How to calculate churn rate for a subscription product
Last updated: July 23, 2026
From the SoleOS answers series — written about our own product space; grounded in published definitions and documented behavior, never invented numbers.
Customer churn rate is the percentage of paying customers you lost in a period, divided by the customers you started with. Revenue churn is the same idea applied to money instead of headcount — MRR lost divided by MRR at the start of the period. They rarely match, and net revenue churn (which nets out expansion revenue) is the one number that can legally go negative — meaning upgrades from existing customers outpaced what you lost to cancellations and downgrades.
Customer churn vs. revenue churn
These answer different questions, and mixing them up is the most common churn-reporting mistake.
Customer churn rate — how many people left:
Customer churn rate = Customers lost in period / Customers at start of period × 100
If you started the month with 100 subscribers and 10 canceled, that's a 10% customer churn rate. Simple, but it treats a $19/mo Solo-plan cancellation the same as a $79/mo Portfolio-plan cancellation, which is misleading once your plans are priced differently.
Revenue churn rate — how much money left:
Gross revenue churn = MRR lost in period / MRR at start of period × 100
MRR lost includes cancellations and downgrades, but not expansion. Say those 10 canceling customers were on Solo ($19/mo), so you lost $190 in MRR. If you started the month at $2,000 MRR, that's a 9.5% gross revenue churn rate — close to the customer number in this case, but only because the canceling customers happened to be priced near your average.
Gross vs. net revenue churn
Gross revenue churn only counts what you lost. Net revenue churn also counts what existing customers added back through upgrades and expansion:
Net revenue churn = (MRR lost to cancellations + downgrades − MRR gained from upgrades + expansion) / MRR at start of period × 100
Continuing the example: if 5 of your remaining customers upgraded from Solo ($19) to Maker ($39) that same month, that's $100 of expansion MRR. Net revenue churn becomes (190 − 100) / 2,000 = 4.5%. If expansion had been $200 instead, net revenue churn would be negative — you'd have grown MRR from your existing base even while some customers left. Negative net revenue churn is the holy grail metric that VCs and operators talk about, precisely because it means your existing customers alone are growing the business, independent of new sales.
| Customer churn | Gross revenue churn | Net revenue churn | |
|---|---|---|---|
| What it measures | People lost | Dollars lost | Dollars lost minus dollars gained from existing customers |
| Can it be negative? | No | No | Yes — and that's a good thing |
| Weighted by price? | No | Yes | Yes |
| Best for | Product/retention health | Revenue risk | Overall account-base health |
Why one blended monthly number lies to you at small scale
A single "churn rate" for the whole period is the number most dashboards show, and it's the one most likely to mislead a small subscription business.
The math problem is simple: at low customer counts, one cancellation is a huge percentage swing. If you have 12 customers and one cancels for a reason that has nothing to do with your product — a card expired, their business closed, they changed roles — your churn rate jumps to 8.3% for the month. Annualized, that sounds alarming. It might just be noise.
The second distortion is subtler: blending new signups and cancellations into one period hides what's happening. Add 15 new customers and lose 15 existing ones in the same month and net customer count is flat — looks like 0% growth. But that says nothing about whether your existing customers are sticking around; it only means gross additions offset gross losses. A business with high churn and equally high acquisition looks identical, on a blended net basis, to one with low churn and low acquisition. Those are very different businesses to run.
Cohort-based measurement: the honest alternative
The fix is to stop measuring "churn this month" and start measuring "what happened to the people who joined in a specific month." Take everyone who started a paid subscription in, say, January, and track what percentage of that January cohort is still paying in February, March, month 6, month 12. Plot the same exercise for the February cohort, the March cohort, and so on.
This isolates churn from new-customer noise entirely — a cohort's size only shrinks over time in this view, so you can't accidentally cancel out losses with new signups. It also reveals the shape of your retention curve: does churn front-load in month one (an onboarding problem) or is it a slow, steady leak that continues for a year (a value or pricing problem)? A blended monthly rate can't tell you that; a cohort curve can.
Annual plans and mid-period signups complicate the math
Two practical wrinkles show up as soon as you have more than a handful of customers:
- Annual plans don't have a monthly renewal event. A naive monthly churn calculation either ignores annual subscribers until their actual renewal date or has to normalize their contribution to MRR (annual price ÷ 12) and treat non-renewal at the anniversary as the churn event. Mixing monthly and annual subscribers without this normalization understates who's actually at risk of leaving.
- Mid-period signups muddy the denominator. Someone who signs up on day 15 wasn't a "customer at the start of the period," so they shouldn't sit in that period's denominator — and if they cancel before day 30, that's a different failure mode (onboarding drop-off) than a six-month customer leaving. Lumping both together obscures which problem you're looking at.
Define your period boundaries and denominator explicitly, and don't let annual normalization or early cancellations quietly distort a number you're using to make decisions.
What's a "good" churn rate?
There's no honest single number to give you here, and any guide that hands you one industry-average benchmark is fabricating it. What counts as "good" depends on your price point (a $19/mo self-serve product structurally churns faster than a $500/mo tool with a signed contract), your customer type (consumer vs. small business vs. enterprise), your billing cadence (annual plans show lower monthly churn because the decision point is rarer), and your stage (early customers behave differently than ones who found you two years in through word of mouth).
The useful discipline is tracking your own trend over time — is gross revenue churn improving or worsening quarter over quarter, is net revenue churn heading toward or below zero, is your cohort retention curve flattening out sooner than it used to. A rate that's stable or improving against your own history tells you more than comparing yourself to a number pulled from someone else's business model. If you want the vocabulary standardized before tracking trends, SoleOS's metrics glossary defines MRR, churn, and related terms consistently, and the Founder Playbook guides go deeper on related topics like how ARR and MRR relate for a small SaaS.
Disclosure: this guide is published by SoleOS, a portfolio dashboard built for solo founders running several subscription products at once. If you run one product with a few dozen customers, a spreadsheet with a cohort tab is genuinely all you need — you don't need a tool for this yet. SoleOS starts earning its keep once you're reconciling churn across multiple apps and don't want to rebuild the same cohort math in every spreadsheet.
Once you've got the measurement right, the next question is what to actually do about it — see this guide on reducing SaaS churn for tactics once you know your real number.
Frequently asked questions
Should I use customer churn or revenue churn to report to myself?
Track both, but lead with net revenue churn if you have any upgrade or expansion activity — it best reflects account-base health since it accounts for plan-price differences and rewards expansion. Customer churn is still worth watching separately since it tells you about product stickiness independent of pricing.
How do I handle a customer who downgrades instead of canceling?
A downgrade counts as revenue churn (you lost MRR) but not customer churn (they're still a customer). Track it inside gross revenue churn's "MRR lost" bucket alongside cancellations, rather than dropping it, so the revenue number stays honest even though the customer count doesn't move.
What period should I use — monthly, quarterly, or annual?
Monthly is standard for early-stage subscription businesses because it's the shortest period that still smooths some noise. If your customer count is under a few dozen, also report quarterly — a single cancellation swings a monthly rate far more than a quarterly one, and you'll chase less noise.
Does a free trial cancellation count as churn?
No — churn is measured against paying customers only. Someone who starts a 7-day trial and never converts never entered your churn denominator; that's a trial-to-paid conversion metric, a different number. Only count someone as churned once they've paid and then stopped.
My churn rate jumped this month because of one cancellation — should I panic?
Not automatically. Check whether it reflects something systemic (a bug, a price change, a competitor) or something idiosyncratic (an expired card, the customer's business closing). One data point isn't a trend — which is why cohort curves and multi-period trends matter more than any single month's blended rate.