Churned Users
The number of active user seats lost in a period because customers cancelled their entire contracts and dropped to zero users.
◆ Count
Formula
Built from
What it measures
The absolute count of user seats that were live and are now gone because a customer cancelled their entire contract — their seat count dropped to exactly zero. Partial downsells (50 seats to 30) do not count; only full account terminations do. Measured per period (month, quarter, year).
Why it matters
You track Churned Users to measure the volume and velocity of full customer cancellations — how many accounts are leaving you completely. Churn is the largest brake on growth: a subscription business cannot grow faster than it replaces the customers it loses. Leadership uses Churned Users to forecast revenue impact, alert customer success to at-risk saves, and trigger product investigations. By isolating full cancellations from partial downgrades, you separate the signal of "contracts ending" from the noise of "a customer dropping a seat or two."
How to read it
Read Churned Users as your full-cancellation volume for the period, never as a single number in isolation. Divide it by Starting Users to get gross user churn rate, and compare this month to last month — a rising count means retention is deteriorating. Compare it to New Users: if new seats acquired fall below seats churned, you are in negative growth even if revenue looks flat. Break it down by cohort and segment to find where the bleeding happens, and always pair it with Churned MRR to separate volume from value — losing 50 enterprise seats hurts far more than losing 500 SMB seats.
What good looks like
Good
Churned Users runs under 5% of Starting Users for the period, and the count is flat or falling month over month — you are retaining the vast majority of your seat base.
Watch
Churned Users sits at roughly 5–10% of Starting Users, or the absolute count is creeping up; dig into which segments and cohorts are dropping to zero.
Bad
Churned Users exceeds 10% of Starting Users, or full cancellations are outpacing new seats acquired so net seat growth turns negative.
Watch-outs
- Conflating user churn with revenue churn. Churned Users measures volume; Churned MRR measures value. One enterprise cancellation can dwarf 200 SMB seats in dollars — always measure both.
- Counting downgrades as churn. If a contract shrinks from 50 to 30 seats, that is a downsell, not 20 churned users. Only count when seats drop to zero.
- Ignoring cohort retention. A spike in churned users might be normal attrition from an old cohort or a red flag from your newest one. Segment by cohort to know which.
- Reporting churned users without comparing to acquisition. Churning 100 seats while acquiring 150 is net growth; churning 100 while acquiring 80 is decline. Always pair it with New Users.
Worked example
Hypothetical
You open the month with 1,200 active seats across 60 contracts. During the month: Contract A (80 seats) cancels fully, Contract B (22 seats) cancels fully, Contract C downsizes from 50 to 35 seats, and Contracts D and E are each 15-seat contracts that cancel in full. Churned Users = 80 + 22 + 15 + 15 = 132 (Contract C's downsell does not count). Gross user churn rate = 132 ÷ 1,200 = 11%.
Variants & windows
The same metric re-expressed by a mechanical transform — a trailing window, a growth rate, a per-unit scaling, or a book/segment cut. Each is computed from Churned Users above.
- Churned Live Users Churned
- Churned Live Users (T3M) Churned · Trailing 3-month
- Churned Live Users (TTM) Churned · Trailing 12-month
- Churned Contracted Users (T3M) Churned · Trailing 3-month · Contracted book
- Churned Contracted Users (TTM) Churned · Trailing 12-month · Contracted book