Document Type: Framework
Status: Active
Authority: HeadOffice
Applies To: Experimentation Brain, Affiliate Brain, Conversion Brain, Research Brain
Parent: HeadOffice
Version: v1.0
Last Reviewed: 2026-04-18
Purpose
The HeadOffice Growth Lever Theme Framework defines the intermediate structure between Growth Levers and individual experiments.
Themes provide structured groupings of related experiments.
They ensure experimentation produces cumulative learning rather than isolated test outcomes.
Themes improve:
learning coherence
experiment prioritisation clarity
pattern detection speed
strategic focus discipline
Without themes, experiment portfolios become fragmented and inefficient.
Themes allow MWMS to test structured hypotheses about system behaviour.
Scope
This framework applies to:
Experimentation Brain hypothesis structuring
Affiliate Brain testing direction
Conversion Brain optimisation focus
Research Brain signal interpretation
This framework governs:
theme definition logic
experiment grouping logic
learning consolidation logic
This framework does not govern:
individual experiment methodology
statistical analysis approach
channel selection decisions
These remain governed by respective frameworks.
Definition
A Theme is a structured hypothesis area within a Growth Lever.
Themes group experiments investigating related behavioural mechanisms.
Themes provide organisational structure for experimentation.
Example:
Growth Lever:
increase landing page conversion rate
Themes:
clarity improvement
trust signal strengthening
information hierarchy improvement
visual communication effectiveness
Each theme contains multiple experiments.
Theme Structure Role
Themes operate as:
middle layer between strategy and execution.
Structure:
North Star Metric
↓
Growth Lever
↓
Theme
↓
Experiment
Themes prevent random experimentation.
Themes support cumulative learning.
Characteristics of Effective Themes
Effective themes:
focus on identifiable behavioural mechanisms
allow multiple experiment variations
produce interpretable learning patterns
align with Growth Lever objective
Themes should not:
be overly broad
be overly narrow
describe single experiment
duplicate Growth Lever definition
Theme Quantity Guideline
Typical structure:
2–4 themes per Growth Lever.
Too many themes:
reduce learning clarity
reduce experiment velocity
reduce insight consolidation
Too few themes:
limit exploration diversity
increase risk of local optimisation bias
Balanced theme selection improves learning efficiency.
Theme Lifecycle
Themes evolve as learning accumulates.
Themes may:
expand when successful
be refined when partially successful
be retired when consistently ineffective
Themes should be evaluated periodically.
Themes are not permanent structures.
Relationship to Experimentation Brain
Experimentation Brain tests hypotheses within themes.
Themes improve hypothesis coherence.
Themes improve experiment prioritisation quality.
Relationship to Growth Lever Framework
Growth Lever defines optimisation direction.
Theme defines investigation focus area within lever.
Themes support systematic learning.
Governance Rule
Experiments should be grouped under themes where possible.
Theme structure should remain interpretable and stable during experimentation cycle.
Version Control
v1.0
Initial definition of Growth Lever theme structuring framework.