MWMS Attribution Driven Decision Framework

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