This glossary defines the shared language used across Measuring Products. Definitions are practical rather than exhaustive. Use them to keep articles, diagrams, templates and interface copy consistent.
editorial/editorial-decisions.md takes precedence where a term is under active revision.
Core measurement terms
Product measurement
The practice of using evidence about product behaviour to support better product decisions. It includes deciding what matters, capturing evidence, defining metrics, interpreting results and maintaining the system over time.
Do not use it as a synonym for dashboards, reporting or analytics tooling alone.
Measurement system
The connected set of questions, workflows, event records, metric definitions, dashboards, ownership and operating practices used to understand and improve a product.
Measurement design
The work of deciding what needs to be understood and specifying the workflows, evidence, events, metrics and interpretation needed to support that understanding.
Measurement operations
The work of implementing, validating, governing, reviewing, maintaining and retiring measurement after it has been designed.
Measurement practice
The habits, responsibilities and routines through which a team designs and operates measurement.
Product question
A specific question about product behaviour, value, quality, risk or change that measurement should help answer.
Examples:
- Where do applicants struggle to continue?
- Are more organisations reaching a usable plan?
- Did the new workflow reduce avoidable rework?
Decision
A choice or action that evidence should inform. A decision gives measurement purpose; the workflow provides the practical structure for designing it.
Evidence
Recorded or observed information that supports understanding. Evidence may come from event records, research, support contacts, operational systems, logs or other sources.
Evidence is not automatically insight.
Insight
An interpretation of evidence that changes understanding or supports action. Insight depends on context, comparison and judgement.
Confidence
The degree of trust a team can reasonably place in a conclusion or decision. Confidence depends on data quality, definition quality, sample size, context, limitations and the strength of the evidence—not only numerical precision.
Workflow terms
Workflow
A sequence of user and system activity through which a person or organisation attempts to achieve a meaningful outcome.
A workflow is not the same as a screen, feature or complete user journey.
Workflow step
A meaningful unit of progress within a workflow. Steps should describe behaviour or state change rather than interface location.
Prefer “confirm contact details” to “visit page three”.
Entry point
The condition or observed activity that marks where analysis of a workflow begins. A workflow may have multiple entry points.
Exit
A point at which a person leaves or pauses the workflow without reaching the selected completion point. An exit is not always a failure.
Completion point
The condition that marks the end of the workflow for a specific measurement question. Completion should be defined explicitly and may differ from receiving value.
User intent
What a person is trying to achieve. Intent is inferred through research and context; it should not be claimed from clickstream data alone.
User journey
A broader experience across time, channels, touchpoints and contexts. A journey may contain several workflows and activity outside the product.
Feature
A product capability. Features enable workflows, but feature boundaries do not necessarily match how users achieve outcomes.
Event and instrumentation terms
Event
A structured record produced when instrumentation observes selected product or system activity.
In lighter prose, “event” may refer to the record and the named event type. Where precision matters, use event record.
Products produce behaviour and system activity. Instrumentation records selected activity as events.
Event name
The stable identifier describing the observed activity represented by an event, such as registration.started or account.first_accessed.
Event property
Context attached to an event record that helps interpret it, such as entry route, account type or validation result.
Properties should be purposeful, stable and privacy-conscious.
Fire condition
The exact condition under which an event should be recorded. It should be observable and testable.
Instrumentation
The technical implementation that records selected product and system activity as event data or other evidence.
Instrumentation requirement
A statement describing what needs to be recorded, when, with which properties and how the implementation will be validated.
Tracking plan
A coordinated plan for implementing or changing instrumentation. Depending on the organisation, it may combine event definitions, specifications, ownership and release scope.
Metric terms
Metric
A calculated measure used to answer a product question, monitor a condition, diagnose a problem, evaluate a change or support a decision.
Metrics are calculated from evidence. Products do not directly generate metrics.
Metric definition
The durable specification of a metric: its purpose, formula, population, time window, exclusions, dimensions, owner and limitations.
Formula
The calculation that turns source data into a metric. A formula is incomplete without its population, time window and exclusions.
Population
The people, sessions, accounts, organisations, items or attempts eligible for inclusion in a metric.
Unit of analysis
The entity being counted or compared, such as a person, session, account, organisation, application or workflow attempt.
Filter
A rule that limits the records included in an analysis or view.
Exclusion
A documented rule removing records that would make the metric misleading or invalid, such as test accounts or known migration artefacts.
Time window
The period over which activity is included or compared. Definitions should distinguish reporting-period metrics from cohort metrics where necessary.
Cohort
A group defined by a shared starting condition or time, followed over an appropriate period. Cohorts help avoid mixing new and long-running activity in misleading ways.
Dimension
A stable attribute used to break down or compare a metric, such as entry route, organisation type or device category.
Segment
A meaningful group of people, accounts or organisations defined using one or more dimensions or behaviours.
Signal
A piece of evidence that may indicate a meaningful product condition or change. A signal is not necessarily conclusive.
Indicator
A metric used as a practical signal of a wider condition that cannot be observed directly or completely.
Leading indicator
A measure expected to change before a later outcome. The relationship is normally a hypothesis and should be tested rather than assumed.
Lagging indicator
A measure that reflects an outcome after the contributing activity has occurred.
Outcome
A meaningful result for a user, organisation, service or business. Outcomes may occur partly or entirely outside the digital product.
Conversion
The proportion of an eligible population that reaches a defined completion point. The denominator, time window and unit must be stated.
Activation
Evidence that a new user or account has reached an early point of meaningful value. Activation is product-specific and should not be reduced to a generic login or registration event.
Engagement
Meaningful use of a product over time. Engagement must be defined in relation to value; raw activity volume is not automatically engagement.
Success
A defined condition indicating that the workflow or product achieved the outcome relevant to the question. Success is contextual and may require more than digital completion.
Failure
A defined condition indicating that a workflow could not continue or achieve its intended outcome. Do not label every exit or pause as failure.
Friction
Effort, confusion, delay or repeated work that makes progress harder. Some friction is protective or necessary, so the term should be interpreted in context.
Metric structures
Metric tree
A hierarchical representation of relationships between measures. State which relationship the tree represents:
- mathematical decomposition;
- hypothesised drivers;
- diagnostic workflow measures.
Do not mix these relationships without making the distinction visible.
North Star metric
A high-level measure intended to represent recurring customer value and provide strategic focus. It should not replace supporting measures, constraints or judgement.
Measurement framework
A coherent structure connecting a question or decision to workflows, evidence, events, metrics, interpretation and action.
Dashboard and reporting terms
Dashboard
A curated view of metrics and supporting context designed for a defined monitoring, diagnostic, evaluation or health purpose.
A dashboard is an output of measurement design, not the strategy itself.
Report
A structured presentation of information for a particular audience, period or decision. A report may include narrative, interpretation and recommendations as well as metrics.
Chart
A visual representation of data. A chart is useful only when its form supports the comparison or interpretation the reader needs.
View
A particular arrangement or filtered presentation of evidence for an audience or task.
Operating terms
Measurement maturity
The demonstrated capability of a team or organisation to design, implement, use and maintain trustworthy measurement.
Maturity should be assessed through observable practices rather than a single vanity score.
Measurement debt
The accumulated cost and risk created by weak, outdated, duplicated, undocumented or unused measurement.
Measurement health
The current reliability, usefulness and maintainability of the measurement system.
Measurement governance
The rules, responsibilities and decision processes used to keep measurement consistent, appropriate and trustworthy.
Ownership
Accountability for a definition, implementation, decision or maintenance action. Ownership does not mean one person performs all the work.
Stewardship
Ongoing care for the quality, relevance and use of measurement artefacts and systems.
Measurement review
A deliberate assessment of whether a workflow, event, metric, dashboard or broader measurement system remains correct, useful and worth maintaining.
Review rhythm
The regular or trigger-based cadence through which measurement is reviewed. Useful triggers include product changes, instrumentation changes, disputed metrics and declining use.
Measurement artefacts
Workflow catalogue
A maintained inventory of important workflows, their boundaries, owners, questions, risks and measurement status.
Event catalogue
The durable record of what each event means, when it fires, which properties it carries, who owns it and what it does not prove.
Metric catalogue
The durable record of metric definitions, formulas, populations, windows, dimensions, owners and limitations.
Instrumentation specification
The implementation contract for adding or changing instrumentation. It translates measurement design into build and validation requirements.
Acceptance criteria
Testable conditions used to decide whether an implementation meets its requirement.
Roles
Product team
The cross-functional group responsible for understanding, delivering and improving a product. Measurement is a shared team responsibility.
Product manager
Typically accountable for product questions, priorities, decisions and whether measurement supports the intended product outcome.
Designer
Contributes understanding of workflows, interactions, states, accessibility and the behaviour the product enables.
UX researcher
Provides evidence about intent, context, experience and explanation that event data alone cannot provide.
Engineer
Implements and validates instrumentation, data contracts and relevant technical safeguards.
Analyst
Develops definitions, checks data quality, performs analysis and supports responsible interpretation.
Product leader
Creates the conditions for useful measurement, aligns it with strategy and ensures that teams act on evidence rather than merely report it.
General data terms
Analytics
The analysis and interpretation of data to understand behaviour, performance or outcomes. Analytics is one part of product measurement.
Data
Recorded facts or representations used as evidence. Data is not automatically accurate, complete, neutral or useful.
KPI
A key performance indicator: a metric selected to represent performance against an important objective. A metric does not become useful merely because it is labelled a KPI.