Learn the net revenue retention formula, what good NRR benchmarks look like, and how to use customer intelligence to improve it. A practical guide for B2B SaaS leaders.
Net revenue retention (NRR) is a SaaS metric that measures the percentage of recurring revenue retained from an existing customer cohort over a set period, after accounting for expansions, contractions, and churn. An NRR above 100% means your existing accounts are generating more revenue than they did at the start of the period — without a single new logo. For B2B SaaS companies, NRR is the clearest signal of product-market fit at scale and the durability of your revenue base.
Net revenue retention — also called NRR net revenue retention or net revenue retention rate — answers one question: if you stopped acquiring new customers entirely, would your revenue grow, hold flat, or shrink?
NRR captures three dynamics in a single number:
• Expansion revenue — upgrades, upsells, seat additions, and usage overages from existing accounts
• Contraction revenue — downgrades and partial cancellations from existing accounts
• Churn revenue — ARR lost when accounts cancel entirely
A company with 105% NRR grows its revenue base from existing accounts alone. A company at 90% NRR must acquire enough new ARR each quarter just to replace what it lost — a treadmill that only accelerates as the base gets larger.
The lesson: NRR is not a customer success vanity metric. It is a forecast input, a retention audit, and a product health indicator all in one number.
The net revenue retention formula is straightforward:
NRR = (Starting MRR + Expansion MRR − Contraction MRR − Churned MRR) ÷ Starting MRR × 100
Most teams measure NRR on a monthly basis and then annualize it, though quarterly and annual NRR calculations follow the same structure. Here is what each variable means:
In this example, NRR is 104%. The company grows its existing revenue base by 4% per period before counting new logos. Over 12 months, compounding at that rate produces meaningful ARR growth from accounts that already exist in the CRM.
NRR vs. Gross Revenue Retention (GRR): GRR excludes expansion — it only measures how much of your starting revenue survived contraction and churn. GRR has a ceiling of 100%. NRR can exceed 100% because expansion is included. Both matter: GRR tells you how well you protect the base; NRR tells you how well you grow it.
NRR benchmarks vary meaningfully by company stage, average contract value (ACV), and go-to-market motion. The table below reflects data from public SaaS benchmarks, including OpenView's 2023 SaaS Benchmarks Report and Bessemer's State of the Cloud data.
The median NRR for public SaaS companies sits around 104–106%, per Bessemer's data. Top-quartile companies — those with strong expansion loops built into the product — consistently post NRR above 115%. Companies below 100% NRR rarely sustain growth rates that satisfy investors, because new ARR must first replace lost ARR before it contributes to growth.
The rule: If your NRR is under 100%, fix retention before you scale acquisition. Pouring new ARR into a leaking bucket accelerates burn, not growth.
Consider a B2B SaaS company with $5M ARR entering Q1 with 80 enterprise accounts. During the quarter:
• 12 accounts expand, adding $120,000 in new MRR
• 5 accounts downgrade, reducing MRR by $30,000
• 3 accounts cancel entirely, removing $50,000 in MRR
Starting MRR: $416,667 (≈$5M ARR ÷ 12). Ending MRR from that cohort: $416,667 + $120,000 − $30,000 − $50,000 = $456,667. NRR: $456,667 ÷ $416,667 × 100 = 109.6%.
That 9.6 percentage-point lift from existing accounts, compounded quarterly, produces more than $480,000 in incremental ARR over a year — before a single new deal closes. The expansion motion — not the retention motion — is doing the heavy lifting here.
This is why product leaders who tie roadmap decisions to expansion use cases tend to move NRR faster than those focused purely on churn prevention. Churn prevention keeps you above water. Expansion is what compounds.
NRR is a lagging indicator — the number you see today reflects decisions made 6 to 18 months ago. The inputs that move it are structural, not tactical.
• Usage-based or seat-based pricing — expansion revenue grows automatically as accounts scale
• Sticky workflows — products embedded in daily operations see lower contraction than those used periodically
• Proactive CS engagement — accounts that receive structured engagement before renewal are less likely to downgrade
• Roadmap aligned to expansion use cases — shipping features that unlock new teams, departments, or workflows inside an existing account drives upsell
• Closed-loop feedback programs — customers who feel heard reduce churn; customers who surface ignored needs quietly leave
• Misaligned onboarding — accounts that never reach full activation are at contraction risk within 90 days
• Feature gaps that competitors fill — lost expansion deals to adjacent products erode NRR without showing up as churn
• Opaque roadmaps — accounts that can't see what's coming have lower renewal confidence
• Reactive support models — problems surface at renewal, not at the moment they emerge
• The black hole effect — when customer feedback disappears without response, trust erodes and contraction follows
The connection between product feedback and NRR is direct, not theoretical. Across the 1,724 feature requests submitted in our feedback portal, the single highest-voted completed request — "Allow users to subscribe to suggestions" (558 votes, 338 supporters) — is a transparency feature. Customers demanded visibility into the roadmap and feedback status. Closing the loop on what gets built, and when, is one of the clearest trust signals a product team can send to an at-risk account.
NRR is simple in formula. It is not simple in practice. These are the measurement traps that cause teams to misread their own number.
NRR should be measured against a fixed cohort — the accounts that existed at the start of a period. Some teams accidentally include new logos in the "expansion" bucket mid-period, which inflates NRR. Lock the cohort on day one of the measurement window and do not touch it.
A seat reduction is not the same as a cancellation. Gross revenue retention (GRR) and NRR tell different stories when contraction is mislabeled as churn. Accurate segmentation matters — especially when diagnosing which accounts need intervention.
Billing data, CRM data, and CS platform data often sync on different schedules. An NRR number calculated on stale billing records overstates retention. Real-time or near-real-time data pipelines into your analytics layer are a prerequisite for reliable NRR tracking.
By the time an account cancels, it has usually sent 6 to 12 months of leading signals — reduced usage, support escalations, unanswered renewal outreach, or feedback submitted with no response. NRR captures the outcome. It does not surface the upstream signals that predicted it. That requires a layer of customer intelligence that most BI stacks don't provide.
Customer success teams own the renewal motion. Product teams own the reason customers renew — or don't. When NRR is tracked only inside CS platforms, product leaders lose visibility into the decisions they control. The features that unlock expansion, reduce friction, and address the gaps that cause contraction are product decisions. NRR belongs on the product team's dashboard too.
Improving NRR requires a strategy across three levers simultaneously: prevent churn, reduce contraction, and expand existing accounts. Most teams focus on the first and underinvest in the third.
The most effective change a product team can make is connecting feature requests to the accounts and ARR behind them. When a VP of Product can show the exec team that the top 10 unshipped requests represent $3.2M in at-risk ARR, prioritization arguments resolve themselves. Opinion-based roadmaps create prioritization fights. Revenue-weighted customer intelligence ends them.
In our feedback portal data, demand is concentrated: the top 10 requests hold 36.4% of all votes among the top 100 submissions. That concentration means a small number of shipping decisions have an outsized impact on the accounts driving the most engagement — and likely the most ARR.
Accounts that submit feedback and receive no response are experiencing what we call the black hole effect. They submitted a request. They heard nothing. At renewal, they have no evidence that the vendor listens — and a contraction or cancellation becomes easier to justify internally. Closing the loop is not about building everything customers ask for. It is about communicating what you are building, why, and when. That distinction matters enormously to a CS team managing a renewal conversation.
Expansion revenue comes from new workflows, new teams, and new capabilities inside existing accounts. If the roadmap is primarily focused on fixing existing functionality, expansion stalls. The highest-impact roadmap decisions for NRR are the ones that open a new department or unlock a new job-to-be-done for an account that already trusts you. Identify those use cases by analyzing which requests come from your highest-ACV accounts — not which requests have the most raw votes.
NRR is a lagging metric. The leading indicators — declining usage, unresolved support escalations, feedback submitted with no response, silence in QBRs — appear months before a cancellation. A customer intelligence layer that aggregates signals from support tickets, CRM notes, and feedback submissions into a single view gives CS and product teams the runway to intervene before it is too late.
Uservoice surfaces these signals by unifying feedback from portals, support, CRM, and sales conversations into revenue-weighted insights. When an account that represents $150,000 in ARR submits three feature requests and receives no status update, that is a retention risk — and it should appear on someone's dashboard before the renewal call.
Integration depth is a stickiness multiplier. Accounts deeply connected to your product through CRM, support, and development tool integrations churn at lower rates because switching costs are real. Among the top 100 requests in our feedback portal, integrations rank second in request volume with 16 entries — including GitHub issue tracking (172 votes) and Azure DevOps on-premises integration (157 votes). Those are not feature requests. They are signals about where accounts want deeper workflow embeddedness.
A public roadmap signals confidence and gives at-risk accounts a reason to stay. The "Public Roadmap" feature in our own feedback portal received 177 votes from 290 supporters before it was completed. Customers actively asked for transparency. Providing it removes a common objection at renewal: "We don't know where this product is going." Accounts that can see the roadmap renew with less friction.
Net revenue retention is the most honest measure of whether a B2B SaaS product creates enough value that customers expand, stay, and trust the vendor with more budget over time. The formula is simple. The management discipline behind it is not.
Companies that consistently post NRR above 110% share a common trait: their product and GTM teams operate from the same revenue-weighted view of what customers want. They know which requests come from which accounts, which gaps are driving contraction, and which roadmap decisions will unlock expansion. They do not guess. They do not rely on anecdotes. They build on signal.
NRR reflects the sum of those decisions — 6 to 18 months after they were made. The time to act on them is now, not at the next renewal cycle.
Net revenue retention (NRR) is a SaaS metric that measures the percentage of recurring revenue retained from an existing customer cohort over a given period, after accounting for expansion revenue, contraction, and full churn. An NRR above 100% means existing accounts are generating more revenue than they did at the start of the period — without any new customer acquisition. It is one of the most important indicators of product-market fit and long-term revenue durability for B2B SaaS companies.
The net revenue retention formula is: NRR = (Starting MRR + Expansion MRR − Contraction MRR − Churned MRR) ÷ Starting MRR × 100. Starting MRR is the recurring revenue from the existing customer cohort at the beginning of the measurement period. Expansion MRR is new revenue from those same accounts — upsells, seat additions, usage growth. Contraction MRR is revenue lost from downgrades. Churned MRR is revenue lost from full cancellations. The result is expressed as a percentage.
For B2B SaaS companies, an NRR between 100% and 110% is generally considered healthy, while anything above 110% is strong and above 120% is elite. The median NRR for public SaaS companies sits around 104–106%, based on Bessemer's State of the Cloud benchmark data. Companies below 100% NRR must acquire new ARR just to offset losses from existing accounts, which creates a significant growth headwind. The right benchmark depends on your ACV, go-to-market motion, and stage.
Gross revenue retention (GRR) measures how much of your starting recurring revenue survived after contraction and churn — it excludes expansion and has a ceiling of 100%. Net revenue retention (NRR) includes expansion revenue, so it can exceed 100%. GRR tells you how well you protect the existing base; NRR tells you how well you grow it. Both metrics matter: a company with high NRR but low GRR is masking heavy churn with aggressive expansion, which is an unstable revenue model.
Low NRR typically reflects a combination of product gaps, misaligned onboarding, and weak expansion motions. Common root causes include accounts that never reach full activation, feature gaps that competitors fill, opaque roadmaps that reduce renewal confidence, and reactive support models that surface problems too late. The 'black hole effect' — where customer feedback disappears with no response — is also a significant driver of silent contraction. By the time an account cancels, leading signals have usually been visible for 6 to 12 months.
Customer feedback is a direct input to NRR because the features a product team ships — or fails to ship — determine whether existing accounts expand, stay flat, or contract. Accounts whose top requests go unaddressed have less reason to expand and an easier internal justification for downgrading at renewal. Revenue-weighted feedback analysis — which ties feature requests to the ACV of the accounts behind them — helps product teams prioritize the roadmap decisions with the highest NRR impact, rather than optimizing for raw vote counts.
Most SaaS companies calculate NRR monthly on a rolling 12-month basis, which smooths out seasonal fluctuations and gives a stable trend line. Quarterly NRR calculations are common for board reporting. Daily or weekly NRR tracking is less useful because the metric lags the decisions that drive it by 6 to 18 months — but monthly visibility gives CS and product teams enough signal to identify deterioration before it compounds. The key is consistency in cohort definition across every measurement period.
Yes. High NRR from a small and declining new-logo pipeline creates a false sense of security. If expansion revenue is masking high churn — for example, one or two large accounts expanding aggressively while many smaller accounts cancel — the revenue base is more concentrated and fragile than the NRR number suggests. Always read NRR alongside GRR, new ARR trends, account concentration metrics, and logo churn rate to get the full picture of revenue health.
Turn scattered user data into meaningful customer intelligence, guiding smarter decisions and creating a better product.
Talk to an Expert
