Gross retention and net retention are the cousins of GDR and NDR — they answer the same fundamental question but can be applied to either customer counts or revenue. The distinction is usually clear from context: SaaS conversations default to dollar (revenue) retention; subscription commerce conversations sometimes mean customer retention. Either way, the gross vs. net split is the same.
The two perspectives
- Gross retention — what fraction of existing customers (or revenue) stayed. Excludes expansion. Caps at 100%.
- Net retention — what fraction stayed plus what existing customers added through upsell or pack expansion. Includes expansion. Can exceed 100%.
Worked example with customer counts
You start the period with 1,000 subscribers. 80 churn and 30 upgrade to a higher tier or larger pack size.
- Gross customer retention = (1,000 − 80) ÷ 1,000 = 92%.
- Net customer retention typically equals gross customer retention, because customer counts do not increase with upsell (the customer is still one customer). For meaningful net retention you need to look at revenue.
For dollar retention, the same 1,000-subscriber base might generate $40,000 MRR, lose $3,500 to churn, gain $4,000 from existing subscriber upgrades, giving gross dollar retention 91% and net dollar retention 101%.
Why the split matters
The same logic as GDR vs NDR. Net retention alone can hide a churn problem; gross retention alone misses the upsell story. Best-in-class subscription companies report both, and investors increasingly require both. The pattern that wins valuation is high gross plus high net — high net with weak gross is expansion-dependent and brittle.
What healthy looks like
- SaaS: Gross 92-95%, Net 110-130%.
- Shopify subscription stores: Less standardly tracked, but applying the framework — gross monthly customer retention 90-95%, net dollar retention modestly higher with meaningful upsell or pack-size growth.
- Public benchmark: SaaS companies with high gross plus high net retention trade at premium multiples, often 2-3x the multiple of weaker-retention peers.