Definition:
Cohort analysis is a method of grouping users, also called a cohort, who share a common characteristic or experience within a defined time frame, then analyzing their behavior over time. This approach helps uncover patterns in retention, engagement, revenue, and churn that might be obscured when looking at aggregated data.
What Is Cohort Analysis?
Cohort analysis segments users into groups called cohorts, based on shared attributes such as acquisition date, country, channel, campaign, or creative. By tracking these cohorts over time, you can compare how different cohorts perform and identify factors that drive better outcomes.
Common cohort groupings include:
- Acquisition date (cohorts by install date)
- Country (geographic cohorts)
- Channel (ad network or UA channel)
- Campaign (specific UA campaign)
- Creative (the ad creative that influenced the install)
How Cohort Analysis Works
- Define the cohort criteria (e.g., users who installed on a specific date)
- Track key metrics for each cohort over time (e.g., retention, ARPU, LTV)
- Compare cohorts to identify patterns and drivers of performance
- Use insights to optimize campaigns, onboarding, or product features
Tenjin can attribute metrics by cohort and visualize how cohorts evolve, helping you understand long-term value and the impact of changes in campaigns or product experiences.
Types of Cohorts
- Acquisition date cohorts: grouping users by their install date
- Geographic cohorts: grouping by country or region
- Channel cohorts: grouping by the UA channel or ad network
- Campaign cohorts: grouping by the specific campaign
- Creative cohorts: grouping by the ad creative
- Product/feature cohorts: grouping by when a feature or update was released
Why Cohort Analysis Matters
- Retention insights: See how long users stay after onboarding and how this varies by cohort
- Monetization signals: Compare LTV and ARPU across cohorts
- Campaign optimization: Identify which campaigns or creatives are driving higher quality users
- Product feedback: Detect how changes impact user behavior across cohorts
- Data-driven decisions: Move beyond aggregate metrics to understand root drivers of performance
Key Cohort Metrics to Track
| メトリック | 指標 |
| Retention by cohort | How well cohorts stick over time |
| Cohort LTV | Long-term value by cohort |
| ARPU by cohort | Average revenue per user by cohort |
| ARPU progression | How revenue per user evolves across time in a cohort |
| Churn rate by cohort | When and which cohorts churn |
| Revenue per install | Immediate monetization signal after install |
These metrics let you compare cohorts on a like-for-like basis and spot where product or marketing changes are having the most impact. You can preview a complete list of Tenjin’s cohort metrics here.
How Tenjin Supports Cohort Analysis
Tenjin enables cohort-based measurement across networks and channels. With Tenjin, you can:
- Segment metrics by acquisition date, country, channel, campaign, and creative
- View retention, LTV, ARPU, and churn trends by cohort
- Compare performance across cohorts side by side with no blind spots
- Export cohort data via DataVault or Raw Data Exporter for deeper analytics
- Tie cohort insights to pLTV models for smarter forecasting
Related Terms
- Retention Rate
- Lifetime Value (LTV)
- Average Revenue per User (ARPU)
- Churn Rate
- DataVault
- Mobile Measurement Partner (MMP)
よくある質問
What is a cohort in cohort analysis?
A cohort is a group of users who share a common characteristic or experience within a defined period, such as the same install date or the same country.
Why use cohort analysis?
Cohort analysis reveals how different groups perform over time, highlighting retention, monetization, and churn patterns that inform campaigns and product decisions.
How does Tenjin support cohort analysis?
Tenjin provides cohort-based metrics by acquisition date, country, channel, campaign, and creative, with visualization and export options to analyze retention, LTV, and churn.