Metric Catalog Starter — Taxonomy, Rules & Example Metrics

Welcome — purpose and quick value

This starter catalog helps teams stop metric proliferation and align measurement with decisions. Use the template to capture the definition, ownership, calculation, source, audience, and decision context for each metric so everyone understands what it means, where it comes from, and how it should be used.

Core columns (what to record and why)

• Metric name: A short, consistent name. Use agreed prefixes or families (see Taxonomy below).
• Owner: Primary accountable person or role (not just a team). The owner approves the definition and is responsible for metric governance.
• Steward / Contact: Person who produces or maintains the data (can be a data engineer, analyst, or operations lead).
• Formula / Definition: Exact calculation, including numerator, denominator, filters, and any transforms. Write it in plain language and, where possible, include SQL/pseudocode or reference to the exact automated calculation.
• Source: System(s) and tables of record (e.g., ERP.orders, WMS.shipments, CRM.opportunities). Note any data latency, batching, or ETL steps.
• Frequency / Cadence: How often it is computed and reviewed (real-time, hourly, daily, weekly, monthly).
• Audience: Who uses this metric (roles, teams, external stakeholders).
• Decision use: The decisions or actions this metric supports (e.g., daily dispatching, monthly budgeting, safety interventions). This links the metric to purpose.
• Target / Thresholds: Desired performance and any thresholds that trigger action (warning/critical) and how they were set (historical, regulatory, benchmarking).
• Units / Format: Percent, count, $USD, minutes, etc., and rounding rules.
• Sample dashboard placement: Suggested dashboard or report and card placement/context (e.g., Operations > Daily OEE board > Top-right).
• Data quality notes: Known caveats, common data issues, coverage gaps, and corrective actions in place.
• Last reviewed / Version: Date the definition was approved and version identifier. Keep change history short and linked if possible.

Naming and taxonomy guidance

Adopt a small, consistent taxonomy to avoid duplicated concepts. Use families and short prefixes to group metrics (examples):

• OPS_ — Operational / throughput (OPS_OEE, OPS_Throughput)
• QLT_ — Quality (QLT_DefectRate, QLT_RMA)
• FIN_ — Financial (FIN_GrossMargin, FIN_AR_Days)
• SAF_ — Safety (SAF_LostTimeInjuries)
• CX_ — Customer / Experience (CX_NPS, CX_OnTimeDelivery)

Keep names readable: use widely understood abbreviations and avoid team-specific jargon. If a metric is derived differently in different areas, treat them as distinct metrics with clear qualifiers (e.g., CX_OnTimeDelivery_Shipping vs CX_OnTimeDelivery_3PL).

Owner & stewardship roles — who does what

• Owner (Accountable): Approves the definition, targets, and decision rules. Ensures metric remains useful and aligned with outcomes.
• Steward (Responsible): Maintains calculation, source mapping, and data quality. First responder for anomalies.
• Consumer (Informed): Teams and roles that rely on the metric to act. Their feedback should shape metric evolution.

Measurement rules — practical conventions

• Prefer objective, system-captured sources over manual spreadsheets where possible.
• Define time windows explicitly (calendar vs rolling periods) and timezone rules for cross-site metrics.
• Document handling of nulls, outliers, adjustments, and returned/cancelled items.
• Include a short unit test or validation query and where it is stored (e.g., a notebooks folder or analytics repo link).
• When a metric is derived from multiple systems, describe reconciliation rules and expected divergence tolerance.

Starter example metrics

Use these examples as copyable starter rows. Each entry lists key fields concisely so teams can paste into their catalog and adapt.

• OPS_OEE
  Owner: Plant Operations Manager
  Steward: Manufacturing Analyst
  Formula: Availability x Performance x Quality (Availability = RunTime / PlannedProductionTime; Performance = ActualOutput / IdealOutput; Quality = GoodUnits / TotalProduced)
  Source: MES.production_events, ERP.production_orders
  Frequency: Daily
  Audience: Plant leadership, continuous improvement teams
  Decision use: Daily focus areas for production losses and CAPEX requests
  Sample dashboard placement: Plant > Daily OEE board
• QLT_DefectRate
  Owner: Quality Manager
  Steward: QA Analyst
  Formula: (DefectiveUnits / TotalInspectedUnits) * 100
  Source: QA.inspections
  Frequency: Weekly
  Decision use: Root cause investigations, supplier quality actions
  Target: < 0.5%
• CX_OnTimeDelivery
  Owner: Logistics Director
  Steward: Logistics Analyst
  Formula: Shipments delivered on or before promised date / Total shipments (%)
  Source: WMS.shipments, TMS.delivery_confirmations
  Frequency: Weekly
  Decision use: Carrier performance, customer communications
• FIN_CashConversionCycle
  Owner: Finance Lead
  Steward: FP&A Analyst
  Formula: DSO + DIO - DPO
  Source: AR, Inventory, AP ledgers
  Frequency: Monthly
  Decision use: Working capital management and vendor negotiation
• SAF_LostTimeInjuryRate
  Owner: Safety Manager
  Steward: EHS Coordinator
  Formula: (Lost time injuries x 200,000) / Hours worked
  Source: IncidentReports system
  Frequency: Monthly
  Decision use: Safety program adjustments, training prioritization
• HR_TurnoverRate
  Owner: HR Director
  Steward: HR Analyst
  Formula: (Separations during period / Average headcount during period) * 100
  Source: HRIS
  Frequency: Quarterly
  Decision use: Retention programs, recruiting planning

How to use this template (practical steps)

1. Gather a small cross-functional working group (owner, steward, key consumers).
2. Inventory existing metrics and find duplicates or near-duplicates. Mark one as canonical; others become documented variants or candidates for retirement.
3. For each metric, fill every core column. If a column is unknown, add an action item and owner to resolve it.
4. Approve definitions with Owners and publish the catalog where teams can view it. Set a cadence (e.g., quarterly) for review.
5. Track change history and require owner sign-off before making definition or source changes that affect downstream decisions.

Metric governance checklist (quick)

• Does the metric map to a decision or outcome?
• Is the owner named and committed?
• Is the formula unambiguous and reproducible?
• Is the source authoritative and accessible to stakeholders?
• Are targets and thresholds documented?

Next steps & optional platform enhancements

This HTML template is a practical starting point. Consider these platform enhancements when you're ready to operationalize the catalog:

• Create an interactive metric-entry form (fields above) so teams can submit and update metric definitions and store them via the platform's submission capability.
• Bundle the catalog as an Ownable Domain/Toolkit so sites or teams can copy and tailor a canonical enterprise taxonomy while preserving inheritance and versioning.
• Connect key metric rows to dashboards, alerts, and data quality checks so consumers see current values, variance, and trust indicators next to the definition.

Use the starter examples as a living seed. The goal is not perfection on day one but clarity about meaning, owners, and decision use so teams can act with shared understanding.