Revenue Churn vs Logo Churn: Why the Two Rates Disagree

What is the difference between revenue churn and logo churn?

Logo churn counts customers lost. Revenue churn counts dollars lost. They measure the same departures against different denominators, and in most SaaS businesses they tell opposite stories.

A company starts the year with 200 customers and $20 million of ARR. It loses 24 customers worth $600,000 combined. Logo churn is 12 percent. Revenue churn is 3 percent. Both figures are accurate descriptions of the same twelve months. Which one you put in the board deck determines whether the year reads as a retention problem or a rounding error.

The right answer is to show both, because the gap between them carries the information. When logo churn runs far above revenue churn, you are losing small accounts. When revenue churn runs above logo churn, you lost something large.

What is logo churn?

Logo churn is the percentage of customers who left during a period, measured against the customer count at the start of the period. Divide customers lost by starting customers.

It treats a $2,000 account and a $2 million account identically. That sounds like a flaw and it is actually the point. Logo churn is a product and fit metric. If customers are leaving in volume, something about onboarding, value delivery, or targeting is broken, and the dollar weighting would hide it.

It is also the metric that predicts the shape of your business two years out. A land and expand motion depends on small accounts surviving long enough to grow. If those accounts are churning at 30 percent annually, the expansion engine has no fuel, and revenue churn will not show it until the cohort that would have expanded is already gone.

What is revenue churn?

Revenue churn is the percentage of recurring revenue lost during a period, measured against the ARR at the start of the period. It includes full churn and, in most constructions, contraction from downgrades.

Revenue churn is the number that drives the financial model. Gross revenue retention is its complement, and net revenue retention takes revenue churn and nets expansion against it. Every retention benchmark quoted in a fundraise references this side of the ledger.

Keeping contraction separate from full churn is worth the extra line. A customer who cut seats by 40 percent and a customer who left entirely both reduce revenue, and they call for entirely different interventions. Contraction is also the earlier signal. Customers usually downgrade before they leave.

How can both rates be true at once?

Because customer count and revenue concentration are independent. Two scenarios, same starting position. Scenario A looks fine on the financial model and is a genuine problem for the product organization. Scenario B looks fine on the customer count and is a genuine problem for the CFO. A company reporting only one of these rates cannot tell which of the two problems it has.

The diagnostic that resolves it is segmentation. Break both rates by ARR band, by acquisition channel, and by tenure. Churn is almost never evenly distributed, and the segment cut usually names the cause in one look.

Which rate should you report?

Report both, plus contraction, on the same view. The table below is the minimum useful set. Net revenue retention belongs on the list because it is the only figure that nets expansion against losses. A company can post 8 percent revenue churn and still grow the base if expansion covers it. That does not make the churn acceptable, and separating the lines is what keeps expansion from masking it.

What actually predicts churn before the rates move?

Support case volume, read as a curve rather than a straight line. The relationship is not what most teams assume.

Accounts with zero support cases are at risk. No cases means nobody is using the product deeply enough to hit a question, and a customer who is not using the product will not renew it. Accounts with seven or more cases in a year are also at risk, for the obvious reason. The safest accounts sit in the middle. Three to five cases in a year, typically tier two or three rather than severe, indicates an engaged customer who is getting support and generally happy.

That U-shaped pattern is useful precisely because it fires early. It moves months before a renewal date and long before either churn rate registers anything. A silence-based signal is also the earliest indicator on the sales side, where the absence of activity on a deal, meaning no stage changes, no date changes, no notes, is a worse sign than bad news.

Both rates are lagging by construction. They tell you what already happened to a cohort you can no longer influence. Any churn program that runs on the rates alone is reacting a full renewal cycle late.

How do churn rates feed the revenue forecast?

Forecast revenue churn against the base and logo churn against the customer count, then reconcile them. They are not interchangeable inputs and using one for both produces a model that drifts.

The retention layer of a revenue forecast is the most predictable part of the model, which is why it deserves its own construction rather than a blended assumption. Start from beginning ARR, apply churn and contraction, add expansion, and you have the base before a single new deal closes. Net revenue retention is the summary coefficient for that whole layer.

Logo churn feeds the model differently. It drives the customer count that determines how much expansion surface you have next year and how many accounts your CS team needs to cover. A forecast that only models dollars will miss a headcount problem building underneath it.

Segmenting both rates also protects forecast accuracy when conditions change. If a competitor enters and pressures pricing, contraction moves first in the segment they target. A blended churn rate averages that away for two quarters. Our guide on how to forecast revenue covers how the retention layer stacks against new business in a full model.