The Churn Number in a SaaS Case Study Depends on Which Churn It Means
A SaaS case study tells you churn went down. Down by which measure. That single word covers three separate calculations, and a campaign can look like a clear win on one while the other two are flat or worsening in the same quarter.
The word ‘churn’ is doing three different jobs
Logo churn counts accounts: the share, or raw number, of customers who cancel in a period, treating a two-seat account and a two-hundred-seat account as identical events. Revenue churn counts dollars, the monthly recurring revenue lost to cancellations and downgrades, so a large account cancelling moves this number far more than a handful of small ones leaving. Net revenue retention, usually shortened to NRR, starts from revenue churn and nets in expansion revenue, the extra money existing customers pay through upsells, seat growth, or plan upgrades. Because that expansion revenue gets added back, NRR can sit above 100 percent in the same period that some accounts are actively cancelling.
The mechanical reason these three drift apart is simple. A small account cancelling barely dents revenue churn, since there was never much revenue there to lose. A wave of upsells to accounts that stayed can outrun new cancellations entirely, pulling NRR up even as the logo count ticks down. Three numbers, one quarter of activity, and no requirement that they move together.
A hypothetical quarter, three different headlines
Picture a hypothetical SaaS company starting the quarter with 500 accounts and 200,000 dollars in monthly recurring revenue. These numbers are invented for illustration only, not a real company or reported result.
In this hypothetical quarter, 20 small accounts cancel, each paying roughly 50 dollars a month, for a combined loss of 1,000 dollars in MRR. At the same time, 15 mid-size accounts downgrade to a cheaper plan, each losing about 300 dollars a month, for a combined loss of 4,500 dollars in MRR. And 40 larger accounts expand, adding roughly 250 dollars a month each, for a combined gain of 10,000 dollars in MRR.
Reported as logo churn, this quarter looks strong: 20 cancellations out of 500 accounts is a 4 percent logo churn rate, and the accounts that left were the smallest in the book. Reported as revenue churn, the picture is flatter: 1,000 dollars lost to cancellations plus 4,500 dollars lost to downgrades is 5,500 dollars of revenue churn against a starting base of 200,000, worse than the logo number suggests, since the downgrades did more damage than the cancellations. Reported as net revenue retention, the quarter looks positive: 200,000 minus 5,500 in churn and downgrades, plus 10,000 in expansion, is 204,500, or just over 102 percent NRR. Same accounts, same quarter, three different stories.
A campaign given credit for ‘reduced churn’ in this hypothetical could point to any one of the three. Cite the logo number and it looks like the campaign kept small customers from leaving. Cite NRR and it looks like an unqualified win. Cite revenue churn, the figure reflecting what was actually lost in dollar terms, and the quarter reads as roughly flat at best. None of the three is wrong. Each answers a different question.
What a case study usually shows you, and what it usually doesn’t
‘Churn’ stated without a qualifier is undefined, and the reader has to resolve it. A case study that says only “churn fell” tells you less than it appears to: you do not know whether fewer accounts left, less revenue was lost, or a net figure includes money from customers the campaign never touched.
NRR is the number most flattered by activity unrelated to the campaign being credited. If accounts expand for reasons having nothing to do with a retention push, that expansion revenue still lands inside NRR and can carry the headline regardless of what happened to the accounts the campaign was trying to keep. A retention-focused campaign should be checked against logo churn or revenue churn, the figures that isolate what left, not against NRR alone.
The three questions that separate the definitions
- Is this a count of accounts, a count of dollars lost, or a net figure folding in expansion revenue from other customers.
- Does the stated period include expansion or upsell activity that could be masking cancellations underneath a headline number.
- Would the number move noticeably if one large account cancelled or expanded, and does the case study say anything about account size distribution.
Why this matters more in SaaS than most other categories
A metric name hiding more than one calculation is not unique to churn. The same pattern shows up in engagement rate is three different metrics wearing one name, where the denominator alone changes what the percentage means. What raises the stakes is that SaaS runs on subscription revenue, so a retention number is often the single headline result of a case study. See also why the best metric on a dashboard is rarely the real story. The reader’s task is not to distrust the number, but to ask which of the three it actually is.
Ready to check a claim against the source?
Browse the case study library, or submit your own. For a broader starting point, see how to read a marketing case study without being misled.
FAQ
What is the difference between logo churn and revenue churn?
Logo churn counts the number or percentage of accounts that cancel in a period, treating a small account and a large account the same. Revenue churn counts the dollars lost to cancellations and downgrades, so a large account cancelling moves this figure far more than several small accounts combined.
How can net revenue retention be above 100 percent if customers are cancelling?
NRR nets expansion revenue from existing customers, upsells, upgrades, seat growth, against the revenue lost to cancellations and downgrades in the same base. If accounts that stay spend enough more, the net figure can rise even while some accounts leave, as in the hypothetical quarter above.
Which churn number should I trust in a case study?
None of them alone. Ask which of the three was reported, and whether the other two were disclosed. A case study reporting only NRR for a retention focused campaign shows the number most flattered by unrelated expansion activity.
Why do SaaS case studies favor one churn definition, and why say churn improved without saying which churn?
Whichever number moved most favorably that period is likeliest to appear in the summary, and the other two go unmentioned. That is a general pattern with any metric having more than one accepted calculation, not evidence of dishonesty in any single case.