Latka logo
By Nathan LatkaFinance & Fintech3 min read

Revenue Churn: Gross vs. Net, the Formulas, and What Good Actually Looks Like

Logo churn tells you who left; revenue churn tells you what it cost. The formulas, the disclosed benchmarks, and the three churn-fighting systems real operators described on tape — adoption gates, no-quota success teams, and day-21 euphoria.

On this page
  1. Disclosed numbers, not survey averages
  2. The three churn-fighting systems on the record
  3. Working benchmarks

Revenue churn is the percentage of recurring revenue you lose in a period from cancellations and downgrades. It’s the dollar-weighted sibling of logo churn, and the more honest of the two: losing ten $100 customers and losing one $10K customer are the same logo-churn event and very different businesses.

Gross revenue churn = MRR lost to cancellations + downgrades ÷ starting MRRNet revenue churn adds expansion back; below zero, it flips into the metric everyone brags about — net revenue retention above 100%.

Gross revenue churn

Measures the leak — every dollar lost to cancellations and downgrades.

Net revenue churn

Measures whether the bucket still fills anyway — expansion added back against the leak.

Gross and net answer different questions — and quoting only net, the near-universal habit, can hide a real problem: 130% NRR over 25% gross churn is a company sprinting on a treadmill.

Disclosed numbers, not survey averages

CompanyGross revenue churnNet retentionWhen disclosed
ShowpadSingle digits annually130%Feb 2018
Salesloft~10–12% annually120%Jul 2019
UserTesting14% annuallyJust over 100%Spring 2020
Outreach“In the 140s”May 2019
Algolia“Net negative”120–125%Aug 2019
1Mind211%Mar 2026

Two structural patterns explain most of the spread. Price point: Webflow’s $40/month designer plans churned 4–5% of customers monthly while its hosting line ran “essentially zero” — prosumer subscriptions leak, infrastructure doesn’t. And segment: every operator above reported enterprise cohorts churning less than SMB, which is why blended numbers need the split attached before they mean anything.

The three churn-fighting systems on the record

The adoption gate

Outreach’s Manny Medina wouldn’t let customer success touch an account until an implementation manager got 70% of purchased seats performing daily “sales positive motions.” “While the deal is hot and the ink is still wet, that’s when you attack the adoption problem — so you don’t have to deal with it at renewal.” His 140s NRR, he insisted, was that gate and nothing cleverer. Full breakdown here.

Separation of church and sales

UserTesting’s Andy MacMillan ran a chief customer officer whose team sold nothing — “gross dollar retention is an outcome metric” of value delivered — while the sales team owned expansion. Salesloft’s Kyle Porter and Medina ran commission-bearing hybrids. The three tapes together are the whole industry debate in primary sources; all three cleared 100%+ NRR, which suggests the discipline matters more than the org chart.

Front-loaded value

Workboard’s Deidre Paknad aimed for sponsor “euphoria” on day 21 via paid coaching-led onboarding — churn prevention purchased by the customer, before the first renewal conversation could exist.

Working benchmarks

From the disclosures above and the wider dataset:

  • Enterprise SaaS — should hold gross revenue churn under ~10% annually; best-in-class runs low single digits.
  • Mid-market — 10–15% is normal and 20%+ is a product problem wearing a finance costume.
  • SMB and prosumer — tolerate 2–4% monthly only when acquisition is near-free, Webflow-style.

Net retention: 100% is the line between compounding and refilling; 120%+ is genuinely good; the 140s are world-class; and numbers like 1Mind’s 211% exist only in young cohorts of a category being born — enjoy them while they last, because the law of large cohorts always arrives. (Her cohort math is still worth studying.)

31%what 3% monthly churn compounds to annually — not 36%

The reporting discipline that makes any of this comparable: measure in dollars, state the period (monthly compounds), disclose gross next to net, and never count contraction as anything but churn. The companies quoted here got quoted because they put real numbers on the record; the numbers survived scrutiny because they were measured honestly.

Compare disclosed retention across thousands of SaaS companies at getlatka.com/saas-companies.

SourcesFounder disclosures on the record: Showpad (Feb 2018), Salesloft (Jul 2019), UserTesting (Spring 2020), Outreach (May 2019), Algolia (Aug 2019), 1Mind (Mar 2026); the Webflow and Workboard tapes; and the wider dataset.

Get the real numbers behind SaaS

CEO-confirmed revenue, growth, and valuation data for thousands of private SaaS companies.

Create Your Free Account →