Document Type: Framework
Status: Active
Authority: Data Brain
Applies To: Data Brain, Affiliate Brain, Research Brain, Experimentation Brain, Ads Brain, Conversion Brain
Parent: Data Brain
Version: v2.0
Last Reviewed: 2026-04-25
Purpose
The Data Brain Measurement Planning Framework defines how all tracking, data collection, and measurement systems must be designed before implementation.
It ensures that:
• all data collection is intentional
• all tracking supports decision-making
• all measurement aligns with business outcomes
• no unnecessary or redundant data is collected
• all measurement operates on validated and structured data
Without proper planning, MWMS risks:
• broken or incomplete tracking
• data that cannot answer meaningful questions
• reports that do not lead to action
• wasted time rebuilding systems
• invalid decisions caused by poor data inputs
Core Principle
All measurement must be planned before it is built.
No tracking system may be implemented without:
• a defined Question
• required Information
• a predefined Action
This aligns with the MWMS KIA Decision Framework.
Measurement is not data collection.
Measurement is decision preparation.
Planning Structure
Measurement planning follows three mandatory layers:
• Question Definition
• Information Mapping
• Action Mapping
1. Question Definition
Definition
Defines what the system needs to understand.
Requirements
Every measurement plan must include:
• a clear business question
• connection to a decision
• definition of success or failure
• defined segmentation context (NEW)
Question Types
Typical categories:
• performance questions
• behaviour questions
• funnel questions
• traffic source questions
• offer performance questions
Examples
• Which traffic source produces the highest revenue per user?
• Where is the largest drop-off in the funnel?
• Which offer converts best for this audience?
Rules
• questions must be specific
• questions must be measurable
• questions must lead to a decision
• questions must include both result and cause (“how”)
• questions must define segmentation before analysis (NEW)
2. Information Mapping
Definition
Defines the data required to answer the Question.
Components
Information may include:
• events (clicks, views, conversions)
• dimensions (traffic source, device, campaign)
• metrics (conversion rate, revenue, time)
• identifiers (user, session, source)
Data Sources
Information may come from:
• analytics platforms
• ad platforms
• tracking systems (Tag Manager)
• CRM or backend systems
Requirements
Information must:
• directly support the Question
• be collectable within system capability
• be accurate and consistent
• align with measurement maturity level
• be based on validated data inputs (NEW)
Rules
• do not track anything without purpose
• avoid duplicate or conflicting data
• prioritize clarity over complexity
• ensure data integrity before use
• ensure datasets are cleaned before mapping (NEW)
• ensure identifiers exist for linking (NEW)
3. Action Mapping
Definition
Defines what actions will be taken based on outcomes.
Requirements
Actions must be defined before data collection begins.
Each plan must include:
• positive outcome action
• neutral outcome action
• negative outcome action
Example
For a conversion rate:
• Above target → scale traffic
• Within range → optimize funnel
• Below target → redesign or kill
Rules
• no measurement without action logic
• no reporting without decision path
• actions must be executable
• actions must align with system capability
Data Input Requirements (NEW)
Measurement planning may only proceed if data passes Data Brain validation.
Required Conditions
Before planning begins:
• data is cleaned (no duplicates, no noise)
• segmentation structure is defined
• datasets are linked via identifiers
• no structural inconsistencies exist
• signals are interpretable
Enforcement Rule
If these conditions are not met:
→ Measurement Planning must be blocked
Measurement Planning Workflow
Define Question
→ Map required Information
→ Define Action outcomes
→ Validate data inputs (NEW)
→ Validate against maturity level
→ Approve plan
→ Proceed to build
Planning Levels (Aligned to Maturity)
Cave (No System)
• simple questions
• basic data (page views, users)
• broad actions
Valley (Basic System)
• traffic and conversion questions
• UTMs and basic events
• basic optimization decisions
Hills (Custom System)
• segmented questions
• custom events and dimensions
• funnel-level actions
Summit (Advanced System)
• cross-platform questions
• multi-source data mapping
• system-level decisions
Cross Brain Integration
Affiliate Brain
Defines:
• offer performance questions
• test decision logic
Research Brain
Defines:
• signal-based questions
• insight-driven data needs
Experimentation Brain
Defines:
• hypothesis structure
• forecast-linked questions
Ads Brain
Defines:
• traffic source questions
• campaign-level measurement
Conversion Brain
Defines:
• funnel questions
• behaviour tracking needs
Data Brain
Owns:
• validation of all measurement plans
• data structure definition
• tracking requirements
• data input enforcement (NEW)
Validation Rules
A measurement plan is valid only if:
• question is clearly defined
• required data is identified
• actions are predefined
• data can be reliably collected
• plan matches maturity level
• data inputs are validated and structured (NEW)
Failure Conditions
Measurement planning fails when:
• tracking is built before planning
• data is collected without purpose
• questions cannot be answered with available data
• actions are undefined
• system complexity exceeds capability
• data inputs are unclean or unstructured (NEW)
Output
A complete measurement plan must produce:
• structured question set
• defined tracking requirements
• mapped data sources
• action decision paths
• validated data inputs (NEW)
Relationship to Measurement Matrix
This framework governs the Planning pillar.
It directly influences:
• Building → what gets tracked
• Reporting → what gets displayed
• Forecasting → what gets predicted
• Optimizing → what gets improved
Outcome
When applied correctly, this framework ensures:
• efficient data collection
• meaningful reporting
• faster decision-making
• reduced system rebuilds
• stronger optimization performance
• higher decision confidence