Cohort analysis is the practice of grouping customers or revenue by the period they were acquired and tracking how their behaviour evolves over subsequent periods. In B2B SaaS, a cohort analysis of ARR retention and expansion shows: how much of the ARR from each acquisition period (month or quarter) is still present 3, 6, 12, and 24 months later; whether churn is concentrated in specific cohorts (a bad cohort can signal a flawed product launch, a bad ICP batch, or a go-to-market change that acquired poor-fit customers); and whether expansion is making up for churn in each cohort over time. Cohort analysis is more diagnostic than aggregate retention metrics because it separates the signal of different periods rather than mixing all customers into a single average.
How B2B SaaS cohort analysis works
A typical B2B ARR cohort analysis: group all ARR signed in Q1 2024 as the "Q1 2024 cohort." Track the ARR from that cohort each subsequent quarter: at Q2 2024 (3 months), Q3 2024 (6 months), Q4 2024 (9 months), Q1 2025 (12 months). Plot retention as a percentage of starting ARR: if the Q1 2024 cohort started with 1 crore in ARR and retains 85 lakhs after 12 months and 92 lakhs after 18 months (because expansion exceeded churn), the cohort shows a "smiling curve" -- initial churn followed by expansion-driven recovery. Cohort curves that decline monotonically (continuing to shrink every period) indicate that churn is consistently outpacing expansion; cohort curves that flatten or rise indicate that expansion revenue from surviving customers is making up for lost accounts.
What cohort analysis reveals
- Retention trend over time: comparing retention of recent cohorts versus older cohorts at the same point in their lifecycle reveals whether product-market fit, onboarding, or customer success effectiveness has improved or degraded
- Bad cohort identification: a cohort that churns dramatically faster than surrounding cohorts indicates a problem in that acquisition period -- a specific campaign that brought in poor-fit customers, a product issue that affected customers onboarded in that period, or a customer success gap
- Expansion trajectory: whether expansion revenue from retained customers is building in each cohort over time; companies with strong expansion show cohorts growing past 100% of their starting ARR
- Payback period by cohort: when cumulative revenue from a cohort crosses the CAC invested to acquire it, the cohort is paid back; cohort analysis makes payback period concrete
- ICP validation: comparing retention of cohorts acquired from different channels, ICPs, or GTM motions reveals which customer profiles retain and expand most reliably
How to run a cohort analysis in B2B SaaS
The simplest cohort analysis requires: a customer list with start date and ARR; a record of monthly or quarterly ARR changes (expansions, contractions, and churn) at the account level. Using this data in a spreadsheet: (1) Group accounts by acquisition quarter; (2) For each subsequent quarter, sum the current ARR of accounts in each cohort; (3) Express each period's ARR as a percentage of the cohort's starting ARR; (4) Plot the table as a heatmap (green for values above 100%, yellow for 80-100%, red for below 80%) to visualise retention trends across cohorts. More sophisticated cohort analysis uses SaaS financial tools (Mosaic, ChartMogul, Paddle Revenue) that generate cohort views automatically from subscription data.
Frequently asked questions
- What is cohort analysis in B2B SaaS?
- Cohort analysis in B2B SaaS is the practice of grouping customers by the period they were acquired and tracking how their revenue behaviour evolves over subsequent periods. A cohort is a group of customers who started in the same period (e.g., all customers who signed in Q3 2024). Tracking the ARR of each cohort over time reveals: how much of the initial ARR is retained after 3, 6, 12, 18, and 24 months; whether expansion revenue from retained customers in the cohort is building over time; and how different cohorts compare to each other at the same point in their lifecycle. Cohort analysis is more diagnostic than aggregate retention metrics (like overall GRR or NRR) because it isolates the performance of specific acquisition periods rather than averaging them together. A company with a flat 90% GRR might have strong recent cohorts and a legacy bad cohort dragging the average down -- cohort analysis reveals this; aggregate metrics do not.
- How do you read a SaaS cohort analysis table?
- A SaaS cohort analysis table has: rows representing acquisition cohorts (Q1 2024, Q2 2024, Q3 2024, etc.); columns representing the age of the cohort (Month 1, Month 3, Month 6, Month 12, Month 18, etc.); each cell showing the ARR retained as a percentage of the cohort's starting ARR. Reading the table: a cell value of 100% means the cohort has retained exactly its starting ARR; above 100% means expansion has exceeded churn (net retention is positive); below 100% means churn has exceeded expansion; the diagonal from top-left to bottom-right shows the most recent data point for each cohort. Key patterns to look for: (1) Column trends: are month-12 retention rates improving for more recent cohorts compared to older ones? (2) Row patterns: does each cohort's retention stabilise after initial churn (typical of SaaS with good ICP fit) or continue declining (churn without stabilisation signals persistent value delivery problems); (3) Expansion curves: cohorts where values increase in later months ("smiling curves") indicate strong expansion revenue.
- Why is cohort analysis important for B2B SaaS?
- Cohort analysis is important for B2B SaaS because aggregate metrics (like overall GRR, NRR, and churn rate) blend the performance of different customer vintages together, masking important patterns. Cohort analysis is essential for: (1) Detecting deteriorating retention: if you acquired many customers in a high-growth quarter but those customers are churning faster than earlier cohorts, aggregate NRR may still look healthy while the underlying problem grows; (2) Validating improvements: if you improved onboarding in Q3, cohort analysis shows whether Q3 and later cohorts retain better than Q1 and Q2 cohorts at the same age; aggregate metrics change slowly and cannot isolate the impact of a specific improvement; (3) Investor conversations: sophisticated SaaS investors (particularly at Series A and above) ask for cohort analysis as part of the data room because it reveals the true durability of the revenue base; a company that can show improving retention in recent cohorts is a stronger investment than one with flat aggregate NRR; (4) CAC payback validation: cohort analysis that tracks cumulative revenue against CAC reveals the true payback period, which aggregate revenue metrics cannot show.
Keep reading
- B2B NRR: what net revenue retention is and how to improve it
- B2B SaaS metrics benchmarks: what good looks like at each stage
- Churn rate: what it is, how to calculate it, and how to reduce it
- B2B expansion revenue: how to grow revenue from existing customers
- SaaS metrics: the key metrics every B2B SaaS team should track