The State of Subscription Fatigue & Churn 2026

Median monthly churn for subscription businesses runs about 3.3 percent, per Recurly's 2023 study. But a single number is close to useless. Consumer subscriptions churn far faster than B2B, and monthly plans churn far faster than annual.

I've spent the last few years consulting on retention for SaaS, e-commerce, and subscription businesses. The number people quote back to me most often is a churn "benchmark" with no source under it.

So we pulled the figures that do disclose their methodology, and lined them up against what we see inside real cancel flows.

Key takeaways

What is a subscription churn rate?

Subscription churn rate is the percentage of active subscribers, or of subscription revenue, that a business loses in a set period, almost always reported monthly.

Across subscription businesses on one billing platform, the median monthly rate is about 3.3 percent (3.27% precisely). That figure comes from Recurly's 2023 study of 1,200-plus of them over the calendar year.

The figure splits two ways, and each split changes what a number means.

Customer churn counts lost accounts. Revenue churn counts lost dollars, which matters more when your accounts aren't all worth the same. It also splits by cause, into voluntary cancels and failed payments, and those two need different fixes.

We don't re-teach the arithmetic here. Our how to calculate churn rate guide has the denominator question and the worked math.

A single blended rate hides more than it shows.

Imagine two businesses both report 5 percent monthly churn. One is a healthy annual-heavy B2B book, the other a failing consumer product. The rate only means something once you cut it by segment and billing model.

How this report sources its numbers

Most churn "benchmarks" online state no sample size and no year, which is why two credible-looking numbers can disagree by threefold. We tag every figure here with its sample and its data year, because that metadata is what makes a number comparable at all.

Any churn rate only compares against another rate measured the same way. Three details carry the weight:

• Denominator. Who is counted, consumer or B2B, one platform's users or all businesses.
• Sample size. How many businesses or subscribers the figure rests on.
• Data year. When the numbers were collected.

The most-cited subscription-churn page in search results reports its rates against "2024 data" with no stated sample and no methodology block. A reader can't tell whose churn it describes.

Stated methodology still doesn't make two numbers interchangeable. A billing vendor's benchmark comes from its own merchant base, which skews toward businesses that already run billing and retention tooling.

So even a well-sourced figure needs its sample bias read before you apply it.

Reading that bias is a habit worth building. Ask who the source sells to, because a benchmark drawn from paying customers of a retention tool describes businesses already trying to retain. 

Your own numbers, if you have not yet invested there, will usually sit worse than the published median, and that gap is expected rather than alarming.

Subscription churn rate benchmarks by segment

Monthly churn ranges from roughly 0.5 percent for enterprise annual contracts to 5 percent-plus for consumer monthly plans, per Recurly's 2023 study. That spread is why the overall median describes almost no one.

The only useful figure is the one matched to your segment and billing model, split three ways below.

By business type and company size

Churn falls as customer size and contract length rise. So the same "3 percent" sits at opposite ends of health depending on who you serve. The bands below come from the same 1,200-plus-business study, measured over the 2023 calendar year:

Picture an enterprise book and an SMB book that both run 3 percent monthly churn. The enterprise book is bleeding, while the SMB book is doing fine. The band matters more than the headline median, because your segment sets what "normal" even means.

One caveat rides on all three bands. They carry the same billing-platform skew noted above. A business with no retention tooling would run higher.

By billing model (monthly vs annual)

Billing model is the single largest lever on churn, and moving customers from monthly to annual is what pulls it down.

Shifting an entire book that way can cut churn by around 80 percent at the extreme, per a ProfitWell analysis of one billing provider's customer cohort. That's the upper bound, an all-annual book against an all-monthly one.

An annual contract compresses twelve monthly cancel decisions into one. Fewer decision points, fewer chances to leave. The effect holds across price bands.

But it's a correlation, not a clean cause. Higher-intent customers may pick annual to begin with.

The gap is wide at every revenue band, per ChartMogul's 2025 SaaS billing report. Under 25 dollars in average revenue per account, annual plans kept 62 percent of customers a year against 41 percent for monthly.

In the 250-to-500-dollar band, ChartMogul's data puts annual revenue retention at 88 percent against 76 percent for monthly. Both figures come from a 2,500-plus-company dataset covering 2024.

Annual billing has a cost. It pulls cash forward and locks in customers who may sour. The retention gain comes with a refund-and-goodwill bill if the product disappoints.

Voluntary vs involuntary split

Voluntary churn is an active cancel and involuntary churn is a failed payment, so the two take different fixes. Of that 3.3 percent median, about 2.4 percent (2.41%) is voluntary and 0.9 percent (0.86%) involuntary.

Our voluntary and involuntary churn guide walks the full distinction.

Involuntary churn hides a second effect. When a card fails and the customer has to re-enter it, many stop and reconsider instead.

One study of ten subscription products found that inattention alone drives 14 to 200 percent-plus more revenue than an attentive baseline. The same study found cancellations spike in the months cards get replaced.

A forced payment update is a cancel prompt in disguise.

That's why involuntary churn is worth fixing quietly. Recover the payment before the customer notices, and you skip the reconsideration entirely.

Where subscription fatigue fits the churn picture

Consumer fatigue and the business-side churn benchmarks are two readings of the same shift. When households trim their stack of monthly charges, consumer subscriptions see voluntary cancels rise first.

I've kept a subscription running for months, just because canceling meant digging through account settings for a buried button. Most people have a few of those.

When a household finally cuts its stack, the ones with the most friction survive a round or two longer. But the decision is already made. That delayed exit is what a consumer subscription reads as rising voluntary churn.

Fatigue is a portfolio decision a consumer makes across every charge they carry.

Thousands of those decisions add up, in aggregate, to the churn rate a business reads each month. So fatigue shows up as more voluntary cancels, in exactly the consumer and monthly-plan segments the benchmarks above already isolate.

The full consumer picture, with the survey figures behind it, is in our subscription fatigue explainer.

The evidence is streaming-specific, so treat it as directional for other categories. A B2B founder should read fatigue as a reason to watch their voluntary cancels more closely. The streaming rate itself stays a consumer number.

What the benchmarks look like inside real businesses

Voluntary churn is the part you can move, and moving it one point compounds. Across five cancel-flow engagements, that was the pattern behind the benchmark ranges.

Lifetime value is roughly average revenue divided by churn. So a single point off the rate stretches how long each customer pays. A classic Harvard Business Review study put the profit from a five-point retention gain at 25 to 95 percent, by industry.

The engagements below span consumer video, fitness, and design tools plus one B2B accounting product. They range from a few hundred customers to tens of thousands. Each is an individual engagement, with its own subscriber count and window.

None is a Churn.io network average. We read them as five separate stories, each true only for the business it came from.

Consumer subscriptions: where a point or two is winnable

In two consumer engagements, a reason-routed cancel flow took roughly a point off monthly churn, and pause did most of the work. Each is one business, with its own numbers:

In both, the biggest single reason for leaving was low usage or a busy stretch of life. Both point to a pause offer, because a discount does nothing for someone who has stopped using the product.

In this engagement's Churn.io data, 62 percent of the pause segment at the fitness app accepted and 52 percent came back paying. That one branch drove more than half of all saves.

Not every reason was worth an offer. Customers leaving for a better-fit competitor were routed to an exit survey with no save attempt. Chasing them wastes the offer, and their answers are worth more.

Those named-competitor answers fed the product roadmap, which cut the switching reason over the following months.

A third consumer engagement, a design tool with about 9,200 paid users, showed a different shape. Nearly half its cancels came in the first two weeks. That is failed activation, and a discount can't fix it.

So the team split the flow by account age. New users went to an onboarding call, everyone else to the standard offers.

Pause isn't a free win, though. Paused customers who resume retain a little worse than never-paused ones. The lift is real, but softer than the headline save rate suggests.

Proving it moved: the one randomized test

One engagement ran a proper randomized test, and segmented saves earned about 4.5 times the lifetime value per save that a blanket discount did.

A B2C meditation app with roughly 35,800 subscribers split traffic 50/50 for six weeks. Half saw the old 60-percent-off-everything offer, half saw a flow routed by cancel reason.

The team pre-registered the design first. Sample sizes, the primary endpoints, and the stop rules were all set before the test ran. So the readout couldn't be massaged after the fact. That is what separates a real result from a dashboard that happens to look good.

The segmented flow raised acceptance about 1.5 times, and the customers it saved stayed. Lifetime value per save moved from around 9 dollars to around 42 dollars, at point below 0.01.

The blanket discount kept buying back price-sensitive customers who left again at expiry. The segmented flow saved people whose reason was fixable.

This is the only case here with a controlled holdout. So it's the one that supports a genuine "this caused that" claim. The others are honest before-and-after readings, because none of them held out a control group.

The counter-example: when a cancel flow cannot help

In one B2B engagement, the cancel flow barely moved churn, because most of the churn was structural.

Per this engagement's Churn.io data, a vertical accounting tool with about 620 customers ran monthly churn of 1.8 percent, already low. The flow pulled it to 1.68 percent against a small cancel sample. In that same Churn.io data, roughly 60 percent of cancels were companies being acquired, going out of business, or forced onto a parent company's tool.

No offer saves an account that no longer exists.

The flow's real return turned out to be upstream. The exit reasons told the sales team which incoming deals carried consolidation risk. They started scoring the pipeline on it and deprioritized prospects sitting inside parent companies with a consolidation pattern.

That value never shows up in the save numbers, and it took a few months of exit data to see. Direct saves alone did not pay for the flow here. The exit intelligence did, but only because sales and RevOps acted on what it surfaced.

The honest read is simple. A cancel flow pays for itself where churn is behavioral and struggles where it's structural. If your churned customers are satisfied and still leaving, no save offer will hold them.

Once you know which part of your churn is behavioral, a cancel flow is where you test a save offer against that gap.

How to use these benchmarks without fooling yourself

A benchmark tells you whether to look, not what to do. Comparing your blended rate to a blended median is the most common way founders reach a wrong conclusion. The comparison hides the segment and billing mix that explain the gap.

The honest use has three steps:

1. Run your own churn rate first, so you have a real number to compare.
2. Match the benchmark to your own cut, so like compares to like.
3. Read the sample bias of the source.
4. Treat the gap as a question about which single change to make.

Say your monthly plans churn at 6 percent and your annual cohort at 1 percent. That gap tells you billing mix is the problem, so that's where to spend. Our operator playbook covers what to do next.

Even a correctly matched benchmark is a lagging read on a market that's still moving. Acquisition and trial-conversion rates have fallen year over year, so this year's median will likely be lower next year.

For the strategic case behind treating retention as the priority, start with our reduce churn playbook.

FAQ

What is a good churn rate for a subscription business?

Under about 1 percent monthly is strong for annual-heavy B2B, while a consumer monthly product near 5 percent can still be healthy. The "good" number is whatever sits at or below the median for your own segment and billing model.

Should I measure churn by customer or by revenue?

Report both, because customer churn counts lost accounts and revenue churn counts lost dollars. If you can only track one, use revenue churn, since it weights the accounts that pay the bills.

How often should I refresh my churn benchmark?

Re-check your comparison figures once a year, because the market keeps shifting and an old benchmark drifts out of date. Read your own churn rate monthly against that annual benchmark.

Does a high churn rate always mean a retention problem?

No, a high rate can trace to a high-mortality segment or your billing mix. Cut it by segment and reason before you blame retention.