NRR is the single best health metric in SaaS. Above 100% means your existing base grows on its own through expansion, even before new sales — the holy grail.
Example: ($100k + $12k expansion − $3k contraction − $5k churn) ÷ $100k × 100 = 104% NRR. Best-in-class is 120%+.
Above 100% = your base is growing without new logos.
Example — NRR in practice
A Bahraini fintech selling expense-management software to SMEs starts the year with $2.0M in ARR from existing customers. Over 12 months, upsells to a premium tier add $300k, while downgrades cost $50k and churn removes $150k. Net Revenue Retention is (2,000,000+300,000−150,000−50,000)/2,000,000 = 105%, meaning the existing base alone grew revenue even before counting a single new logo.
تبدأ شركة تقنية مالية بحرينية تبيع برنامج إدارة المصروفات للشركات الصغيرة والمتوسطة العام بإيراد متكرر سنوي قدره 2.0 مليون دولار من عملائها الحاليين. خلال 12 شهراً، تضيف الترقيات إلى الباقة المميزة 300 ألف دولار، بينما تخصم التخفيضات 50 ألف دولار ويُفقد بسبب التسرب 150 ألف دولار. صافي معدل الاحتفاظ بالإيراد = (2,000,000+300,000-150,000-50,000)/2,000,000 = 105%، أي أن القاعدة الحالية وحدها زادت الإيراد قبل حتى إضافة عميل جديد واحد.
NRR, properly understood
NRR measures revenue change from the existing customer base only — no new logos in the calculation. Start with beginning-of-period MRR (or ARR) from that cohort, add expansion revenue (upsells, seat growth, cross-sell), subtract contraction (downgrades) and subtract churned revenue (customers who cancelled), then divide by the starting MRR: (Start + Expansion − Contraction − Churn) ÷ Start × 100. Data sources are the billing system (Stripe, Chargebee, a custom subscription ledger) filtered to a fixed customer cohort at the start of the period — the single most common calculation error is letting new customers who signed up mid-period leak into the numerator, which artificially inflates the score.
NRR is one of the clearest signals of whether a Gulf B2B SaaS business is actually sticky, because regional sales cycles are long and CAC is high — a company that can expand revenue inside existing accounts is far less dependent on a brutal new-logo pipeline. Watch for FX and currency-mix effects if a book of business spans SAR, AED, and USD-denominated contracts, since a currency move can shift the reported number without any change in underlying usage. Ramadan and the summer slow season (many Gulf B2B buyers are on reduced schedules July-August) can compress the timing of renewal conversations, so a quarterly NRR read taken right after either window can understate the real trend — a trailing 12-month NRR is more reliable than any single quarter for a Gulf-based book.
A single large enterprise account renewing or expanding can swing NRR dramatically for an early-stage company with a small logo count — always check NRR alongside logo count and concentration (top-3-account revenue share) before trusting the aggregate number. NRR over 100% is good but doesn't mean the business is healthy in isolation: a company can post 115% NRR while still burning cash if new-logo acquisition and gross margin are both weak. And NRR calculated on gross revenue (ignoring the cost to serve that expansion) can mask a business where expansion revenue is unprofitable — pair it with a margin view of the expanded accounts, not just the topline retention number.
Pair NRR with Gross Revenue Retention (GRR — the same formula but excluding expansion, so it isolates pure churn/contraction without upsell masking it), with logo retention (customer count, not revenue), and with the Rule of 40 and CAC payback for a full health picture. A company can have strong NRR and weak GRR simultaneously — a few accounts expanding hugely while the median account is actually shrinking — which only shows up when both metrics are tracked side by side.
Put it to work
- Fix the cohort at period start — never let mid-period new logos leak into the numerator.
- Calculate GRR alongside NRR so expansion in a few accounts can't mask churn in the base.
- Check concentration: exclude or flag your top 1-3 accounts and recompute — does NRR still hold?
- Use trailing 12-month NRR, not single-quarter, for GCC books affected by Ramadan/summer renewal timing.
- Track expansion revenue's gross margin separately — 100%+ NRR on unprofitable upsells isn't a win.
- Reconcile FX-driven MRR moves separately from usage-driven moves for multi-currency books.
- Report NRR alongside logo count in the same cohort — a rising percentage on a shrinking customer base is a different story than a rising percentage on a stable one.
Turn the theory into real pipeline.
Get a free 60-second growth audit of your site, or talk to a strategist about your funnel.