The eight churn terms SaaS teams use interchangeably and should not: churn rate and retention rate, customer (logo) churn and revenue churn, gross and net, monthly and annual, negative churn, attrition, and how churn rate feeds customer lifetime value. Each with its formula, the conversion between them, and the mistake each one invites.
Short answer
What is the difference between churn rate and retention rate?
Retention rate = 100% − churn rate, on the same cohort over the same period. The traps are elsewhere: logo churn counts accounts, revenue churn counts money, and they diverge whenever customers differ in size; annual churn is 1 − (1 − monthly)^12, not monthly × 12; and negative churn is only possible for revenue, never for logos.
Eight terms, four pairs, and a meeting that goes wrong when two people in it are using different ones. This is the short reference: each definition, the formula, the conversion where there is one, and the mistake each term invites.
Churn rate is the share of a starting population lost during a period. Retention rate is the share kept. On the same cohort and period they sum to 100%, so they carry the same information. Which one you quote is a framing choice: retention sounds better, churn is easier to act on because it names the loss.
The mistake: quoting retention monthly and churn annually, or the reverse, and comparing. 95% monthly retention sounds like a strong number and is 54% annual retention, which is not.
Logo churn = accounts cancelled ÷ accounts at start. Every account counts once, whatever it pays. Revenue churn = recurring revenue lost to cancellations and downgrades ÷ recurring revenue at start. A $50 a month account and a $5,000 a month account are one logo each and a hundred-fold apart in revenue.
| What happened (100 accounts, $100K MRR at start) | Logo churn | Revenue churn |
|---|---|---|
| Ten $200 accounts cancel | 10% | 2% |
| One $20,000 account cancels | 1% | 20% |
| Nobody cancels; five accounts downgrade by $2,000 each | 0% | 10% |
The mistake: reporting one and believing it is the other. A self-serve product with many small accounts will always show logo churn well above revenue churn; an enterprise product will show the opposite. The third row is the one that hides: downgrades produce revenue churn with zero logo churn, and a team watching logos only sees nothing.
Gross revenue churn counts only losses: cancellations and downgrades. Net revenue churn subtracts expansion from existing customers. Gross is the leak; net is the leak minus the inflow. Net can be negative. Gross cannot. The retention forms of the same two numbers are GRR and NRR.
The conversion is annual = 1 − (1 − monthly)^12. Each month's churn applies to the survivors of the previous months, so the base shrinks and the losses compound rather than add.
| Monthly churn | × 12 (wrong) | Compounded (right) | Annual retention |
|---|---|---|---|
| 1% | 12% | 11.4% | 88.6% |
| 2% | 24% | 21.5% | 78.5% |
| 3% | 36% | 30.6% | 69.4% |
| 5% | 60% | 46.0% | 54.0% |
| 8% | 96% | 63.2% | 36.8% |
The mistake: multiplying by twelve. It overstates annual churn, by 14 points at 5% monthly, and since most published benchmarks are annual and most dashboards are monthly, a team can spend a quarter fixing a number that was fine. The reverse conversion, from annual to monthly, is 1 − (1 − annual)^(1/12).
Negative churn is net revenue churn below zero: expansion from existing customers exceeded what cancellations and downgrades took away, so the base grew without a single new sale. It is the same fact as net revenue retention above 100%, stated the other way round: −10% net churn is 110% NRR.
The mistake:saying “negative churn” about logos. It cannot happen. Accounts either stay or go; a cohort cannot end the year with more members than it started with. Negative churn is a revenue statement only.
In the simplest model, LTV = ARPA × gross margin ÷ churn rate, with ARPA (average revenue per account) and churn on the same period. The division works because 1 ÷ churn is the expected number of periods an account survives: at 2% monthly churn, 50 months. The relationship is inverse and steep. Halving churn doubles LTV; going from 3% to 2% monthly lifts it by half.
The mistake: computing LTV from logo churn for a book where revenue churn is the binding number, or the reverse. Use revenue churn and revenue per account together, or logo churn and account count together. Mixing them produces a lifetime for a customer that does not exist. The fuller model uses net revenue churn, which allows expansion to lengthen the effective lifetime and is the reason negative-churn businesses show LTV figures that look impossible.
Same shape, different fields. Churn is the SaaS and subscription term for lost customers. Attrition is the HR term for employees who leave, and older telecoms usage for subscribers. In a SaaS conversation, say churn; if you are reading about attrition rate, check whether the page is about people.
All eight terms share one requirement: a fixed starting population. The denominator is the accounts, or the revenue, on day one of the period, and nobody who joined later is in it. The most common distortion in every one of these metrics is a denominator that includes the period's new sales. If the data lives in a billing system, the Stripe measurement notes cover the six ways that goes wrong; for the benchmarks by segment, see the churn benchmarks, and to run the arithmetic on your own figures, the churn calculator.
Churn rate and retention rate are complements: retention rate is 100% minus churn rate, measured on the same cohort over the same period. A 5% monthly churn rate is a 95% monthly retention rate. The two are not interchangeable across periods, though: 95% monthly retention is 54% annual retention, so a monthly retention figure always looks better than the annual one it implies.
Customer churn (logo churn) counts accounts that cancelled as a share of accounts at the start of the period, regardless of size. Revenue churn counts the recurring revenue lost to cancellations and downgrades as a share of starting revenue. They diverge whenever customers differ in size: a SaaS business losing many small accounts has high logo churn and low revenue churn, and one losing a single large account has the reverse.
Annual churn is 1 minus (1 minus monthly churn) to the twelfth power, not monthly churn times twelve. 3% monthly churn is 30.6% annual, not 36%; 5% monthly is 46%, not 60%. The compounding form is correct because each month's churn applies to the customers who survived the previous months, so the base shrinks as the year goes on.
Negative churn, or negative net revenue churn, is when expansion revenue from existing customers exceeds the revenue lost to cancellations and downgrades in the same period, so the existing customer base grows in revenue terms with no new sales. Negative churn of minus 10% annually is the same thing as net revenue retention of 110%. Logo churn can never be negative; only revenue churn can.
Logo churn in SaaS is customer churn counted by account: the number of customer accounts (logos) that cancelled in a period divided by the number active at the start. It ignores how much each account paid. Logo churn is the right metric for understanding whether the product retains customers; revenue churn is the right one for understanding whether it retains money.
Customer lifetime value is inversely proportional to churn rate: in the simplest model, LTV equals average revenue per account per period, times gross margin, divided by churn rate per period, because 1 divided by churn is the expected lifetime in periods. Halving churn doubles LTV. That is why a point of churn is worth more than a point of acquisition efficiency in most SaaS models.
Churn rate and attrition rate describe the same thing, the share of a population lost over a period, but the words belong to different fields. Churn is the SaaS and subscription term for customers; attrition is the HR term for employees and the older telecoms term for subscribers. In a SaaS context, use churn; a reader who searches for attrition rate usually wants the employee metric.
Keep reading
Customer success
10 min read · Sep 17, 2026
Retention metrics
9 min read · Sep 17, 2026
Retention metrics
10 min read · Sep 17, 2026
First scores in 15 minutes. Full accuracy in 24 hours. From $299/mo. Logo and revenue churn read from billing, involuntary churn separated, saves drafted for approval.
Not ready to switch? Size your leak from your MRR and churn rate, with no billing access at all.