SaaS churn rate: how to calculate it correctly and 2026 benchmarks
SaaS churn rate is the percentage of customers or recurring revenue you lose over a period, almost always measured monthly. It splits into two units and four numbers. Customer churn counts accounts lost. Revenue churn counts lost MRR, and it comes in two forms: gross churn (losses only) and net churn (losses minus expansion and reactivation), which can go negative when upgrades outrun cancellations. For SMB SaaS, healthy monthly churn runs 3 to 5 percent. The best teams hold below 1 to 2 percent.
Most teams report one blended logo number and call it churn. That single figure hides the diagnosis. The gap between the four churn variants is where the money and the fix live, and two of those numbers move when a customer never makes a decision at all.
The two units: customer churn vs revenue churn
There are two ways to count churn, and they answer different questions.
Customer churn (also called logo churn) counts how many accounts you lost. Every customer weighs the same, whether they pay you $50 or $5,000 a month.
Revenue churn counts how much MRR you lost. Each account is weighted by its dollar value, so one whale leaving hits harder than ten minnows.
Here are the formulas, side by side.
Customer churn rate = customers lost during the period / customers at the start of the period
Gross revenue churn rate = (churn MRR + contraction MRR) / MRR at the start of the period
The denominator rule trips up nearly everyone. Do not count customers or revenue you acquired during the period. Freeze the starting base, then measure what leaks out of it.
Work a number. You start the month with 900 customers and lose 30. Customer churn is 30 / 900 = 3.33 percent. The denominator is 900, not 900 plus the new signups you closed that month.
Now the dollars. Those same 900 customers carried $450,000 in starting MRR. During the month, accounts worth $13,500 cancelled outright (churn MRR), and another $4,500 in downgrades came off existing accounts (contraction MRR, the revenue lost when a customer drops a tier without leaving).
Gross revenue churn is ($13,500 + $4,500) / $450,000 = 4 percent.
Same month, two different stories. The two numbers diverge whenever customer sizes vary. Lose ten tiny accounts and keep your whales, and customer churn can read 5 percent while revenue churn reads 1 percent. Lose one big account and the reverse happens.
Revenue churn is the better health read for most SaaS, because it reflects the money actually at risk. The churn rate calculator runs both from your numbers.
Gross MRR churn vs net MRR churn: the leaky bucket
Gross MRR churn measures how fast the bucket leaks. Net MRR churn measures whether the bucket is still filling from your existing base, because it subtracts the expansion and reactivation revenue gross churn ignores.
Net revenue churn rate = (churn MRR + contraction MRR − expansion MRR − reactivation MRR) / starting MRR
Two new inputs. Expansion MRR is revenue from existing customers who upgraded or added seats. Reactivation MRR is revenue from previously churned customers who came back, the input that quietly breaks naive churn math when someone cancels and resubscribes mid-period.
Take the same $450,000 starting base. Churn MRR was $13,500 and contraction was $4,500, so gross churn is 4 percent. During the month, existing customers also added $22,500 in upgrades (expansion) and $2,000 in returning accounts (reactivation).
Net MRR churn is ($13,500 + $4,500 − $22,500 − $2,000) / $450,000 = −$2,500 / $450,000 = −0.56 percent.
The bucket leaked 4 percent and refilled past full. Gross churn can never go below zero, because it only counts losses. Net churn goes negative whenever expansion plus reactivation beats churn plus contraction.
That negative number has a name on the retention side. Net revenue retention (NRR) is 1 minus net MRR churn, so this base sits at 100.56 percent. Gross revenue retention (GRR) is 1 minus gross churn, here 96 percent. The net revenue retention calculator handles the mirror math if you prefer thinking in retention.
The four churn types at a glance
Churn is not one number. It is four, and each one diagnoses a different problem.
| Churn type | Formula | Can it go negative? | What it tells you |
|---|---|---|---|
| Customer (logo) | Customers lost / starting customers | No | How many accounts walked, value-blind |
| Gross MRR | (Churn + contraction) / starting MRR | No | How leaky the bucket is, in dollars |
| Net MRR | (Churn + contraction − expansion − reactivation) / starting MRR | Yes | Whether the base grows on its own |
| Involuntary | Failed-payment churn / total churn | No (a share) | How much churn is a billing problem, not a product one |
Report all four and the story writes itself. High gross churn with negative net churn means strong expansion is masking real logo loss, which is fragile. A high involuntary share means your churn is mostly a dunning problem, not a roadmap problem.
What is a good monthly churn rate in 2026?
Healthy monthly customer churn depends on who you sell to. SMB-focused SaaS runs 3 to 5 percent a month (some sources stretch to 7). Mid-market lands around 1.5 to 3 percent. Enterprise sits at 1 to 2 percent. The best companies hold monthly churn below 1 to 2 percent, roughly under 5 percent a year.
| Segment | Healthy monthly customer churn | Source basis |
|---|---|---|
| SMB-focused | 3 to 5 percent (up to 7) | Vena 2025, Vitally |
| Mid-market | 1.5 to 3 percent | Segment data |
| Enterprise | 1 to 2 percent | Segment data |
| Best in class (any) | below 1 to 2 percent | ChartMogul |
On the revenue side, ChartMogul ties the target to average revenue per account (ARPA). Keep monthly gross MRR churn below 3.5 percent if your ARPA is under $100, and below 2.5 percent if ARPA is above $100. Bigger accounts should be stickier, so the bar is higher.
Zoom out to the year and the private-SaaS data lands here: median annual gross dollar churn is about 14 percent, and median net revenue retention is about 106 percent (SaaS Capital 2025). That 106 splits hard by segment, which is why a blended number lies. The full breakdown is in the NRR benchmarks guide.
One warning on the customer-churn benchmark. SMB churn concentrates early. Roughly 43 percent of SMB customer losses happen in the first 90 days after purchase. If your blended churn looks fine, check whether new cohorts are bleeding before the old ones mask it with cohort analysis.
Negative churn: when churn becomes a growth engine
Net negative churn means your existing base grows revenue with zero new customers, because expansion and reactivation outrun cancellations and downgrades. It is the structural prize in SaaS, which is why investors pay up for it.
The example above already crossed the line. Net churn of −0.56 percent is NRR of 100.56 percent. Push expansion harder and the gap widens. About 53 percent of companies with ARPA above $500 reach net negative churn (ChartMogul). The best public SaaS runs 120 to 125 percent NRR or higher (KeyBanc 2025).
Segment matters here too. SMB-segment median NRR is about 97 percent, while enterprise median is about 118 percent (SaaS Capital 2025). A 97 percent NRR is on target for an SMB product and a crisis for an enterprise one. Compare inside your band, not against the blended median. Expansion revenue is the one input you can grow without a ceiling, so most of the work to get there lives on that side.
The 20 to 40 percent of churn you are probably ignoring
Between 20 and 40 percent of total SaaS churn is involuntary. It comes from failed payments and expired cards, not from a customer deciding to leave. This is the most actionable churn number you have, and almost nobody splits it out.
Involuntary churn is the share of your churn caused by billing failures. The formula is simple: involuntary churn / total churn.
The point is the fix, not the formula. Voluntary churn is a product and pricing problem you solve with the roadmap. Involuntary churn is a billing problem you solve with dunning and card retries. If you report one blended churn number, you cannot tell which one you have, and you burn roadmap cycles on a problem that a payment retry sequence would close.
Run the split before your next planning cycle. If a third of your churn is failed payments, the highest-ROI churn work this quarter is not in your product backlog.
The annualization trap: why monthly churn times 12 is wrong
Do not multiply a monthly churn rate by 12 to annualize it. Churn compounds. Each month’s loss applies to a smaller base, so the right formula is annual churn = 1 − (1 − monthly rate) raised to the 12th power.
The linear shortcut overstates badly, and the truth is still alarming. An 8 percent monthly churn does not annualize to 96 percent. It compounds to 1 − (0.92 to the 12th power) = about 63 percent of customers gone in a year.
The version that should scare you is the small number. A 5 percent monthly churn feels survivable. Run it: 1 − (0.95 to the 12th power) = 46 percent of your customers quietly erased over twelve months. That is the gap between “we lose a little each month” and a business that has to replace nearly half its base every year just to stand flat.
Tracking all four churn numbers in real time
The reason most teams get churn wrong is the decomposition. You have to split raw billing events into churn, contraction, expansion, and reactivation, exclude mid-period new signups from the denominator, and resolve the case where someone cancels and resubscribes inside the same month. Done by hand in a spreadsheet, that is exactly where the errors creep in.
Mowt computes all four churn variants directly from raw Stripe data in real time. It resolves mid-period reactivations and contraction automatically, so customer, gross, net, and involuntary churn come from one source of truth instead of a monthly rebuild. The churn analytics view splits voluntary from involuntary, so you know whether the next move is a roadmap or a dunning sequence.
Pull your last month and split it four ways. The blended logo number is the least useful of the four. The one that matters is the gap, and whether it is closing.
FAQ
How do you calculate churn rate?
Pick a period (usually a month) and a unit. For customer churn, divide customers lost during the period by the customers you had at the start: 30 lost out of 900 is 3.33 percent. For revenue churn, divide lost MRR (churned plus contracted accounts) by starting MRR. Never count customers acquired during the period in the denominator, and never multiply a monthly rate by 12 to annualize it, because churn compounds.
What is the difference between customer churn and revenue churn?
Customer churn counts how many accounts you lost, treating every customer equally. Revenue churn counts how much MRR you lost, weighting each account by its dollar value. They diverge when account sizes vary: losing ten tiny accounts might show 5 percent customer churn but only 1 percent revenue churn. Revenue churn is the better health signal because it reflects the money actually at risk.
What is a good monthly churn rate for SaaS?
For SMB-focused SaaS, healthy monthly customer churn is about 3 to 5 percent. Mid-market runs 1.5 to 3 percent and enterprise 1 to 2 percent. The best companies hold monthly churn below 1 to 2 percent, under roughly 5 percent annually. On revenue, ChartMogul recommends keeping monthly gross MRR churn below 3.5 percent (ARPA under $100) or below 2.5 percent (ARPA above $100).
What is the difference between gross and net MRR churn?
Gross MRR churn counts only lost revenue (churn plus contraction over starting MRR), so it can never be negative and is always higher than net churn. Net MRR churn subtracts expansion and reactivation from existing customers, showing whether upgrades offset losses. Gross churn tells you how leaky the bucket is. Net churn tells you whether it is still filling from your existing base.
Can churn rate be negative?
Net revenue churn can be negative. Customer churn and gross revenue churn cannot. Net negative churn happens when expansion plus reactivation MRR from existing customers exceeds the MRR lost to cancellations and downgrades, which is the same as net revenue retention above 100 percent. It means your base grows revenue with zero new customers. About 53 percent of companies with ARPA above $500 reach it, and the best SaaS runs 120 to 125 percent net revenue retention.
About the Author
Matt Smith
Serial entrepreneur and former big 4 consultant turned SaaS operator. Built and scaled analytics and data warehouses platforms at multiple enterprise Stripe companies before founding Mowt. Passionate about making complex metrics accessible to every founder.