Metric, target and observation
Support traceable performance assessment and target monitoring while preserving the meaning and comparability of observations across revisions.
Bundle → Layer → Finding → Questions Filled
2 bundles · 4 layers · 6 findings · 12 questions
Definition Defines the quantity, method and breakdowns used by a metric.
Metric and method
Separates metric meaning from versioned calculation methods.
Metric definition
Defines the measured quantity, unit and aggregation rules.
- Which quantity and unit does this metric definition specify?
- Which aggregation rule does this metric definition specify across breakdowns and periods?
Method version
Records calculation changes and their implications for comparison.
- What calculation change does this method version record relative to its predecessor?
- What comparison with its predecessor does this method version permit or forbid, with what justification?
Breakdowns
Defines the dimensions and population boundaries of a metric.
Dimension definitions
Records the breakdown vocabulary and population definition.
- Which dimensions and code list versions does this metric definition reference?
- Which population inclusion and exclusion rules does this metric definition specify?
Observation and target Separates measured results, their revisions and intended target values.
Observation
Records the result and the evidence needed to interpret its revision.
Observation record
Pins the observation context and distinguishes values from missingness.
- Which metric, method version, period, population and breakdown does this observation revision pin?
- Does this observation revision contain a measured value, including zero, or a missing value with a stated reason such as unknown or not applicable and a separate encoding such as null?
Recalculation
Preserves distinguishable observation revisions and comparison decisions.
- What revision identifier, retained predecessor reference and recalculation reason does this observation revision record?
- What comparison with its predecessor does this observation revision permit or forbid, with what justification?
Target
Records the intended value and the basis on which performance is assessed.
Target value
Identifies the target authority, period and calculation basis.
- Which value, period and setting authority does this target record specify?
- Which metric and method version does this target record require for assessment?
Classifiers Filled
- Family
- Enterprise profiles
- Category
- Enterprise subject
- Entry kind
- subject
- Domain
- EnterpriseDAT
- Industry
- Cross-industry
- Tags
- EM-DAT-05W1subjectMetricDefinitionMeasurementMethodDimensionDefinitionObservationTargetValue
What it is Filled
Definition of the measured quantity, unit, breakdowns, calculation, observation and target value. The formula and the measured result have different versions.
Why it exists Filled
Support traceable performance assessment and target monitoring while preserving the meaning and comparability of observations across revisions.
Distinguishing features Filled
- The value of a metric lives in observations, never as a field of the metric definition.
- Each observation pins the method version, the period and the population it describes.
- Units and aggregation rules are explicit, so headcount and FTE are never summed together.
- Zero is a measured value, while unknown and not applicable are distinct semantic reasons for missingness that must be recorded separately from an encoding such as null.
What robots and AI may and may not do Filled
Must not
- Sum headcount and full-time equivalent values together.
- Compare observations from different method versions without a documented comparability justification and any required adjustments.
- Replace a missing value with zero.
- Overwrite a published observation when recalculating it.
Only with a human decision
- Obtain human approval before changing the formula of a key performance indicator used for targets or pay.
- Obtain human approval before setting or revising targets.
May
- Compute observations and record method version, period and population.
- Check whether two observations are comparable before charting them together.
- Flag values whose method changed between periods.
Moral aspects Filled
- Metrics used to judge people or units can be unfair when methods or populations differ.
- Misleading figures can drive harmful budget and staffing decisions.
Who is affected
- Employees and teams measured by the metric
- Managers who set targets
- Readers of reports
Owners Filled
Steward
The data owner or head of analytics who answers for data assets and their fitness for use.
Master systems
- Data catalogue
- Data warehouse
- BI platform
- Data orchestrator
Links to other meta-models Filled
neighbor
- EM-DAT-06
- EM-STR-02
references
- WM-XCT-025 - conceptual-candidate
- WM-DAT-010 - conceptual-candidate
What else AI and robots need to interact with it Filled
Identity and identifiers required Filled
- A metric has a stable identifier, and its calculation methods have separate version identifiers.
- An observation is identified by metric, period, population and breakdown, with a revision identifier distinguishing each retained result.
Direct properties not applicable Not applicable
Not applicable
Enterprise record contour: physical properties belong to referenced world models.
Recognition required Filled
- A metric record has a unit, an aggregation rule and a versioned method.
- A metric definition can be confused with a single dashboard number or its target.
Capabilities and actions required Filled
- Pinned methods and populations support comparability assessment across a time series.
- Recalculations can be traced alongside original values.
Hazards and failure modes required Filled
- Silent formula changes can break trend comparability.
- Mixing units or populations can produce invalid totals.
- Substituting zero for an unknown value misrepresents the evidence.
Standards and interfaces optional Filled
- The RDF Data Cube Vocabulary from the World Wide Web Consortium supports observations and dimensions.
- Statistical Data and Metadata eXchange from the SDMX sponsoring organizations supports statistical data and metadata exchange.
- The 80000 series from the International Organization for Standardization and the International Electrotechnical Commission provides standards for quantities and units.
Context of use required Filled
- Metric records support performance management, regulatory statistics and operational dashboards.
- Public statistics and regulated disclosures follow defined methodologies.
Sources Filled
- W3C DCAT: dataset, distribution, catalog
- W3C PROV and Data Cube: provenance and observation
- SDMX and BI/data catalog practice: indicators, breakdowns, definitions and releases
- W3C The RDF Data Cube Vocabulary - Checked live (HTTP 200) on 2026-10-07.
- W3C PROV-O: The PROV Ontology - Checked live (HTTP 200) on 2026-10-07.
- SDMX: Statistical Data and Metadata eXchange - Checked live (HTTP 200) on 2026-10-07.
Open questions
- How to ensure comparability of periods and populations?
- How to distinguish null, zero, not-applicable and unknown?
- What changes when a KPI gets a new formula?
- Установить границу и решение reuse/extend/new по действующим спецификациям.
- Подтвердить semantic crosswalk, права и source mastership.
- Выбрать immutable refs; провести проверки fixtures до заявления о публикационной готовности.
Machine files
Provenance
enterprise research programme · published-partial
Built from: enterprise/models/em-dat-05/brief.json, ver-cy/world-models/card-supplements/em-dat-05.json
Still in Russian: suggested owner, blocking decisions, vercy candidates. Translate in research/enterprise/i18n/units.en.json.