion architecture is fantasy.

For a company at $80,000 MRR, the right peer set might be: B2B, $500 to $2,000 monthly ARPA, annual contracts, founder led sales, seat pricing, and customers with 20 to 200 employees. That is narrow enough to guide a target.

4. Pinpoint segments

An average can improve while the business gets worse.

Imagine that your oldest cohort retains 95% annually, but the last two quarters retain 72%. The large survivor base keeps the blended number respectable. New customer quality is deteriorating underneath it.

Segment churn at minimum by:

Acquisition cohort
ARPA band
Plan or product
Acquisition channel
ICP segment
Contract duration
Voluntary versus involuntary loss

The last split matters for low ARPA products. Failed payments are an operations problem. Deliberate cancellations are usually a value, fit, or priority problem. Combining them produces a blended rate with no clear owner.

I also separate early churn from mature churn. A cancellation in the first 90 days points toward qualification, expectation setting, onboarding, or activation. A cancellation after two years points toward habit decay, competitive replacement, champion change, budget pressure, or a missing expansion path.

We covered the behavioral side in the 72 hour SaaS activation framework. If new users never complete the core action, the retention problem exists before the first invoice becomes old enough to appear in a churn report.

Segment pattern	Likely problem	First investigation
First 90 day churn is high	Poor fit or weak activation	Sales promise, setup friction, time to first value
Large account revenue churn exceeds logo churn	Concentration risk	Executive relationships, value proof, renewal process
Logo churn exceeds revenue churn	Small accounts leaving	Low ARPA economics, self service onboarding, payment recovery
One channel churns twice as fast	Acquisition quality	Channel message and ICP filters
NRR is strong while GRR falls	Expansion masks base loss	Gross churn causes and expansion concentration

This segmentation turns “our churn is 3.2%” into a decision. Perhaps annual customers are healthy at 92% GRR while monthly starter accounts lose 7% each month. The fix is not a company wide customer success initiative. It may be a tighter starter plan promise, faster activation, or a pricing change.

5. Establish the economic ceiling

External averages describe what other companies tolerate. Your unit economics decide what you can afford.

A rough lifetime estimate for a stable subscription is one divided by monthly customer churn. At 2% monthly churn, expected lifetime is about 50 months. At 5%, it is about 20 months. This shortcut ignores cohorts, gross margin, expansion, and time value, but it exposes the scale of the problem.

Suppose ARPA is $800, gross margin is 80%, and monthly gross revenue churn is 3%. The simple gross profit LTV estimate is:

$800 multiplied by 80%, divided by 3%, equals about $21,333

If fully loaded CAC is $9,000, the LTV to CAC ratio is about 2.4. A founder may find a 3% churn benchmark and feel average. The economics still fail the usual 3 to 1 threshold, before overhead and discounting.

A rate can be normal and still be unacceptable for your acquisition cost, margin, cash position, and growth plan. That conflict is common, and the economics win every time.

Set the internal target backward from four constraints:

Constraint	Question	Target implication
CAC payback	How many months until gross profit recovers CAC?	Retention must exceed the payback window by a wide margin
LTV to CAC	Does gross profit LTV clear the required return?	High CAC requires lower churn or higher expansion
GRR	How much base revenue can disappear before growth stalls?	Set a hard floor before giving credit for expansion
NRR	Can the installed base grow without new logos?	Build an expansion target suited to the pricing model

The 2026 SaaS Capital median of 91% GRR and 103% NRR is useful context for a $3 million to $20 million bootstrapped B2B company. If your model needs 95% GRR to recover a 20 month CAC payback safely, 91% is not permission. Your ceiling is set by the cash equation.

Expansion deserves its own operating system. I outlined that in the SaaS expansion revenue framework. GRR tells you whether the floor is solid. NRR tells you whether the installed base can compound on top of it.

A practical SaaS churn benchmark scorecard

Build one scorecard per customer segment. Do not bury Starter, Growth, and Enterprise in one blended row.

Field	Example entry
Segment	Mid market B2B, 50 to 250 employees
ARPA	$1,200 monthly
Contract	Annual, billed annually
Pricing	Seats plus fixed platform fee
Monthly customer churn	1.1% equivalent
Trailing twelve month GRR	91%
Trailing twelve month NRR	106%
First 90 day customer retention	94%
Best peer benchmark	Annual B2B, $500+ ARPA, similar ARR stage
Internal GRR floor	93%
Main gap	Contraction in accounts below 30 seats
Next action	Interview ten contracted accounts before pricing review

Review the scorecard monthly, but judge annual contract segments on renewal cohorts and trailing twelve month results. Use a three month moving average for monthly plans so one small base fluctuation does not trigger a false alarm.

Run the analysis in this order:

Calculate customer churn, GRR, and NRR from the same starting cohort.
Convert every external benchmark to the same cadence.
Filter peers by ARPA, ARR, segment, contract, and pricing.
Split your result into cohorts and customer segments.
Set the target from unit economics, then use peers as a reasonableness check.

If the number misses the internal target for two consecutive periods, move from benchmarking to diagnosis. Review churn reasons, product behavior, champion activity, and commercial signals. Our SaaS churn prediction signal stack covers how to identify the specific accounts likely to leave before the revenue disappears.

Three benchmark mistakes I keep seeing

Treating median as healthy

Median means half the sample performed worse. It does not mean the result supports your business model. In a weak market, the median can describe a broad problem.

Use median to locate yourself. Use the top quartile to understand feasible performance. Use internal economics to set the target.

Comparing your best segment with someone else’s whole company

Founders sometimes compare enterprise GRR with a broad SaaS average, or compare overall NRR with a public company’s large customer metric. The favorable comparison survives because the denominators stay hidden.

Put the metric definition, segment, and time period beside every benchmark in the board deck. If those three labels are missing, remove the number.

Waiting for a perfect benchmark

You will never find a dataset containing twenty companies with your exact ARPA, product, stage, pricing model, geography, sales motion, and contract. That does not make benchmarking useless.

Use the narrowest credible peer set available, state the remaining differences, then anchor the final target in your own economics and cohort trend. Direction matters more than false precision.

The number should force a decision

The average SaaS churn rate is useful only after you remove the ambiguity. Define the metric. Convert the cadence. Match the peer group. Split the segments. Check the economics.

Then ask the question a benchmark is supposed to answer: are we retaining customers well enough to support the company we are trying to build?

If the answer is no, another industry report will not fix it. The next step is a retention diagnostic that connects lost revenue to acquisition source, activation behavior, product usage, account relationships, and renewal process.

If your churn dashboard looks acceptable but growth still feels like a treadmill, book a free growth audit with Momentum Nexus. We will map the retention math, show what the blended number hides, and build a focused 90 day plan around the segment causing the leak.