Skip to main content

Cohort Analysis

Quick definition: Cohort analysis groups customers by acquisition date (e.g., "Q1 2025 cohort") and tracks their subsequent behavior—retention, expansion, churn—to isolate whether newer customers are higher-quality or lower-quality than older ones.

Key Takeaways​

  • Cohort analysis is the single best diagnostic tool for evaluating SaaS business health because it separates acquisition quality from retention and expansion quality
  • A company with all high-quality cohorts (strong retention and expansion across all acquisition periods) is sustainable; one with declining cohort quality is heading toward trouble
  • Declining retention in newer cohorts signals potential product issues, market saturation, or customer acquisition quality degradation
  • Improving expansion in newer cohorts (higher ACV growth, faster upsells) indicates improving product fit or pricing leverage
  • Cohort curves are the foundation of LTV calculations and growth modeling, determining whether a company can profitably acquire customers at scale

Understanding Cohort Grouping​

A cohort is simply a group of customers acquired during the same time period. Most companies organize cohorts monthly or quarterly. A "Q1 2025 cohort" includes all customers who started paying in January, February, or March 2025.

Once defined, the cohort is tracked forward. How many of the Q1 2025 customers are still active in April 2025? In July 2025? In January 2026? What was their monthly churn rate? What was their average expansion rate—the percentage by which they increased spending?

Cohort analysis reveals whether a company's business model is working consistently across time. If the Q1 2025 cohort has identical retention and expansion curves to the Q4 2024 cohort, the company has found a repeatable process. If Q1 2025 cohort retention is significantly lower, something changed—the product, the market, the customer segment being acquired, or the sales process.

Retention Curves and Cohort Quality​

The most visible cohort metric is the retention curve: the percentage of each cohort still active as weeks or months pass since acquisition.

A healthy B2B SaaS retention curve for mid-market shows roughly 95% of the cohort surviving Month 1, 92% by Month 3, 85% by Month 6, and stabilizing around 75-80% by Month 12. This means the company loses roughly 2% of each cohort in Month 1 (early-stage churn from onboarding failures), then 1-2% monthly thereafter.

By contrast, a weak retention curve might show 90% Month 1, 80% Month 3, 60% Month 6, and 40% Month 12. This company is losing customers rapidly and continuously, indicating either product issues or poor customer fit.

Comparing retention curves across cohorts reveals whether business health is improving or deteriorating. A company where every cohort shows identical 95% Month 1 retention is consistent. One where Q4 2024 showed 95% Month 1 retention but Q1 2025 shows 88% is experiencing a deterioration—possible signals include product quality issues, a shift toward lower-quality customers, or competitive pressure.

The best growth companies show improving retention curves over time. Q4 2024 cohort: 95% Month 1 retention. Q1 2025 cohort: 97% Month 1 retention. This reveals that the company is getting better at onboarding, product fit, or customer selection.

Expansion Revenue and Customer Growth​

Beyond retention, cohorts are tracked by their expansion revenue—the revenue generated from existing customers beyond their initial contract value. A customer acquired for $5,000 annually who is now paying $7,000 annually has contributed $2,000 in expansion revenue.

Expansion revenue is often the unsung hero of SaaS growth. A company with 90% gross retention but 15% expansion revenue (existing customers increasing spend by 15% on average annually) actually has 103.5% net retention (90% of customers stay, and they increase spend by 15%, creating net positive growth from the existing base).

Cohort analysis reveals whether expansion is consistent across cohorts or declining. A company where the Q4 2024 cohort achieved 12% expansion by Month 12 and the Q1 2025 cohort is tracking to similar levels has consistent expansion. One where Q4 2024 achieved 12% but Q1 2025 is tracking to 8% is concerning—either older customers are more "stickier" and easier to upsell, or the newer cohort is lower-quality and less engaged.

The most important expansion metric is net revenue retention (NRR). If a company had 100 existing customers in Month 1, all paying $1,000 monthly, and 90% remain in Month 2 (all still paying $1,000) plus the remaining 10 customers increase to $1,100, the cohort's MRR would have moved from $100,000 to $100,000 (90 * $1,000 + 10 * $1,100). Net revenue retention is 100%, showing that the cohort is generating zero net new value.

But if the 90 remaining customers all increase to $1,050 monthly, MRR grows to $94,500 + $1,000 = $95,500. Net retention is 95.5% (declining due to churn). However, with expansion included, it might reach 100-105%, showing that despite some churn, the cohort is net positive due to expansion.

Cohort Quality and Customer Acquisition Efficiency​

Cohort analysis directly informs whether a company can sustain its growth and spend on customer acquisition.

Consider a company acquiring customers at $10,000 CAC (customer acquisition cost) with a one-year payback period. If the cohort has 90% Year 1 retention and the customer pays $1,000 monthly, the company recovers the acquisition cost ($10,000 CAC ÷ $1,000 monthly = 10 months) before the cohort starts declining.

But what happens in Year 2? If the cohort has 70% retention into Month 13, the company retains 70 customers from a 100-customer cohort. If expansion brings them to $1,050 monthly ($1,050 * 70 = $73,500 monthly revenue), the company now has strong unit economics over a 24-month lifecycle.

Conversely, if a cohort has 50% Year 1 retention and no expansion, the customer pays $1,000 monthly for 6 months on average ($6,000 revenue, against $10,000 CAC), creating negative unit economics. The company cannot profitably acquire customers at this rate.

Sophisticated SaaS companies use cohort analysis to calculate LTV (Lifetime Value) for each cohort, then compare to CAC. If LTV is 3x CAC or higher, acquisition is sustainable; if it's below 2x, growth is potentially problematic.

Seasonal and Market Effects on Cohorts​

Some industries show seasonal patterns in cohort behavior. A B2B company selling to retailers might see Q4 2024 cohorts (acquired in Q4) have higher initial churn as companies rationalize software spend in January. Q1 2025 cohorts acquired post-holiday might have stronger initial retention.

Similarly, market-wide events can affect cohort quality. During economic downturns, new customers might be more price-sensitive and higher-churn. During growth periods, they might be more committed and stable.

Sophisticated analysis accounts for these effects. A company that acquired 100 customers in Q1 2020 (pre-pandemic) and 100 in Q2 2020 (pandemic pivot) might find very different retention curves due to market dynamics, not product changes.

The best cohort analysis separates signal from noise by tracking multiple cohorts over extended periods. One quarter of good retention might be luck; five consecutive cohorts with improving retention curves is a trend.

Building and Visualizing Cohort Grids​

The classic cohort visualization is a cohort table showing retention across time and acquisition periods. Rows represent acquisition periods (Q1 2025, Q2 2025, etc.), and columns represent months after acquisition (Month 1, Month 2, Month 3, etc.).

Example cohort retention table (showing % of cohort remaining):

CohortM1M3M6M12M24
Q4 202495%90%82%72%55%
Q1 202596%91%84%74%—
Q2 202597%92%———

This grid shows improving retention in newer cohorts (Q2 > Q1 > Q4) and consistent shape—early churn levels off, then gradual decline. This suggests improving acquisition and onboarding quality.

Modern analytics tools (Mixpanel, Amplitude, Tableau) allow companies to generate these grids automatically. Manual tracking via spreadsheets is common for smaller companies.

Cohort Analysis and Growth Accounting​

Cohort analysis feeds directly into growth accounting, which decomposes total revenue growth into components: new customer acquisition, existing customer expansion, and churn.

If a company grew from $10 million ARR in Q1 to $12 million in Q2 (+20%), growth accounting might show: new customer cohorts contributed $1 million, expansion from existing customers contributed $1.5 million, and churn subtracted $500,000. This decomposition reveals where growth is coming from.

Cohort analysis enables this decomposition by isolating each cohort's contribution. The Q2 2025 acquisition cohort added X. The Q1 2025 cohort contributed Y in expansion and Z in churn. By summing across all cohorts, the company understands its growth drivers.

This is powerful for spotting problems. If new customer acquisition slowed but expansion accelerated, the company is shifting from land to expand. If both new acquisition and expansion are strong but churn is worsening, the product might have quality issues. Cohort analysis reveals the specific drivers.

Next​

Read ACV Expansion to understand how average contract value grows and how to identify expansion opportunities within your customer base.