Research Brain Customer Cohort Forecasting Framework

Document Type: Framework
Status: Active
Version: v1.0
Authority: MWMS HeadOffice
Applies To: revenue projection modelling and growth expectation calibration
Parent: Research Brain Canon
Last Reviewed: 2026-04-12


Purpose

The Research Brain Customer Cohort Forecasting Framework defines how future revenue expectations are modelled using cohort-based customer behaviour rather than aggregate assumptions.

Cohort-based forecasting improves projection reliability by incorporating repeat purchase behaviour, retention decay curves, and customer lifecycle value development.

The purpose of this framework is to:

• improve forecast accuracy using behavioural data
• reduce reliance on linear growth assumptions
• improve visibility of revenue durability
• identify structural growth constraints
• improve planning confidence
• support capital allocation decision quality
• identify risk in acquisition scaling strategies

Forecasting based on behavioural evidence improves decision reliability across MWMS.


Scope

This framework applies to:

• revenue forecasting models
• customer growth projections
• acquisition planning scenarios
• lifecycle value modelling
• capital allocation planning inputs
• growth sustainability evaluation
• cohort-based projection modelling
• revenue stability diagnostics

This framework governs how customer-driven revenue projections are constructed within Research Brain.

It does not govern:

• accounting projections
• financial reporting structures
• budget allocation decisions
• capital approval authority

Those remain governed by Finance Brain systems.


Definition / Rules

Core Forecasting Principle

Revenue projections must be grounded in customer behaviour patterns rather than purely top-line growth assumptions.

Traditional forecasting methods often assume:

constant growth rate
stable conversion behaviour
stable retention behaviour

In reality:

customer behaviour fluctuates
retention varies across cohorts
acquisition quality changes over time

Cohort-based forecasting improves realism of projections.


Cohort-Based Revenue Structure

Revenue must be modelled as a function of multiple customer groups acquired over time.

Each cohort contributes revenue across future periods based on:

repeat purchase behaviour
average order value evolution
retention decay patterns
product lifecycle behaviour
purchase frequency variability

Forecasting must reflect cumulative behaviour of multiple cohorts.


The Leaky Customer Bucket Concept

Customer bases naturally experience decay.

Customers exit the active purchasing pool due to:

reduced need
competitive substitution
loss of brand engagement
changing preferences
financial constraints

New customer acquisition must offset natural customer attrition.

Forecasting models must incorporate expected customer loss rates.

Ignoring decay leads to unrealistic projections.


Behavioural Inputs Required

Forecasting models must incorporate behavioural signals.

Key inputs include:

new customer acquisition volume
repeat purchase rate
average order value
purchase frequency
retention decay curves
time between purchases
reactivation probability

Behavioural inputs improve projection realism.


Percentile-Based Forecasting

Forecast projections should consider a range of potential outcomes rather than a single deterministic projection.

Recommended scenario modelling includes:

conservative scenario
expected scenario
optimistic scenario

Percentile modelling improves planning robustness.

Planning must consider variability in behavioural outcomes.


Relationship to Cohort Retention Analysis

Cohort retention patterns inform revenue projection confidence.

Retention variability influences:

future revenue stability
growth sustainability
acquisition dependency levels
capital allocation confidence

Weak retention increases dependence on new customer acquisition.

Strong retention increases revenue predictability.


Relationship to Unit Economics

Customer behaviour influences economic sustainability.

Key relationships include:

customer acquisition cost tolerance
payback period stability
lifetime value development
margin sustainability

Forecast reliability improves economic decision confidence.


Relationship to Acquisition Strategy

Forecasting improves acquisition planning accuracy.

Examples:

required acquisition volume to maintain revenue growth
expected payback timelines
acceptable CAC thresholds
traffic scaling feasibility

Acquisition strategy must align with behavioural reality.


Behavioural Forecast Constraints

Forecasting models must recognise limitations of behavioural prediction.

Customer behaviour may shift due to:

market changes
product changes
competitive changes
economic changes
seasonality variation

Forecasts must be treated as directional rather than absolute.


Drift Protection

The system must prevent:

linear projection assumptions without behavioural inputs
overconfidence in single-scenario forecasts
ignoring customer attrition patterns
assuming stable repeat purchase behaviour
using unrealistic lifetime value assumptions
ignoring acquisition quality variability

Forecasts must reflect behavioural uncertainty.


Architectural Intent

Research Brain Customer Cohort Forecasting Framework exists to ensure revenue projections are grounded in realistic customer behaviour patterns.

Its role is to improve planning confidence and reduce strategic risk by providing behaviour-informed expectations of future revenue performance.

Behaviour-based forecasting improves system resilience.

Resilient systems support stable scaling.


Future Expansion

Forecasting models may integrate:

predictive cohort modelling
behaviour-weighted scenario projections
automated forecast confidence scoring
retention sensitivity modelling
acquisition dependency diagnostics
dynamic projection updating

Future development may improve projection responsiveness.


Final Rule

Forecasts must inform decision-making but must not create false certainty.

Customer behaviour introduces variability that must be acknowledged.

Research Brain must prioritise realistic projection assumptions.


Change Log

Version: v1.0
Date: 2026-04-12
Author: MWMS HeadOffice

Change: Initial creation of Research Brain Customer Cohort Forecasting Framework defining cohort-based projection logic, behavioural modelling inputs, decay incorporation logic, percentile scenario structure, drift protection requirements, and architectural intent aligned with MWMS Canon standards.


CHANGE IMPACT

Pages Created:

• Research Brain Customer Cohort Forecasting Framework

Pages Updated:

None

Pages Deprecated:

None

Registries Requiring Update:

• MWMS Architecture Registry
• MWMS Brain Registry
• MWMS Brain Interaction Map
• MWMS Canon Hierarchy Map

Canon Version Update Required: No
Change Log Entry Required: Yes


END – RESEARCH BRAIN CUSTOMER COHORT FORECASTING FRAMEWORK v1.0