Document Type: Framework
Status: Structural
Version: v1.1
Authority: HeadOffice
Applies To: Data Brain, Experimentation Brain, Ads Brain, Affiliate Brain, Research Brain
Parent: Data Brain
Last Reviewed: 2026-04-22
Purpose
The Signal Design Specification Framework defines how measurement signals are structured before implementation.
It ensures that tracking architecture is:
consistent
interpretable
decision-aligned
experimentation-ready
scalable
technically stable
This framework converts measurement strategy into an actionable signal structure.
It prevents:
ambiguous event design
inconsistent parameter usage
duplicated signals
unclear conversion definitions
fragmented measurement logic
unstable tracking structures
Signal structure must be defined before implementation begins.
Well-structured signals improve interpretability.
Improved interpretability strengthens optimisation decisions.
Structured signals improve experiment reliability.
Structured signals improve learning continuity across MWMS.
Scope
This framework governs:
event structure design
parameter structure design
metric definition structure
dimension definition structure
conversion signal definitions
signal naming conventions
signal reuse logic
signal expansion logic
signal validation preparation
recommended vs custom event selection discipline
parameter reuse structure discipline
signal hierarchy discipline
This framework applies to:
website tracking
landing page tracking
advertising signals
funnel signals
conversion signals
behavioural signals
experiment measurement signals
server-side tracking signals
tag-managed signals
Core Principle
Signals must be intentionally designed.
Signal design must support decision-making.
Signal design must support experimentation.
Signal design must support interpretation.
Signal design must support consistency across time.
Signal design must be documented before implementation.
Signal structure must reflect behavioural meaning rather than technical implementation detail.
Signal structure must support reliable interpretation across environments.
Signal Design Hierarchy
Signal structures must follow a layered hierarchy.
Business Objective
→ Business Question
→ Decision Requirement
→ Signal Requirement
→ Event Definition
→ Parameter Definition
→ Metric Definition
→ Dimension Definition
→ Conversion Definition
→ Validation Requirement
Each level must align with the level above.
Signal definitions must not exist without decision relevance.
Signals should not be created without interpretation purpose.
Signal hierarchy ensures measurement aligns with decision requirements.
Event Structure Principles
Events represent observable behavioural interactions.
Events must:
represent meaningful behavioural actions
avoid duplication
avoid semantic ambiguity
avoid unnecessary fragmentation
avoid over-specific naming
avoid over-general naming
avoid platform-dependent terminology
avoid implementation-specific naming
Event names should:
represent behaviour
remain stable across time
remain interpretable by humans
remain comparable across campaigns
remain consistent across experiments
Examples of event types:
page_view
cta_click
form_submit
video_start
video_complete
scroll_depth
add_to_cart
checkout_start
checkout_complete
outbound_click
button_click
generate_lead
purchase
Events must represent behavioural meaning, not implementation detail.
Behavioural meaning improves interpretability.
Interpretability improves optimisation direction.
Recommended vs Custom Event Discipline
Signals should prioritise existing event structures where possible.
Event selection hierarchy:
automatically collected events
recommended events
custom events
Recommended events improve:
cross-platform comparability
standardised interpretation
future compatibility
Custom events should be used when:
no suitable recommended event exists
behaviour is business-specific
behaviour represents unique decision relevance
Custom events should not duplicate existing event meaning.
Duplicate event logic reduces signal clarity.
Signal clarity improves decision confidence.
Parameter Structure Principles
Parameters provide contextual information about events.
Parameters allow interpretation of behavioural meaning.
Parameters should:
describe context
describe characteristics
describe classification
describe value
describe segmentation characteristics
Examples:
traffic_source
page_type
offer_id
campaign_angle
device_type
user_status
checkout_step
content_category
interaction_type
interaction_location
interaction_context
Parameters should remain reusable across events when possible.
Reusable parameters reduce structural complexity.
Reusable parameters improve signal continuity.
Reusable parameters improve cross-campaign comparability.
Reusable parameters reduce parameter quota pressure.
Parameter structure should prioritise interpretability over granularity.
Parameter Consistency Discipline
Parameter meaning must remain stable across environments.
Changing parameter meaning creates interpretation drift.
Parameter consistency improves:
report reliability
historical comparability
experiment comparability
decision clarity
Stable parameters improve learning continuity.
Required Parameter Awareness
Some recommended events require specific parameters.
Examples:
generate_lead requires value and currency
purchase requires transaction structure
Missing required parameters alters event classification behaviour.
Incorrect parameter structure may reduce interpretability.
Parameter completeness improves signal usability.
Parameter completeness improves platform compatibility.
Metric Definition Principles
Metrics quantify behavioural magnitude.
Metrics measure frequency, quantity, or intensity.
Examples:
event_count
conversion_count
click_count
session_count
add_to_cart_count
purchase_count
lead_count
Metrics must:
remain consistently defined
avoid redefinition across time
avoid conflicting interpretation
Metrics must clearly represent numerical meaning.
Metric definition stability improves trend reliability.
Metric consistency improves experiment comparability.
Dimension Definition Principles
Dimensions provide contextual segmentation of metrics.
Dimensions enable comparison across user characteristics or behavioural categories.
Examples:
traffic source
campaign type
device type
geographic region
content category
user segment
offer type
Dimensions must:
remain stable
remain interpretable
remain reusable across reports
avoid redundant naming
avoid semantic duplication
Dimensions enable structured comparison.
Structured comparison improves optimisation clarity.
Conversion Signal Definition
Conversion signals represent meaningful outcome events.
Conversions represent behavioural completion of value-generating actions.
Examples:
purchase completed
qualified lead generated
registration completed
trial started
application submitted
booking completed
strategy call booked
Conversions must:
align with business objectives
align with decision thresholds
align with experiment success criteria
Conversion signals must be clearly defined before implementation.
Conversion signals must remain stable across experiments when possible.
Changing conversion definitions reduces data continuity.
Stable conversion definitions improve experiment comparability.
Stable conversion definitions improve optimisation consistency.
Signal Reuse Principle
Signals should be reused where meaning remains consistent.
Signal reuse improves:
data comparability
historical continuity
reporting clarity
experiment comparability
Examples:
“page_type” parameter reused across multiple events
“offer_id” reused across funnels
“traffic_source” reused across campaigns
“interaction_type” reused across engagement events
Reuse reduces unnecessary parameter proliferation.
Reuse reduces structural fragmentation.
Reuse improves learning continuity.
Signal Naming Consistency
Signal names must remain:
consistent
interpretable
stable
platform-independent
human-readable
Naming must avoid:
unnecessary abbreviations
inconsistent formatting
inconsistent casing
ambiguous terminology
naming duplication
consistent naming improves signal clarity.
Signal clarity improves interpretability.
Interpretability improves optimisation confidence.
Signal Expansion Logic
Signals may evolve over time.
Signal expansion must follow governance discipline.
Expansion must:
maintain backward interpretability
avoid breaking existing reporting
avoid renaming existing signals unnecessarily
extend signal structure logically
Signal evolution must preserve interpretability.
Signal evolution must preserve comparability.
Uncontrolled expansion reduces signal clarity.
Controlled expansion supports long-term scalability.
Signal Documentation Requirement
Signal structures must be documented prior to implementation.
Documentation must include:
event definitions
parameter definitions
metric definitions
dimension definitions
conversion definitions
expected values
signal relationships
Documentation ensures:
consistent implementation
consistent validation
consistent interpretation
Documentation supports cross-brain coordination.
Documentation reduces structural drift risk.
Validation Preparation Requirement
Signal design must prepare for validation.
Signal design must allow verification of:
correct event firing
correct parameter population
correct metric behaviour
correct conversion triggering
correct event sequencing
correct routing destination
Validation logic must be definable before implementation.
Validation clarity improves implementation reliability.
Validation readiness improves signal trustworthiness.
Relationship to Data Layer Architecture
Signal definitions must align with data layer structure.
Events and parameters must map clearly to data layer schema.
Data layer implementation must reflect signal design.
Signal structure must remain independent of implementation technology.
Signal meaning must remain stable even if tools change.
Structured data layer architecture improves signal reliability.
Relationship to Experimentation
Signal design must support experiment measurement.
Experiments require:
primary success metrics
diagnostic supporting signals
behavioural interpretation signals
Signal structure must allow causal interpretation.
Signal clarity improves experiment reliability.
Signal clarity improves learning speed.
Relationship to Ads Brain
Signal design supports campaign optimisation.
Ad signals require:
consistent conversion definitions
consistent click signals
consistent engagement signals
consistent funnel progression signals
Signal consistency improves optimisation speed.
Signal clarity improves campaign comparability.
Relationship to Affiliate Brain
Signal design supports offer evaluation.
Affiliate signals require:
consistent conversion definitions
consistent funnel progression signals
consistent engagement signals
consistent performance benchmarks
Signal structure supports scaling decisions.
Signal clarity improves opportunity evaluation confidence.
Governance Requirements
Signal structures must:
align with MWMS naming discipline
maintain interpretability across brains
avoid duplication
avoid structural fragmentation
avoid uncontrolled expansion
Signal structure changes must follow structured review.
Signal structure stability supports reliable learning.
Drift Prevention
Signal drift occurs when:
signal definitions change meaning
parameter values change interpretation
conversion definitions change without documentation
event naming changes inconsistently
parameter meaning evolves without visibility
Signal drift reduces data reliability.
Signal drift reduces decision confidence.
Signal drift must be prevented through disciplined signal design.
Stable signal definitions improve learning continuity.
Stable signal definitions improve optimisation reliability.
Minimum Viable Signal Set Principle
Initial implementations should prioritise:
decision-critical signals
conversion-critical signals
diagnostic signals required for interpretation
Additional signals may be added when justified.
Minimal viable signal structure enables faster implementation.
Minimal viable signal structure enables faster learning cycles.
Minimal viable signal structure reduces structural complexity risk.
Change Log
Version: v1.1
Date: 2026-04-22
Author: HeadOffice
Change:
Expanded framework to include:
recommended vs custom event discipline
parameter reuse discipline
validation readiness structure
signal hierarchy clarity
conversion stability logic
alignment with:
Custom Event Design Framework
GTM Signal Governance Framework
Debugging and Validation Framework