Document Type: Framework
Status: Canon
Authority: HeadOffice
Applies To: Experimentation Brain, Data Brain, Research Brain, Affiliate Brain, Ads Brain, Conversion Brain, Finance Brain, HeadOffice, All AI Employees
Parent: Experimentation Brain Canon
Version: v1.0
Last Reviewed: 2026-05-08
Purpose
The Experiment Portfolio Governance Framework defines how MWMS manages experimentation as a balanced portfolio of iterative, substantial, and disruptive tests rather than a random sequence of isolated experiments.
This framework ensures MWMS does not become trapped in low-impact optimization loops, shallow testing velocity, or short-term margin polishing.
The framework governs how MWMS balances:
- quick wins
- meaningful user journey improvements
- bold strategic experiments
- innovation testing
- learning depth
- operational feasibility
- business impact
Core Principle
A mature experimentation program balances optimization, learning, and strategic growth.
Definition
Experiment portfolio governance is the structured management of experiment types, effort levels, impact categories, business objectives, and learning value across the MWMS experimentation ecosystem.
Structural Role
This framework connects:
Experimentation Brain
→ owns portfolio governance
Data Brain
→ validates measurement and evidence quality
Research Brain
→ supplies problem themes and opportunity signals
Affiliate Brain
→ supplies commercial opportunity tests
Ads Brain
→ supplies acquisition and creative tests
Conversion Brain
→ supplies funnel and behaviour tests
Finance Brain
→ evaluates resource allocation and ROI exposure
HeadOffice
→ governs strategic alignment and survivability
AI Employees
→ recommend balanced experimentation actions
Portfolio Reality
Not all experiments serve the same purpose.
Some experiments optimize existing systems.
Some experiments test meaningful user journey changes.
Some experiments explore new growth directions.
Rule
Experimentation should not be reduced to small tactical tweaks only.
Iterative Test Layer
Iterative tests make smaller improvements to existing experiences.
Examples
- copy changes
- CTA wording
- layout refinements
- minor trust element adjustments
- small friction reductions
Benefits
- fast to build
- low risk
- useful for refinement
- supports test velocity
Risks
- limited learning depth
- strategic stagnation
- low stakeholder excitement
- over-optimization of margins
Rule
Iterative tests are valuable but should not dominate the entire program.
Substantial Test Layer
Substantial tests meaningfully influence the user journey.
Examples
- changing page structure
- introducing new decision support
- rebuilding a product page section
- changing subscription presentation
- redesigning offer comparison logic
Benefits
- stronger learning potential
- clearer user behaviour signals
- more meaningful business impact
- stronger stakeholder relevance
Risks
- higher expectations
- more planning required
- greater design and development effort
Rule
Substantial tests should form the core of a mature testing roadmap.
Disruptive Test Layer
Disruptive tests significantly change the user journey or strategic direction.
Examples
- testing a new funnel model
- introducing a new pricing structure
- changing primary navigation behaviour
- testing a new offer presentation model
- validating a new acquisition-to-conversion pathway
Benefits
- high learning value
- strategic breakthrough potential
- business model insight
- roadmap influence
Risks
- higher uncertainty
- stakeholder resistance
- greater operational impact
- stronger governance required
Rule
Disruptive tests should be protected as strategic learning assets.
Impact And Effort Separation Layer
Impact and development effort are not the same thing.
A low-effort change can create high impact if it affects a critical user touchpoint.
A high-effort change can create low impact if it changes something users do not care about.
Rule
Experiment priority should not be judged by build effort alone.
Critical Touchpoint Layer
Small changes at high-leverage touchpoints may produce major impact.
Examples
- checkout CTA
- pricing selection
- subscription choice
- lead form submission
- navigation decision point
- product selection area
Rule
Touchpoint importance influences impact potential.
Portfolio Balance Layer
MWMS should maintain a healthy mix of experiment types.
Suggested default balance:
- 60% iterative
- 30% substantial
- 10% disruptive
This ratio may change depending on:
- business maturity
- traffic level
- resource availability
- strategic urgency
- risk exposure
Rule
The portfolio should balance safety, learning, and breakthrough potential.
Velocity Layer
Test velocity matters, but not at the expense of learning quality.
Examples
Weak velocity:
- many small tests with low strategic value
Strong velocity:
- consistent tests across iterative, substantial, and disruptive categories
Rule
Experiment velocity should be quality-adjusted.
Learning Value Layer
Experiments should be assessed by what they teach, not only whether they win.
Examples
- customer motivation insights
- friction discovery
- trust barrier identification
- offer positioning clarity
- pricing sensitivity learning
Rule
Losing tests can still create valuable strategic intelligence.
Business Impact Layer
Experiment portfolios should support business growth, not only interface polish.
Examples
- revenue improvement
- lead quality improvement
- subscription growth
- retention lift
- reduced drop-off
- improved customer confidence
Rule
Experimentation should remain connected to business outcomes.
Problem Statement Layer
Strong experiments begin with clear problem statements.
Examples
- users do not understand the offer
- customers hesitate at checkout
- traffic clicks but fails to convert
- subscribers do not trust flexibility claims
Rule
Problem clarity improves experiment quality.
Hypothesis Layer
Every experiment should have a structured hypothesis.
Required Elements
- problem or opportunity
- proposed change
- expected user behaviour shift
- primary KPI
- supporting diagnostic metrics
- risk conditions
Rule
Experiments without hypotheses weaken learning quality.
Portfolio Risk Layer
A portfolio with only safe tests becomes stagnant.
A portfolio with too many disruptive tests becomes unstable.
Rule
Strategic balance protects both growth and survivability.
Stakeholder Layer
Substantial and disruptive tests often require stronger stakeholder communication.
Requirements
- explain why the test matters
- explain what risk is controlled
- explain what learning is expected
- explain how success will be measured
Rule
Higher-impact experiments require stronger narrative and governance.
AI Governance Layer
AI Employees should:
- classify experiments as iterative, substantial, or disruptive
- estimate effort separately from impact
- identify portfolio imbalance
- recommend stronger problem statements
- preserve strategic experimentation diversity
- flag overreliance on low-impact tweaks
Rule
AI systems must protect portfolio balance.
Reporting Layer
Experimentation reports should communicate:
- test type
- expected impact
- build effort
- learning value
- business metric connection
- survivability risk
- portfolio balance status
Rule
Experimentation portfolios should remain operationally visible.
Escalation Layer
Portfolio imbalance may require governance review.
Escalation Conditions
- too many minor tests
- no disruptive learning
- excessive risky testing
- weak problem statements
- tests not tied to business outcomes
- low stakeholder trust in experimentation
Rule
Portfolio imbalance should trigger strategic correction.
Measurement Layer
MWMS should monitor:
- percentage of iterative tests
- percentage of substantial tests
- percentage of disruptive tests
- winner rate by test type
- learning value by test type
- business impact by test type
- effort versus impact relationship
Rule
Portfolio quality must remain measurable.
AI Decision Boundary Layer
AI Employees may:
- recommend portfolio balance changes
- classify experiment types
- suggest higher-impact test opportunities
- identify low-value experimentation patterns
AI Employees must not:
- flood the roadmap with random tests
- prioritize effort alone
- eliminate disruptive testing capacity
- optimize only for short-term test velocity
Rule
Experiment portfolio governance constrains experimentation authority.
Cross Brain Integration
Experimentation Brain
→ owns experiment portfolio governance
Data Brain
→ validates evidence and metric reliability
Research Brain
→ supplies problem themes and customer insights
Affiliate Brain
→ supplies offer and commercial test opportunities
Ads Brain
→ supplies acquisition and creative test opportunities
Conversion Brain
→ supplies funnel and UX test opportunities
Finance Brain
→ evaluates ROI, effort, and allocation exposure
HeadOffice
→ governs strategic alignment and survivability
AI Employees
→ operate within portfolio-balance governance boundaries
Failure Modes Prevented
This framework prevents:
- endless small-test stagnation
- experimentation theatre
- weak learning velocity
- effort-based prioritization errors
- no breakthrough testing
- stakeholder disengagement
- random spaghetti testing
Drift Protection
The system must prevent:
- confusing test volume with experimentation maturity
- overvaluing low-effort tweaks
- ignoring disruptive learning opportunities
- separating experiments from business outcomes
- allowing unbalanced roadmaps
- AI test-generation spam
Architectural Intent
This framework transforms MWMS experimentation from:
→ isolated test execution
into:
→ strategic experimentation portfolio management.
It ensures MWMS develops:
- balanced testing roadmaps
- stronger business impact
- better learning depth
- protected innovation capacity
- survivability-aware experiment governance
- long-term experimentation maturity
Final Rule
A mature experimentation program does not only ask:
“Can we test this?”
It asks:
“What role does this test play in the whole portfolio?”
Change Log
Version: v1.0
Date: 2026-05-08
Author: HeadOffice
Change:
Created Experiment Portfolio Governance Framework defining iterative, substantial, and disruptive experiment balance, impact-effort separation, portfolio visibility, and strategic experimentation maturity governance.
Change Impact Declaration
Pages Created:
Experimentation Brain Experiment Portfolio Governance Framework
Pages Updated:
None
Pages Deprecated:
None
Registries Requiring Update:
MWMS Architecture Registry
Experimentation Brain Page Registry
Canon Version Update Required:
No
Change Log Entry Required:
Yes