Gross vs Net Revenue Retention

Gross Counts What Survived; Net Adds What Expanded

Gross revenue retention answers one question: of the recurring revenue you had from a defined set of customers a year ago, how much is still there after churn and downgrades? Expansion is deliberately left out, so the figure cannot exceed 100%. Net revenue retention asks a second question on the same cohort: after those losses, how much did the remaining customers spend or commit, including upsell, more seats, more consumption and price? That figure can sit well above 100%.

Mixing the two in a peer table is the usual error. A 98% renewal rate and a 125% net retention rate are not two readings of the same dial.

The Dictionary Filers Use

CrowdStrike and Workday publish definitions that draw the line in plain language. They sit outside this guide’s peer set but give the cleanest filed wording of the split.

Both houses start from an existing cohort and refuse to credit new logos. The only mechanical difference that matters for modelling is whether expansion is inside the retained base.

Attrition and Renewal Are Gross-Style Measures

Salesforce and ServiceNow do not publish a net retention percentage in the filings used here. They publish loss-side metrics that behave like gross retention under other names.

An 8% attrition rate implies that about 92% of the measured base was not lost to attrition, before any expansion. A 98% renewal rate is the same idea stated from the keep side. Neither figure tells you how much the survivors expanded. For that you need a net measure, or a separate expansion assumption in the model.

The definitions hub places these next to Snowflake’s two-year net retention and the three filers that disclose no retention rate at all.

Why the Split Matters in a Model

A recurring-revenue roll-forward needs both pieces, even if the filer only prints one number. Gross retention (or attrition / renewal restated as a keep rate) sets how much of the opening base survives. Expansion is then either implied by net retention minus gross retention, or typed as its own line. If you feed a net retention figure into a slot that expected a gross rate, you will double-count expansion wherever the model also has an upsell term.

Where Comparisons Break

Putting Salesforce attrition beside a peer’s net retention without a footnote implies a quality gap the filings do not support. One is a loss rate; the other is a growth rate on the retained book. ServiceNow’s renewal rate is closer to gross than to net, and still sits on lost ACV rather than on ARR the way CrowdStrike’s gross retention does.

Usage-based pricing stretches the same vocabulary further: consumption can raise or cut net retention without any logo churn, which is why a net figure on a capacity cohort swings harder than a seat-based renewal rate.