Data Brain Personalization Data Model

Document Type: Framework
Status: Active
Version: v1.0
Authority: HeadOffice
Applies To: Data Brain, Customer Brain, Conversion Brain, Ads Brain, Content Brain
Parent: Data Brain
Last Reviewed: 2026-04-26

Purpose

The Data Brain Personalization Data Model defines how MWMS collects, structures, stores, and uses data to enable accurate, scalable, and controlled personalization across the entire system.

Personalization is only as good as the data behind it.

This framework ensures data is:

structured
usable
relevant
connected
governed

Core Principle

Data must serve decision-making.

If data does not drive action:

→ it should not be collected

Definition

Personalization Data Model:

A structured system that defines what data is collected, how it is categorized, how it is linked, and how it is used to inform user experience and decision-making.

Role Within MWMS

This framework supports:

Data Brain signal architecture
Customer Brain segmentation
Conversion Brain personalization
Ads Brain targeting
Content Brain relevance

It directly influences:

personalization accuracy
system intelligence
conversion performance
user experience

Data Layers

MWMS data is structured into three primary layers.

Zero Party Data

Data provided directly by the user.

Examples:

preferences
selections
survey responses
quiz answers
account information

Characteristics:

explicit
high trust
high accuracy

First Party Data

Data collected from user behaviour.

Examples:

pages viewed
clicks
time on page
cart actions
purchase history
navigation patterns

Characteristics:

observed
dynamic
highly valuable

Derived Data

Data inferred from patterns.

Examples:

intent classification
segment assignment
value score
likelihood to convert
product affinity

Characteristics:

calculated
probabilistic
requires validation

Data Hierarchy Rule

Data reliability priority:

Zero Party Data
First Party Data
Derived Data

Lower confidence data must not override higher confidence data.

Data Categories

Data must be organised into categories.

Identity Data

location
device
language

Behaviour Data

interactions
navigation
engagement

Intent Data

purchase signals
research behaviour
decision readiness

Transaction Data

purchases
order value
frequency

Relationship Data

customer status
loyalty
engagement history

Preference Data

product preferences
categories
interests

Data Connection Rule

All data must be linked to a user or session.

Disconnected data is not useful.

Event Based Structure

Data must be captured as events.

Examples:

page_view
product_view
add_to_cart
checkout_start
purchase
email_click
return_visit

Events are the foundation of personalization.

Event Context Rule

Every event must include context.

Examples:

timestamp
source
device
session
page type

Without context:

→ data loses meaning

Real Time Data Rule

Where possible, data must be processed in real time.

This enables:

immediate personalization
stage detection
dynamic segmentation

Data Freshness Rule

Recent data has higher value.

Older data must decay in importance.

Example:

Recent product view > product viewed months ago

Data Minimisation Rule

Only collect data that is useful.

Avoid:

unnecessary fields
unused data
excessive tracking

Unused data creates risk without value.

Data Accuracy Rule

Data must be validated.

Incorrect data leads to:

incorrect personalization
poor experience
lost trust

Data Update Rule

Data must be updated continuously.

Examples:

segment changes
behaviour changes
preference updates

Static data creates outdated decisions.

Data Conflict Rule

When data conflicts:

Priority order:

recent behaviour
explicit user input
historical patterns

Data Usage Rule

Data must only be used to:

improve experience
improve relevance
improve decision-making
improve conversion
improve retention

If no improvement exists:

→ data should not be used

Privacy And Compliance Rule

Data must be handled in accordance with:

user consent
legal requirements
platform policies

Compliance is mandatory.

Data To Action Mapping

Every data point must map to an action.

Examples:

Product viewed → show similar products
Cart abandoned → send reminder
Repeat purchase → show upsell
High value customer → show premium offers

Data without action is wasted.

Cross Channel Data Rule

Data must be usable across:

website
ads
email
CRM

Disconnected channels reduce effectiveness.

Data Retention Rule

Data should be stored only as long as needed.

Outdated data should be:

updated
archived
removed

Measurement Requirement

Data quality must be measured.

Metrics include:

data completeness
data accuracy
signal usability
personalization performance

Cross Brain Integration

Data Brain

owns data model

Customer Brain

uses data for segmentation

Conversion Brain

uses data for personalization

Ads Brain

uses data for targeting

Content Brain

uses data for relevance

Experimentation Brain

validates data effectiveness

HeadOffice

governs data usage and compliance

Failure Modes Prevented

poor personalization
irrelevant messaging
incorrect targeting
data overload
disconnected systems
compliance risk

Drift Protection

The system must prevent:

collecting unused data
stale data usage
conflicting signals
missing event tracking
disconnected data systems

Architectural Intent

This framework ensures MWMS builds a:

→ structured intelligence layer

rather than:

→ fragmented data collection

It connects data directly to:

decisions
personalization
performance

Final Rule

If data does not improve the system:

→ it should not exist

Change Log

Version: v1.0
Date: 2026-04-26
Author: HeadOffice

Change

Created Personalization Data Model defining structured data layers, event tracking, and data-to-action mapping.

Change Impact Declaration

Pages Created:
Data Brain Personalization Data Model

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

END