Customer churn rate equals customers lost divided by customers at the start of the period, times 100. Gross MRR churn equals lost and downgraded revenue divided by starting MRR, times 100. Net MRR churn subtracts expansion and reactivation from that same numerator, and it can go negative. Use customer churn to track how many logos you’re losing. Use revenue churn, gross and net, to see the dollar impact, especially when deal sizes vary widely across your customer base.
TL;DR:
- Customer churn should be calculated using the initial customer count at the period’s start, not an average, to ensure comparability over time.
- Revenue churn distinguishes between gross loss from cancellations and downgrades and net loss after factoring in upsells and customer reactivations.
- Multiplying monthly churn rates by 12 overstates annual churn; use the compounding formula to accurately estimate yearly revenue loss.
- Always exclude customers who join and churn within the same period to avoid inflating churn metrics artificially.
- Segment churn results by cohort, plan, and channel to identify specific areas of retention issues rather than relying solely on aggregate percentages.
Table of Contents
- Churn Rate Formula Definitions You Can Copy Directly
- How to Calculate Churn Step by Step
- Calculating Gross and Net Revenue Churn From MRR
- Why You Can’t Just Multiply Monthly Churn by 12
- Common Denominator and Reporting Mistakes
- What Counts as a Good Churn Rate
- Worked Examples You Can Replicate
- What the Research Says About Getting This Right
- Why Founders Get the Churn Number Right and the Story Wrong
- Sources
- FAQ
Churn Rate Formula Definitions You Can Copy Directly
Customer churn, sometimes called logo churn, measures how many accounts you lost relative to how many you started with. The formula: customers lost during the period ÷ customers at the start of the period × 100. The denominator is the cohort that existed before the period began, not an average, not an end-of-period count. That distinction determines whether your number is even comparable month over month.
Revenue churn works on dollars instead of accounts, and it splits into two versions that answer different questions:
- Gross MRR churn = (churned MRR + contraction MRR) ÷ MRR at the start of the period × 100. This counts everything you lost, full cancellations plus downgrades, and ignores any gains.
- Net MRR churn = (churned MRR + contraction MRR − expansion MRR − reactivation MRR) ÷ MRR at the start of the period × 100. This nets out upsells and win-backs, which is why net revenue churn can drop below zero when expansion outpaces losses.
Report churn per period, monthly is standard for SaaS, and always state the period alongside the percentage. A “5% churn rate” means nothing without knowing if that’s monthly or annual.
How to Calculate Churn Step by Step
Getting a churn number that means something requires more discipline than plugging two figures into a formula. Analysts who skip steps end up with a metric that swings wildly for reasons that have nothing to do with actual customer behavior.
- Define the churn event. Decide whether churn means cancellation, non-renewal, or a defined period of inactivity, and write that definition down before you calculate anything.
- Freeze the starting cohort. Lock in the customer count or MRR figure as of the first day of the period. This is your denominator and it does not move.
- Exclude new customers from both sides. Someone who signs up and cancels inside the same month should not distort your churn percentage; Stripe’s methodology explicitly strips new customers out of the numerator and denominator before dividing.
- Count losses correctly. Include cancellations and non-renewals. For one-off purchase businesses, define an inactivity or repurchase window (say, 90 days) before you call a customer “churned,” since there’s no renewal event to trigger the count.
- Run customer churn and revenue churn separately. Don’t blend a logo-count metric with a dollar metric into one headline number.
- Segment after you compute the topline. Once the headline is set, break it out by plan, cohort, and acquisition channel to find where the damage is concentrated.
This sequence mirrors the calculation framework Yale SOM outlines for attrition and revenue analysis: define the event and period, freeze the cohort, exclude new adds, calculate logo and revenue churn independently, then segment.
Pro Tip: Keep a running log of every assumption, your inactivity window, your definition of “new,” your treatment of pauses or trials, in the same spreadsheet as the calculation. Six months from now, you’ll need it to explain why this quarter’s number doesn’t match last quarter’s methodology.
Calculating Gross and Net Revenue Churn From MRR
Revenue churn tells you something logo churn can’t: whether the customers walking out the door were paying you $50 a month or $50,000. A SaaS company can lose 10 small accounts and barely dent revenue, or lose one enterprise account and wipe out a quarter’s growth. That’s why ChartMogul’s revenue churn framework treats gross and net as separate, mandatory views rather than one blended figure.
Gross MRR churn is your retention floor. It answers “how much recurring revenue did we lose or shrink,” full stop, with no credit for anything won back. If gross churn is climbing, something in your product or pricing is actively driving customers away or down.
Net MRR churn tells a different story. It answers “after accounting for upsells and win-backs, how much revenue did we actually lose.” The math:
- Churned MRR: revenue from customers who canceled entirely
- Contraction MRR: revenue lost from downgrades among customers who stayed
- Expansion MRR: additional revenue from upsells, seat additions, or usage growth among existing customers
- Reactivation MRR: revenue from customers who churned previously and came back
Subtract expansion and reactivation from the sum of churned and contraction MRR, divide by starting MRR, and multiply by 100. When expansion is large enough, this number goes negative, meaning your existing customer base is growing revenue even as some accounts leave. Pull these four line items straight from your billing system’s subscription change log rather than estimating them; most billing platforms tag upgrade, downgrade, and cancellation events automatically, which makes this calculation mechanical rather than judgment-based — just as tools like Vetros · the Data Product Builder help build data products and pipelines for precise churn metric tracking.
Why You Can’t Just Multiply Monthly Churn by 12
Multiplying monthly churn by 12 is the single most common error in churn reporting, and it consistently overstates annual loss. The correct formula compounds the retained portion of your base across all twelve months: annual churn = 1 − (1 − monthly churn)^12.
The gap between the two methods widens fast as monthly churn increases:
The compounded figure is roughly 46%, because each month’s churn applies to a shrinking base, not the original one. Use the compounding formula for board decks, annual forecasts, and any target you’re setting for next year. The naive multiplication makes your churn problem look worse than it is, which sounds harmless until it drives a panicked pricing decision based on a number that was never real.
Common Denominator and Reporting Mistakes
Most bad churn numbers trace back to the denominator, not the math. Using the beginning-of-period active cohort as your base is what Wharton’s attrition research recommends for subscription businesses, because it anchors the metric to customers who were actually at risk of churning, not an average or an end-of-period count that already includes new signups.
A few specific traps show up constantly in churn reporting:
- Customers who join and churn in the same period inflate churn if you count them in the denominator; exclude new adds from both numerator and denominator, per Stripe’s calculation method.
- One-off purchase businesses have no natural renewal event, so you need an explicit inactivity window (60, 90, 120 days) before calling someone churned, rather than borrowing a subscription-style definition that doesn’t fit.
- Blending logo churn and revenue churn into a single headline hides which one is actually driving the trend. A 2% logo churn rate next to a 6% revenue churn rate tells you your biggest accounts are the ones leaving.
- Averaging the customer count across the period instead of freezing it at period-start understates churn by diluting the base with people who weren’t there when the period began.
Pro Tip: If your churn number jumped or dropped sharply between reporting periods, check the denominator before you check the customers. A methodology change, switching from average to start-of-period counts, for example, produces swings that look like customer behavior but are actually accounting artifacts.
What Counts as a Good Churn Rate
There’s no single acceptable churn rate. It depends heavily on average revenue per account (ARPA) and company scale, and treating a benchmark as universal is how founders talk themselves into bad decisions.
- Median customer churn tends to stabilize around 3 to 4% monthly as SaaS companies scale, with best-in-class performers under 2%.
- Higher ARPA correlates with lower churn. Enterprise accounts paying five figures a month get more implementation support and switching costs than a $19/month self-serve tool, so they leave less often and expand more.
- Negative net MRR churn is the target for top-performing SaaS companies, and it’s usually achieved through product-driven expansion loops, usage-based pricing, seat growth, upsell paths, not discounting.
A logo churn rate of 4% monthly next to a net revenue churn rate of negative 2% tells a very different story than the same 4% next to positive 8% revenue churn. The first company is losing small accounts while its remaining base grows. The second is losing accounts that matter.
Once you have a headline number, segment it by cohort, plan, and acquisition channel rather than reacting to the aggregate. A blended 3% churn rate can hide a 12% churn problem in one plan tier and a 0.5% rate everywhere else. The average tells you almost nothing about where to act.
Worked Examples You Can Replicate
Run these three calculations in a spreadsheet with your own numbers to confirm you’re computing churn the same way each period.
1. Customer churn. You start the month with 500 customers. During the month, 20 cancel and 15 new customers sign up. New customers are excluded from both sides of the equation. Churn = 20 ÷ 500 × 100 = 4% monthly customer churn.
2. Gross and net MRR churn. Starting MRR is $100,000. During the month: $4,000 in churned MRR (full cancellations), $1,000 in contraction MRR (downgrades), $3,000 in expansion MRR (upsells), and $500 in reactivation MRR (win-backs).
Report both; a board member seeing only the 1.5% figure has no idea $5,000 in MRR walked out the door.
**3.
Label every input, starting count, starting MRR, exclusion rules, in the same file as your formulas. Reproducibility is what separates a churn report someone trusts from one they have to recheck line by line.
What the Research Says About Getting This Right
Academic and industry sources converge on the same core discipline: anchor churn to a fixed starting cohort, separate customer counts from revenue, and never trust a single blended percentage.
Churn should be measured against customers active at the end of the prior period, not an average or a rolling count. Segmenting by cohort, plan, and channel turns a single misleading percentage into an actionable diagnosis.
That framing, drawn from Wharton’s retention modeling work and Yale’s attrition research, is the backbone of every formula in this guide.
Why Founders Get the Churn Number Right and the Story Wrong
The formulas in this guide aren’t the hard part. Any analyst can divide lost customers by starting customers. The hard part is what happens after the number exists, and that’s where most founders undersell themselves in front of investors.
They usually do.
The conventional advice, “know your churn rate”, isn’t wrong, it’s incomplete. Knowing the number matters less than knowing how to defend it under pressure: why you chose that denominator, why your net churn diverges from gross, what you’re doing about the segment that’s bleeding. That’s a rehearsal problem as much as a math problem, and it’s exactly the kind of adversarial questioning that separates founders who survive a partner meeting from those who get picked apart on slide four. If you want to see what that kind of scrutiny actually looks like applied to a real revenue model, the Nexus take-rate teardown walks through a forensic read of a board memo the same way a skeptical investor would.
Get the formula right first. Then get ready to explain it to someone whose job is to find the crack in your story.
— D
Sources
- Yale SOM — On the Nature of Customer Attrition and Revenue Analysis (2025)
- ChartMogul — Revenue churn (net and gross revenue churn rate)
- Stripe — How to calculate monthly churn
FAQ
What Is a Standard Churn Rate?
There’s no universal standard, it depends on ARPA and company stage. Median SaaS customer churn stabilizes around 3 to 4% monthly at scale, with higher-ARPA companies typically running lower, and best-in-class performers reaching negative net MRR churn.
What Is the Formula for Churn Rate in SaaS?
Customer churn = (customers lost ÷ customers at the start of the period) × 100. For revenue, gross MRR churn = (churned MRR + contraction MRR) ÷ starting MRR × 100, and net MRR churn subtracts expansion and reactivation from that numerator, which lets it go negative.
How Do You Convert Monthly Churn to Annual Churn?
Use the compounding formula: annual churn = 1 − (1 − monthly churn)^12. Multiplying monthly churn by 12 overstates the annual figure; at 5% monthly churn, the real annual loss is roughly 46%, not 60%.
Should I Track Customer Churn or Revenue Churn?
Track both, but for different purposes. Customer churn tells you how many accounts you’re losing; gross and net revenue churn tell you the dollar impact, which matters more when your customer base has widely varying deal sizes.
What’s the Difference Between Gross and Net MRR Churn?
Gross MRR churn counts every dollar lost to cancellations and downgrades with no offsets. Net MRR churn subtracts expansion and reactivation revenue from that same total, and it can turn negative when upsells outweigh losses.
Leave a Reply