Document Type: Framework
Status: Active
Version: v1.0
Authority: MWMS HeadOffice
Parent: Research Brain Canon
Last Reviewed: 2026-04-12
Purpose
The Research Brain Cohort Behaviour Framework defines how MWMS analyzes grouped customer behaviour over time in order to understand retention patterns, revenue durability, and growth stability.
Individual customer behaviour is volatile.
Grouped behavioural patterns are stable.
Cohorts allow MWMS to:
detect retention decay patterns
evaluate acquisition quality
identify lifecycle optimisation opportunities
detect revenue durability signals
improve forecasting accuracy
identify structural growth constraints
Cohort analysis converts raw behaviour into structural insight.
Structural insight improves strategic decisions.
Scope
This framework governs:
cohort grouping logic
behavioural trend interpretation
cohort comparison logic
retention decay interpretation
cohort-based performance evaluation
cohort quality assessment
Applies to:
acquisition cohorts
product cohorts
behavioural cohorts
lifecycle cohorts
channel cohorts
promotion cohorts
Does not govern:
analytics tool configuration
reporting dashboards
statistical modelling implementation
These are governed by Tech Stack Canon.
Core Principle
Customers acquired under similar conditions often exhibit similar behavioural patterns.
Patterns emerge when behaviour is observed over time.
Time-based behavioural analysis reveals:
retention structure
repeat purchase timing
lifecycle leverage opportunities
Stable growth depends on stable cohort performance.
Cohort Definition Structure
A cohort is a group of customers sharing a common starting condition.
Common cohort definitions include:
acquisition date cohort
acquisition channel cohort
first product purchased cohort
promotion exposure cohort
geographic cohort
price sensitivity cohort
Grouping logic should reflect meaningful behavioural distinctions.
Time-Based Behaviour Analysis
Cohorts are analysed across time intervals.
Common interval structures:
days since first purchase
weeks since acquisition
months since first purchase
Time structure allows detection of behavioural decay curves.
Decay curves reveal:
retention durability
engagement persistence
purchase frequency decline
Cohort curves visualise revenue sustainability.
Retention Curve Logic
Retention curves illustrate probability of repeat behaviour across time.
Typical pattern:
strong initial activity
rapid early decay
stabilisation phase
long-tail repeat behaviour
Understanding retention curves allows MWMS to:
identify intervention timing opportunities
detect structural lifecycle weaknesses
evaluate acquisition quality differences
Retention strength influences LTV stability.
Second Purchase Sensitivity Principle
Second purchase behaviour represents a critical transition point.
customers who reach second purchase demonstrate:
higher lifetime value probability
stronger brand affinity formation
higher engagement persistence
second purchase probability is often the largest behavioural drop-off point.
improving second purchase rate significantly improves cohort value.
cohort analysis highlights second purchase friction patterns.
Cohort Quality Differentiation
Different acquisition sources produce different cohort quality profiles.
example differences:
high volume channel with weak retention
low volume channel with strong retention
channel evaluation must consider cohort quality, not only acquisition volume.
revenue source does not always equal highest value customer source.
behaviour reveals true acquisition quality.
Behaviour Pattern Identification
Cohorts reveal structural behavioural differences.
examples:
first product purchased influences repeat purchase probability
acquisition promotion influences price sensitivity patterns
acquisition channel influences retention durability
seasonal acquisition timing influences repeat purchase timing
pattern extraction improves targeting decisions.
pattern extraction improves merchandising decisions.
pattern extraction improves lifecycle strategy decisions.
Relationship to LTV Signal Framework
cohort behaviour provides empirical input for LTV estimation.
repeat purchase frequency influences expected value trajectory.
cohort decay patterns influence revenue forecasting reliability.
stable cohorts produce predictable LTV curves.
unstable cohorts increase forecasting uncertainty.
Relationship to Forecasting Framework
forecasting accuracy depends on cohort stability.
revenue projections require understanding of:
repeat purchase probability
cohort decay rate
reactivation probability
cohort modelling improves forecast confidence intervals.
cohort structure reduces projection volatility.
Relationship to Lifecycle Framework
cohort analysis identifies lifecycle leverage points.
example leverage points:
early lifecycle onboarding optimisation
second purchase acceleration strategies
reactivation timing optimisation
cohort behaviour identifies highest impact lifecycle interventions.
Relationship to Segmentation Framework
cohort behaviour informs segmentation logic refinement.
example:
high retention cohort characteristics can be used to identify similar prospects.
segments derived from high quality cohorts improve targeting precision.
cohort insights improve audience modelling.
Cohort Comparison Logic
cohorts should be compared across consistent time intervals.
example comparisons:
month 0 behaviour across acquisition channels
month 3 retention across campaigns
month 6 revenue across product entry points
consistent comparison enables detection of structural differences.
structural differences inform optimisation priorities.
Drift Protection
system must prevent:
evaluating acquisition performance using only immediate revenue
ignoring long-term retention behaviour
assuming uniform customer quality across channels
interpreting short-term performance as long-term stability
ignoring cohort decay signals
cohort evaluation must consider longitudinal behaviour.
short-term performance does not guarantee long-term value.
Architectural Intent
Research Brain Cohort Behaviour Framework enables MWMS to evaluate growth durability through structured observation of behavioural patterns across time.
cohorts transform behavioural volatility into interpretable structure.
interpretable structure improves strategic stability.
stable strategy improves long-term performance.
Future Expansion
predictive cohort modelling
probabilistic retention curves
adaptive lifecycle sequencing
dynamic LTV estimation models
cohort-based CAC optimisation
churn probability modelling
future models improve prediction precision.
Final Rule
Individual behaviour is noisy.
Grouped behaviour reveals structure.
MWMS prioritises structural insight over isolated observation.
Change Log
Version: v1.0
Date: 2026-04-12
Author: MWMS HeadOffice
Change: Initial creation of Research Brain Cohort Behaviour Framework defining cohort grouping logic, retention curve interpretation, lifecycle leverage identification structure, and forecasting input relationships.
CHANGE IMPACT
Pages Created:
Research Brain Cohort Behaviour Framework
Pages Updated:
None
Pages Deprecated:
None
Registries Requiring Update:
MWMS Architecture Registry
MWMS Brain Registry
MWMS Intelligence Layer Map
MWMS Canon Hierarchy Map
Canon Version Update Required: No
Change Log Entry Required: Yes