A SaaS board that tracks only NRR can be lulled into complacency by heavy expansion covering a churn problem. GRR exposes the churn. Both numbers side by side tell the real story.
NRR (Net Revenue Retention) and GRR (Gross Revenue Retention) both measure how much of your existing recurring revenue you keep over a period. NRR includes expansion (upsells, cross-sells); GRR excludes it. NRR can exceed 100 percent if expansion outweighs churn. GRR cannot exceed 100 percent. Together, the pair tells you whether retention is strong because expansion is masking churn, or because the base is genuinely sticky.
Six things to understand about NRR and GRR: what each measures, how to calculate them, the healthy benchmarks, why mature teams track both, how expansion masks churn, and the common calculation mistakes.
NRR measures the recurring revenue you kept from a cohort over a period, including expansion (upsells, cross-sells, usage increases) minus downgrades and churn. If a cohort started at $1M and ended at $1.1M after expansion and churn, NRR is 110 percent.
GRR measures the recurring revenue you kept from a cohort over a period, excluding any expansion. Only downgrades and churn count. If a cohort started at $1M and ended at $900K after churn (ignoring any expansion), GRR is 90 percent. GRR is bounded by 100 percent.
Healthy SaaS benchmarks: NRR above 110 percent (expansion outweighs churn), GRR above 90 percent (churn is low). World-class SaaS runs NRR above 120 percent and GRR above 95 percent. The gap between NRR and GRR shows how much of retention comes from expansion versus pure stickiness.
A business with 110 percent NRR looks healthy. If GRR is 85 percent, retention looks less healthy: 15 percent churn is being masked by 25 percent expansion. The expansion is impressive, but the churn risk compounds. Boards that track only NRR miss the pattern until expansion stalls.
GRR is driven mostly by the core product value: do customers get enough value to renew. NRR adds the account expansion motion: upsell, cross-sell, usage-based growth, multi-product adoption. A strong GRR plus a strong expansion motion equals a strong NRR. Fix GRR first.
The most common error is calculating retention on inconsistent cohorts: starting MRR vs ending ARR, or including new customers in the ending period. The denominator should be the starting MRR or ARR of the cohort. The numerator should be the ending revenue of the same cohort, with no new customers included.
The formulas are simple. The inputs require care: which revenue, which cohort, which period, and whether expansion counts. Below is the practical calculation side by side.
NRR = (starting ARR + expansion - contraction - churn) / starting ARR. Starting ARR is the revenue from the cohort at the beginning of the period. Expansion is upsells, cross-sells, and usage increases within the cohort. Contraction is downgrades. Churn is lost customers. The ratio is usually quoted as a percentage.
GRR = (starting ARR - contraction - churn) / starting ARR. Same starting ARR as NRR. Expansion is excluded. GRR cannot exceed 100 percent because expansion is left out. A GRR of 100 percent would mean zero churn and zero contraction.
Both metrics measure the same cohort: customers in the business at the start of the measurement period. Customers acquired during the period do not count toward NRR or GRR. The point is to measure retention of existing revenue, not blended growth.
The standard period is 12 months. Shorter periods (quarterly) have more volatility and less meaning. Longer periods are rare because they aggregate too much market change. Both metrics are typically reported annually, with quarterly tracking of underlying components (expansion, churn, contraction).
Both metrics are dollar-based: they measure revenue retained. Logo retention (percent of customers kept, regardless of size) is a separate metric. Dollar-based NRR and GRR capture the economic impact. A business can have 90 percent logo retention and 110 percent dollar retention if the retained customers are the biggest.
NRR and GRR vary by segment. SMB often has lower retention (higher churn) than enterprise. Enterprise expansion drives NRR higher. Reporting one blended average hides the segment where unit economics are breaking. Boards expect to see segment-level retention in mature reporting.
The benchmarks below reflect what investors look for in SaaS at various stages. Early-stage companies are given more latitude. Mature companies with poor retention lose multiple-expansion opportunities in financing rounds.
GRR below 80 percent means 20+ percent of revenue is churning annually. The product is not sticky enough for the price being charged. The fix is almost always product-side: better onboarding, deeper usage, stronger stickiness features. Expansion cannot compensate indefinitely.
Common range for mid-market SaaS. Retention is okay but could be better. Investors look for a path to 90 percent plus. Churn-reduction investment (customer success, onboarding, product stickiness) is usually the highest-leverage move.
Standard "good" GRR range. Churn is low enough that growth comes mostly from expansion and new sales rather than fighting attrition. Enterprise B2B SaaS often lives here. Mid-market SaaS targeting this range usually has above-average onboarding and support.
NRR below 100 percent means the cohort shrinks over time: more revenue is lost to churn than gained from expansion. The business needs new acquisition to grow, with no help from the existing base. Not sustainable at scale.
The cohort retains its original revenue and adds a bit. Common for mature SaaS without strong land-and-expand motion. Fine, but boards want to see higher. Expansion motion (upsell, cross-sell, usage pricing) is usually the lever.
The cohort grew 20 percent despite any churn. Reflects strong product, pricing power, and expansion motion. SaaS companies with NRR above 120 percent at scale tend to see premium valuation multiples because growth compounds from the base, not from acquisition spend.
Improving GRR and NRR requires different investments. GRR is a product and onboarding problem. NRR adds expansion motion on top. Below is how to pull each lever.
The GRR lever. Customers churn when they do not get value. Fix onboarding so time-to-value drops. Build usage features that create stickiness (data imports, custom fields, integrations the customer configures). Mature programs track time-to-value and act when it extends.
Customers who downgrade mid-cycle hurt both GRR and NRR. Identify downgrade drivers (feature not used, seat reduction, usage drop) and intervene before the renewal conversation. Customer success teams who proactively reach out at the first downgrade signal save material revenue.
The NRR lever. Multi-product companies grow NRR by cross-selling (CRM customer buys Marketing, then Messaging). Tier-based companies grow by upselling. Usage-based pricing creates automatic expansion as customers scale. A deliberate expansion motion (not accident) is what pushes NRR above 120.
Pricing structures that scale with customer success (per-seat, per-usage) create automatic NRR growth. Flat-rate pricing produces zero expansion from the same customer: you grow NRR only through tier upgrades or cross-sell. Pricing design is a strategic retention lever.
Strkr tracks every account relationship, usage pattern, and expansion opportunity in one data model. NRR, GRR, churn drivers, and expansion motion visibility all run on the same data.
NRR (Net Revenue Retention) includes expansion revenue (upsells, cross-sells, usage increases). GRR (Gross Revenue Retention) excludes expansion. NRR can exceed 100 percent if expansion outweighs churn. GRR cannot exceed 100 percent. Together, they show whether retention is driven by expansion or by a genuinely sticky base.
Healthy SaaS runs NRR above 110 percent. World-class runs above 120 percent. Below 100 percent means expansion does not cover churn, which is unsustainable at scale. The right benchmark depends on business model: expansion-focused models (usage pricing, multi-product) run higher than flat-rate ones.
Healthy SaaS runs GRR above 90 percent. World-class runs above 95 percent. Below 80 percent is a product-stickiness problem that expansion cannot compensate for indefinitely. Enterprise B2B typically has higher GRR than SMB because switching costs are higher.
NRR alone can mask churn. A business with 110 percent NRR and 85 percent GRR has strong expansion but a churn problem. Boards tracking only NRR miss the pattern until expansion stalls. Reporting both numbers side by side exposes the real retention picture.
NRR = (starting ARR + expansion - contraction - churn) / starting ARR. Measure the cohort at the start of the period, track movement over 12 months. Expansion includes upsells, cross-sells, and usage increases from the cohort. Contraction and churn come out. The result is usually a percentage above or below 100.
GRR = (starting ARR - contraction - churn) / starting ARR. Same cohort as NRR, same period, but expansion is excluded. GRR cannot exceed 100 percent because the only movements counted are negative (churn and contraction). GRR of 95 percent means the business kept 95 percent of the starting revenue, ignoring any upsells.
Strkr unifies CRM, marketing, messaging, and projects on one data model, so customer success teams see every signal (usage, support activity, deal history) in one record. Churn drivers surface earlier, expansion opportunities become visible without a separate account-planning tool, and retention metrics live next to pipeline.
No sales call, no migration consultant, no four-month implementation. Enter your card, get 14 days of the full Pro tier, cancel any time before day 14 with zero charge. Spin up a workspace, import your CSV, and have something useful before lunch.
