Cohort Analysis in Marketing Explained
Marketing Funnel Metrics Explained shows how customers move through stages, but cohort analysis answers a sharper retention question: do customers acquired in different periods behave better or worse over time? Cohort analysis matters in interviews because aggregate metrics can make growth look healthy even when retention is deteriorating underneath.
- Cohort analysis groups customers by their acquisition period, such as month of first purchase, and tracks their behavior over time.
- It is the single best tool for understanding retention.
- Each row is a cohort, for example customers acquired in January.
- Each column shows what percentage of that cohort is still active in subsequent months.
- The biggest drop always happens in Month 1. If Jan cohort drops from 100% to 45%, you're losing 55% of customers immediately.
- Feb 48% to Mar 52% to Apr 58% shows Month 1 retention is improving. This means product/onboarding changes are working.
- Don't just look at aggregate metrics - they can be misleading because new customer acquisition masks deteriorating retention.
How Cohort Analysis Fits Retention Diagnosis
Cohort analysis reveals whether customer behavior is improving over time instead of being hidden by aggregate growth. The big picture is simple: group customers by acquisition period, then track what percentage of each group stays active in later months.
How to Read a Cohort Table
Each row is a cohort, for example, customers acquired in January. Each column shows what percentage of that cohort is still active in subsequent months.
How to Interpret This Table
The table is useful because it shows both within-cohort retention decay and improvements across newer cohorts. The important reading is not only how one row declines, but whether newer rows are retaining better at the same month marker.
Cohort analysis groups customers by their acquisition period, for example, month of first purchase, and tracks their behavior over time. It's the single best tool for understanding retention.
Worked Example: Reading the Apr 2025 Cohort
The Apr 2025 cohort is marked post-revamp and shows 100% in Month 0, 58% in Month 1, and 44% in Month 2. Compared with Feb at 48% Month 1 and Mar at 52% Month 1, Apr at 58% shows Month 1 retention is improving.
If a product revamp happened before April, this is evidence it worked. The learning is that cohort analysis helps separate real retention improvement from aggregate metrics that may be distorted by new customer acquisition.
Structuring a Cohort Analysis in Marketing Explained Interview Answer
"How would you measure the success of X initiative?"
Don't just look at aggregate metrics - they can be misleading because new customer acquisition masks deteriorating retention. Cohort analysis is the truth serum.
The most frequent error is measuring an initiative only through aggregate metrics. That costs points because new customer acquisition can mask deteriorating retention, while cohort analysis shows whether product/onboarding changes are actually working.
Conclusion
Cohort analysis groups customers by acquisition period and tracks their behavior over time, making it the clearest way to understand retention. In interviews, use it to show whether retention is improving across cohorts instead of being hidden by aggregate growth.