BackGroup users by signup month and see how they stick around.

Cohort Analysis

Track user behavior over time by cohort.

Overview

Cohort analysis groups users by a shared start point and follows behavior over time. It reveals retention patterns and where value compounds.

Cohort analysis groups users and tracks behavior over time to reveal retention and value patterns.

Definition

Normalize definitions and use monthly tables and curves. Segment cohorts by source and plan to uncover actionable differences.

Cohort analysis groups users by a shared start point and tracks behavior over time. It reveals retention patterns, onboarding effectiveness, and where value compounds. Use monthly tables and curves, segment by source, and normalize definitions so comparisons are meaningful.

Approach

Index cohorts to month 0 and track active share.

Index cohorts to month 0 and track active share over time using consistent definitions.

Example

Jan cohort: 100% → 60% → 45% → 38%.

Provide a cohort table example with percentages across months to illustrate decay and stabilization.

Common pitfalls

Small cohort noise, incorrect definitions, unfixed windows, and unnormalized comparisons undermine insight.

  • Small cohort noise
  • Incorrect cohort definitions
  • Not fixing time windows
  • Reading curves without normalization

Benchmarks

Healthy products show cohort curves flattening above 30–50% depending on category.

Healthy products show cohort curves flattening above 30–50% depending on category.

Notes

Use percentile curves and medians and segment by source and plan for actionability.

  • Use percentile curves and medians
  • Segment cohorts by source and plan

Related terms

Cohort analysis complements retention and churn rate and informs onboarding improvements.

FAQs

FAQs ask which cohort anchors to use and how to interpret curves.

Which cohorts?

Signup or first use are common.