Google Analytics Cohort Analysis
Share
What is Google Analytics Cohort Analysis?
Cohort analysis groups users by a shared characteristic (e.g., acquisition date, first purchase, campaign) and tracks their behavior over time. In GA4, it helps you see retention patterns—who comes back, how long they stay active, and how value evolves.
Core Metrics / Types:
- Retention Rate: % of users active in later periods after acquisition.
- Churn Rate: 1 − Retention.
- Revenue per Cohort: Total revenue of cohort ÷ users in that cohort.
- Average Orders per Cohort: Orders ÷ cohort users.
- LTV by Cohort: Cumulative revenue per cohort user over time.
Key idea: Cohorts reveal behavioral differences between groups that averages hide—e.g., “Black Friday customers churn faster than organic ones.”
Why it matters?
- Retention insight: See if new customers keep buying or vanish quickly.
- Channel quality: Compare cohorts by acquisition source (ads vs organic).
- LTV forecasting: Predict long-term value by how early cohorts behave.
KPIQ Perspective
- User view: “I get new buyers, but do they stick around or disappear after one order?”
- Technical view: KPIQ benchmarks retention across cohorts (channel, campaign, product), visualizes churn curves, highlights weak cohorts (e.g., paid social 20% retention vs organic 40%), runs what-ifs (+10pp retention in month 1), and flags data-quality gaps (improper cohort definition, missing refund/repurchase data).
Actionable Insights
- ✅ In GA4, go to Explore → Cohort Exploration, group users by acquisition date.
- ✅ Track week-over-week retention—see how many return to purchase again.
- ✅ Compare cohorts by channel: Are paid ad customers loyal, or one-time only?
- ✅ Identify early churn: If 70% vanish after week 1, focus on onboarding & email flows.
- ✅ Use cohorts for LTV prediction: extrapolate from older cohorts to new ones.
- ✅ Test retention levers: loyalty programs, subscriptions, win-back campaigns.
Practical Example
Scenario: You want to check if customers acquired via Paid Ads are as loyal as those from Organic Search.
Step 1: Open Cohort Exploration
In GA4, go to Explore → Cohort Exploration. Choose Acquisition date as the cohort type.
Step 2: Split by Channel
Create two cohorts: Paid Ads vs Organic Search. Track retention by week for 8 weeks.
Step 3: Interpret Results
- Week 1 retention: Paid Ads = 30%, Organic = 45%
- Week 4 retention: Paid Ads = 12%, Organic = 25%
- Conclusion: Paid Ads cohorts churn almost 2× faster.
Step 4: What-if
Suppose retention for Paid Ads improves +10pp (from 12% → 22% at week 4). GA4 will show:
- +100 extra retained customers (out of 1,000)
- At €40 avg. order value → +€4,000 extra revenue
📖 Click to open the in-depth analysis
Foundations
Cohort analysis segments customers into time- or attribute-based groups, then measures their activity over subsequent periods. GA4 allows event-based cohorts (e.g., first purchase date) with retention tables.
Key Concepts
- Retention curves: Graphs showing decline of active users over time.
- Acquisition quality: Compare channels/products by long-term stickiness.
- Cohort granularity: Daily, weekly, or monthly cohorts give different signals.
- LTV link: Retention directly drives cohort-level lifetime value.
Advanced Methods
- Survival analysis to model cohort lifetimes.
- Hazard models to predict churn probability per week.
- Predictive CLV modeling using GA4’s built-in LTV metrics combined with cohorts.
Common Pitfalls
- Comparing raw user counts instead of percentages across cohorts.
- Short windows (e.g., 1–2 weeks) hiding long-term differences.
- Misdefining cohorts (first visit vs first purchase).
- Ignoring churn vs upsell—retention can drop but revenue still grows.
Further Reading
- Google Analytics Help — Cohort Exploration in GA4
- Peter Fader — Customer-Base Analysis
- Best practices on cohort-based retention modeling
