TL;DR
- A SaaS cohort analysis template groups customers by subscription start month and tracks what percentage of their original MRR remains active each subsequent month.
- The template has two core tables: a gross revenue retention (GRR) cohort showing revenue remaining after churn and contraction, and a net revenue retention (NRR) cohort that adds expansion back in.
- Green benchmark: GRR above 90%, NRR above 110%. Yellow: GRR 80–90%, NRR 100–110%. Red: GRR below 80%, NRR below 100%.
- A healthy SaaS retention curve drops steeply in months one and two, then flattens into a plateau from month three onward. A curve that keeps declining at a consistent rate after month six signals a product-market fit problem.
- The most actionable read: compare each new cohort's Month 3 GRR to the same point in prior cohorts. If it is declining, something has changed — pricing, ICP fit, onboarding, or product quality.
Most SaaS operators track a blended net revenue retention number. Few track it at the cohort level, which means they cannot tell you whether Q1 customers retain better than Q3 customers, or whether a pricing change in March degraded long-term retention. The cohort table is what makes those questions answerable.
This template covers the revenue cohort table structure, the retention rate formulas for both GRR and NRR, how to interpret the shape of the retention curve, red/yellow/green benchmarks, and the three reads that turn the table into operating decisions.
What a SaaS cohort analysis template measures
Definition
A SaaS cohort analysis template groups customers by their subscription start month and tracks how much of their original MRR — and how many of their accounts — remain active over subsequent months. Each row is a cohort. Each column is a month of elapsed time. Each cell shows retention as a percentage of the cohort's original MRR or account count.
Unlike a blended NRR calculation, the cohort table separates each acquisition group onto its own timeline. This means a February cohort's Month 6 is directly comparable to a January cohort's Month 6. Blended NRR cannot give you that comparison. It averages across cohorts of different ages and mixes early-stage customers with mature ones, which masks the signal entirely.
The SaaS cohort table has two layers. The first is the logo (account) retention layer, which shows what percentage of accounts from each cohort are still active. The second is the revenue retention layer, which tracks MRR from each cohort — showing both GRR (churn and contraction only) and NRR (after expansion). Both layers are necessary because they answer different questions.
The revenue cohort table structure
The template below is the standard SaaS revenue cohort format. Rows are acquisition cohorts by month. Columns are months since first payment (M0 through M12+). Each cell shows the percentage of the cohort's original MRR still active in that period.
| Cohort | Starting MRR | M1 | M2 | M3 | M6 | M9 | M12 |
|---|---|---|---|---|---|---|---|
| Jan 2025 | $42,000 | 97% | 94% | 93% | 91% | 90% | 89% |
| Feb 2025 | $38,500 | 96% | 93% | 92% | 90% | 88% | -- |
| Mar 2025 | $51,200 | 94% | 89% | 85% | 83% | -- | -- |
| Apr 2025 | $44,800 | 93% | 87% | 82% | -- | -- | -- |
| May 2025 | $47,100 | 91% | -- | -- | -- | -- | -- |
Reading down the M3 column tells the story: January and February cohorts both held above 92% at Month 3 (green). March dropped to 85% (yellow). April slipped to 82% (red). Something changed between February and March — whether pricing, ICP targeting, onboarding, or product quality — and the cohort table surfaces it within 90 days.
Retention rate formulas: GRR and NRR
The cohort table uses two retention formulas. Each answers a different question.
Gross Revenue Retention (GRR)
GRR measures how much of a cohort's original MRR survives after accounting for churn and contraction, excluding expansion. It can never exceed 100% because it only measures losses.
GRR (Month N) = (Starting MRR − Churned MRR − Contracted MRR) ÷ Starting MRR × 100
Example: Cohort starts at $42,000. Month 6: $2,100 churned, $1,800 contracted. No expansions counted.
GRR = ($42,000 − $2,100 − $1,800) ÷ $42,000 × 100 = 90.7%
Net Revenue Retention (NRR)
NRR adds expansion MRR (upsells, seat additions, plan upgrades) back into the numerator. When expansion exceeds churn and contraction, NRR exceeds 100% — meaning the cohort is growing even as some customers leave.
NRR (Month N) = (Starting MRR − Churned MRR − Contracted MRR + Expanded MRR) ÷ Starting MRR × 100
Example: Same cohort. Month 6: $2,100 churned, $1,800 contracted, $6,300 expanded.
NRR = ($42,000 − $2,100 − $1,800 + $6,300) ÷ $42,000 × 100 = 105.7%
A cohort with 90% GRR and 106% NRR tells you that while 10% of the original MRR churned or contracted, the customers who stayed expanded enough to grow the cohort's total revenue by 6%. This is the pattern that produces compounding revenue even in the presence of churn.
Red, yellow, green benchmarks
Use the table below to benchmark each cohort's GRR and NRR at key elapsed months. Benchmarks are based on composite B2B SaaS data from SaaS Capital, Bessemer, and public company reporting.
| Metric | Green | Yellow | Red | What it signals |
|---|---|---|---|---|
| GRR at M12 | ≥ 90% | 80–90% | < 80% | Revenue durability after one renewal cycle |
| NRR at M12 | ≥ 110% | 100–110% | < 100% | Net expansion vs. churn in the base |
| Logo retention at M12 | ≥ 85% | 70–85% | < 70% | Account survival rate — customer outcomes |
| Monthly churn rate (SMB) | ≤ 3% | 3–7% | > 7% | Post-stabilization churn rate by segment |
| Monthly churn rate (Mid-market) | ≤ 1% | 1–3% | > 3% | Post-stabilization churn rate by segment |
Benchmark context
The 2025 median B2B SaaS GRR is approximately 90%, with the top quartile exceeding 95%. Median NRR is 101% — barely above flat — meaning most SaaS companies are only just covering gross churn with expansion. Best-in-class companies (often enterprise-focused) sustain NRR above 120%. If your NRR is flat, the problem is not churn: it is that expansion is not offsetting it.
How to read the retention curve
The retention curve is the visual representation of a single cohort's row in the table, plotted over time. The shape of the curve is the diagnostic signal — not just the level.
Shape 1: Steep drop then plateau (healthy)
A cohort that drops sharply in months one and two, then flattens from month three onward, is showing healthy behavior. The early drop represents customers who never fully activated — they signed up, evaluated the product, and left before building a workflow dependency. Once those customers are gone, the remaining cohort stabilizes because they have integrated the product into their operations. A plateau above 85% GRR by month three is a strong signal.
Shape 2: Steady linear decline (activation failure)
A cohort that declines at roughly the same rate each month — 2–3% every month with no plateau — signals that customers are not finding durable value. They are churning at a consistent rate that does not slow down as the cohort ages. This pattern is common in products that work for short-term use cases but lack the depth to drive long-term workflow dependency. The fix is typically in onboarding and product depth, not pricing or sales.
Shape 3: Cliff at month 12 (renewal friction)
A cohort that holds reasonably well through months 1–11, then drops sharply at month 12, typically signals annual contract friction. Customers on annual plans are re-evaluating at renewal time and deciding the product does not justify renewal. This pattern is common when sales teams close deals with heavy discounts or mismatched ICPs that are not caught until renewal. The fix is in renewal management, not product.
Key insight
A curve trending toward zero by month 12 is a product-market fit problem, not a churn problem. Labeling it as churn leads to wrong fixes — more CSMs, more check-ins, more discounts at renewal. The right fix is upstream: redefine the ICP, rebuild the onboarding, or narrow the product scope to the use cases that actually drive retention.
Three reads that turn the table into decisions
The cohort table is a diagnostic, not a report. Three specific reads generate the most operating decisions.
Read 1: Vertical comparison at Month 3
Read down the M3 column across all cohorts. Month 3 is the first meaningful post-activation checkpoint — customers who make it to month three have integrated the product into their workflow. If the M3 column is declining across successive cohorts, something has changed in the top of funnel: your ICP has drifted, your messaging is attracting the wrong buyers, or a product change has reduced early stickiness. This read tells you within 90 days of a change whether it is working.
Read 2: GRR vs. NRR gap within a cohort
Compare GRR and NRR for the same cohort at the same elapsed month. A wide gap (e.g., 88% GRR but 115% NRR) means your expansion motion is strong enough to offset significant churn. This is a viable operating position but a fragile one — it hides the churn from aggregate metrics. A narrow gap (e.g., 91% GRR and 93% NRR) means expansion is minimal. Customers are staying but not growing. That signals a missed upsell motion, limited product depth, or a pricing model that does not scale with customer value.
Read 3: GRR vs. logo retention divergence
If GRR falls faster than logo retention across a cohort, you are losing your highest-paying customers first. This is often a product-ICP mismatch at the high end of your customer base. If GRR holds better than logo retention, smaller accounts are churning while larger ones stay — a common pattern in upmarket-shifting companies. Each divergence pattern points to a different action.
How Fairview tracks cohort retention automatically
Building and maintaining a manual cohort table is a monthly project that most operators deprioritize within two quarters. Fairview connects to Stripe, HubSpot, and your billing system and builds the revenue cohort table automatically, refreshing as new subscription data arrives. Each cohort's GRR and NRR are calculated at every elapsed month, color-coded against benchmarks, and surfaced alongside the other operating metrics your team reviews weekly.
The Cohort LTV Tracker extends the analysis further: it segments cohorts by acquisition channel, plan type, and company size so you can compare whether enterprise cohorts retain at materially different rates than SMB cohorts. When a new cohort's Month 3 GRR falls below the threshold of the prior three cohorts, Fairview flags it as an anomaly in the weekly operating brief so your team can investigate before it compounds.
The three reads described above — vertical M3 comparison, GRR vs. NRR gap, and GRR vs. logo divergence — are surfaced as pre-built views inside the platform. Operators who previously spent a full day building a cohort table each quarter run the same analysis in minutes and redirect that time to the investigation the table demands.
Key takeaways
- The SaaS cohort template tracks what percentage of each cohort's original MRR remains after churn/contraction (GRR) and after expansion (NRR). Both metrics are required — one without the other hides half the picture.
- Green benchmarks: GRR at M12 above 90%, NRR at M12 above 110%, logo retention above 85%.
- The shape of the curve is diagnostic: plateau after month 3 is healthy, linear decline is an activation failure, and a cliff at month 12 is a renewal problem.
- The most actionable read is the vertical M3 column. If successive cohorts are declining at Month 3, something changed upstream within the past 90 days.
- Track GRR and NRR separately for the same cohort. A large GRR-to-NRR gap can mask churn problems that will surface when the expansion motion slows.