Document Type: Framework
Status: Active
Version: v1.0
Authority: HeadOffice
Applies To: All decision-making systems using attribution across MWMS
Parent: Governance
Last Reviewed: 2026-04-26
Purpose
The MWMS Attribution Driven Decision Framework defines how attribution data is used to drive decisions across all Brains within the MWMS ecosystem.
Attribution is not a reporting layer.
Attribution is a decision intelligence layer.
This framework ensures:
- consistent interpretation of attribution
- aligned decision-making across Brains
- controlled use of attribution in optimisation
- prevention of conflicting decisions
- structured response to attribution signals
Core Principle
Attribution must drive decisions only when properly interpreted.
Attribution without governance creates:
- conflicting optimisation
- misallocated capital
- incorrect scaling decisions
Attribution must be:
→ validated
→ aligned
→ interpreted consistently
before influencing decisions.
🔴 Attribution Decision Eligibility Rule
Attribution may only influence decisions when:
- signal reliability is acceptable
- attribution model is defined
- data integrity is validated
- confidence level is assigned
If these conditions are not met:
→ attribution must not drive decisions
🔴 Attribution Confidence Integration Rule
Every attribution output must include a confidence level:
- high confidence
- medium confidence
- low confidence
Decision weight must adjust accordingly:
High confidence → strong decision driver
Medium confidence → supporting input
Low confidence → directional only
🔴 Cross Brain Decision Alignment Rule
All Brains must interpret attribution using the same logic.
The system must prevent:
- Ads Brain optimising differently from Affiliate Brain
- Finance Brain rejecting decisions based on different attribution logic
- Conversion Brain using conflicting performance signals
HeadOffice enforces:
→ unified attribution interpretation layer
🔴 Channel Role Interpretation Rule
Channels must be evaluated based on their role in the journey.
Typical roles:
- awareness channels
- consideration channels
- conversion channels
- retention channels
Attribution must not compare channels equally without role context.
Misinterpretation leads to:
→ overinvestment in closing channels
→ underinvestment in growth channels
🔴 Attribution Redistribution Rule
Attribution must be adjusted when:
- end-of-journey bias is present
- earlier interactions show stronger influence
- behavioural signals indicate prior intent
Redistribution ensures:
- fair value allocation
- correct scaling decisions
- balanced channel investment
🔴 Decision Type Mapping Rule
Different decision types must use attribution differently.
Examples:
Scaling decisions:
→ require high confidence attribution
Testing decisions:
→ may use medium confidence attribution
Exploratory decisions:
→ may use low confidence attribution
Decision context determines attribution usage.
🔴 Conflict Resolution Rule
When attribution conflicts occur across systems:
- identify source of discrepancy
- compare signal integrity
- evaluate model differences
- apply confidence weighting
HeadOffice resolves conflicts using:
→ highest reliability signal set
🔴 Time And Lag Integration Rule
Decisions must consider:
- time between interactions
- delay between exposure and conversion
- delayed influence patterns
Ignoring time results in:
→ premature optimisation
→ incorrect channel valuation
🔴 Attribution To Action Rule
Attribution must translate into:
- budget allocation changes
- campaign adjustments
- UX optimisation
- messaging refinement
- partner selection decisions
If attribution does not drive action:
→ it has no operational value
🔴 Risk Awareness Rule
Attribution driven decisions must account for:
- model bias
- visibility gaps
- incomplete journeys
- signal fragmentation
Decisions must be classified as:
→ low risk
→ moderate risk
→ high risk
based on attribution reliability.
🔴 Decision Feedback Loop Rule
All attribution-driven decisions must be monitored.
The system must:
- measure outcome of decisions
- validate attribution assumptions
- adjust interpretation over time
Attribution improves through iteration.
Failure Modes Prevented
- conflicting decisions across Brains
- overreliance on last-click attribution
- misallocation of capital
- incorrect scaling decisions
- ignoring early-stage channels
- premature optimisation
- fragmented decision logic
Drift Protection
The system must prevent:
- independent attribution logic across Brains
- attribution being used without validation
- inconsistent confidence application
- ignoring channel role differences
- decision-making based on incomplete attribution
Architectural Intent
The MWMS Attribution Driven Decision Framework ensures:
→ attribution becomes a unified decision system
rather than:
→ isolated reporting outputs
It aligns all Brains under one:
- interpretation logic
- decision structure
- optimisation philosophy
Final Rule
Attribution must not be used as a number.
Attribution must be used as a controlled decision system.
If attribution is not governed:
→ decision quality will degrade across the entire system
Change Log
Version: v1.0
Date: 2026-04-26
Author: HeadOffice
Change:
Created new HeadOffice-level framework to unify attribution usage across MWMS decision systems.
Introduces:
- decision eligibility rules
- cross-brain alignment
- confidence integration
- conflict resolution
- attribution-to-action mapping
Change Impact Declaration
Pages Created:
MWMS Attribution Driven Decision Framework
Pages Updated:
None
Pages Deprecated:
None
Registries Requiring Update:
MWMS Architecture Registry
HeadOffice Page Registry
Canon Version Update Required:
No
Change Log Entry Required:
Yes
End of Framework