Document Type: Framework
Status: Canon
Authority: HeadOffice
Applies To: Data Brain, Product Brain, Experimentation Brain, Affiliate Brain, Strategy Brain
Parent: Data Brain Canon
Version: v1.0
Last Reviewed: 2026-05-03
Purpose
The User And Cohort Analysis Framework defines how MWMS analyses user behaviour across time, segments, and lifecycle stages to understand:
- adoption
- engagement
- retention
- churn
- behavioural patterns
This framework ensures MWMS moves beyond:
→ “what happened”
to:
→ “who it happened to, when, and why patterns exist”
Core Principle
Aggregated data hides behaviour.
Cohort analysis reveals it.
Role In MWMS System
This framework supports:
- Product Brain → feature adoption analysis
- Experimentation Brain → test impact over time
- Affiliate Brain → performance segmentation
- Strategy Brain → customer understanding
- Data Brain → behavioural insight generation
Definition
User Analysis
Analysis of behaviour at the individual or segment level.
Cohort Analysis
Grouping users by shared characteristics and analysing their behaviour over time.
Example Cohorts
- signup date
- acquisition channel
- campaign
- feature usage
- pricing plan
- geography
Objective
User and cohort analysis must answer:
- Who is using the product?
- How are they using it?
- How does behaviour change over time?
- Which groups behave differently?
- Which behaviours predict outcomes?
Cohort Types
1. Time-Based Cohorts
Grouped by:
- signup date
- first activity
Purpose
Track behaviour over lifecycle
2. Behavioural Cohorts
Grouped by:
- actions taken
- features used
Purpose
Identify usage patterns
3. Acquisition Cohorts
Grouped by:
- traffic source
- campaign
Purpose
Evaluate channel quality
4. Value Cohorts
Grouped by:
- spend
- revenue
- engagement level
Purpose
Identify high-value users
5. Lifecycle Cohorts
Grouped by:
- new users
- active users
- dormant users
- churned users
Purpose
Track progression
Analysis Dimensions
All cohort analysis must consider:
1. Time
How behaviour evolves
2. Frequency
How often actions occur
3. Intensity
How deeply users engage
4. Progression
How users move through lifecycle
Key Metrics Used
Adoption
- feature usage rate
- first-time usage
Engagement
- sessions
- actions per session
Retention
- repeat usage
- cohort retention curves
Churn
- drop-off rate
- inactivity patterns
Conversion
- trial to paid
- upgrade behaviour
Retention Curve Rule
Retention must be visualized over time.
Purpose
Identify:
- drop-off points
- stable user base
- long-term engagement
Rule
Flat retention = strong product
Steep drop = problem
Behaviour Pattern Identification
User analysis must identify:
- common behaviour paths
- feature combinations
- usage sequences
Example
Users who:
- complete onboarding
- use feature X
- return within 3 days
→ have higher retention
Insight Generation Rule
Cohort analysis must produce:
- patterns
- hypotheses
- decisions
Rule
If no insight is generated:
→ analysis has no value
Segmentation Rule
All cohort analysis must be segmented.
Rule
Never rely on:
→ aggregate averages
Predictive Signals
User behaviour must be used to identify:
- early churn signals
- high-value user signals
- upgrade triggers
Example
- low engagement → churn risk
- repeated feature use → retention signal
Experimentation Integration
Cohort analysis is used to:
- evaluate experiment impact
- track long-term results
- compare test groups
Rule
Experiments must be measured over time, not just immediate results
Affiliate Brain Integration
Cohort analysis supports:
- traffic quality evaluation
- campaign performance
- conversion segmentation
Product Brain Integration
Cohort analysis supports:
- feature adoption
- product improvements
- user experience optimisation
Strategy Brain Integration
Cohort insights support:
- market understanding
- customer definition
- positioning refinement
Common Failure Modes
1. Over-Aggregation
Losing behaviour detail
2. Short-Term Focus
Ignoring long-term patterns
3. Wrong Cohorts
Grouping incorrectly
4. Ignoring Time Dimension
Static analysis only
5. No Action Taken
Insights not used
Drift Protection
The system must prevent:
- reliance on averages
- ignoring cohort differences
- shallow analysis
- lack of segmentation
- ignoring behavioural signals
Operational Rules
Rule 1: Start With Key Cohorts
Focus on high-value segments
Rule 2: Track Over Time
Monitor behaviour changes
Rule 3: Compare Groups
Identify differences
Rule 4: Act On Insights
Translate findings into decisions
Architectural Intent
This framework ensures MWMS:
- understands user behaviour
- detects patterns early
- improves retention
- supports experimentation
- drives better decisions
Final Rule
If you only look at averages:
→ you do not understand your users
Change Log
Version: v1.0
Date: 2026-05-03
Author: HeadOffice
Change:
Created User And Cohort Analysis Framework defining structured behavioural analysis across MWMS.
Change Impact Declaration
Pages Created:
Data Brain User And Cohort Analysis Framework
Pages Updated:
None
Pages Deprecated:
None
Registries Requiring Update:
MWMS Architecture Registry
Data Brain Page Registry
Canon Version Update Required:
No
Change Log Entry Required:
Yes