ISO 22400 KPI definitions should not sit with a single function. In regulated, long-lifecycle environments, they are best owned through a formal cross-functional governance model, with a clear operational steward.
Recommended ownership model
A practical pattern is:
- Cross-functional governance body as the owner of definitions
Typically a performance management council, data governance board, or similar forum with representation from:
- Operations / Manufacturing engineering (how work is actually executed)
- Quality (traceability, auditability, and alignment with QMS)
- IT/OT / MES team (data sources, integration, system changes)
- Supply chain / Planning (where KPIs touch scheduling and materials)
- Finance / Controlling (cost-related KPIs and financial alignment)
This group owns the authoritative KPI catalog, including ISO 22400 mappings, calculation logic, and change control.
- Operations as day-to-day steward
Operations leadership (plant manager, value stream leader, or equivalent) should typically be the steward for production KPIs such as OEE and NPT: they rely on them for decisions and are accountable for results. They do not own the IT implementation in isolation, but they own how the KPI is used operationally.
- IT/OT as technical custodian
IT/OT teams (often the MES/SCADA data owners) are responsible for implementing the agreed definitions in systems, guarding data lineage, and documenting interfaces. They should not change KPI logic unilaterally; they execute changes approved by the governance body.
Why single-function ownership usually fails
Common failure modes when one group “owns” ISO 22400 KPIs in isolation:
- IT-only ownership: KPIs are technically consistent but disconnected from how work is actually done. Plants bypass official dashboards and create shadow spreadsheets, breaking traceability.
- Operations-only ownership: Each plant tweaks definitions (e.g., what counts as planned vs unplanned downtime), destroying comparability across sites and complicating audits.
- Quality-only ownership: Definitions may be robust from a compliance standpoint but too slow to change or too narrow for real-time performance management.
ISO 22400 is a common language for manufacturing KPIs. In brownfield environments with mixed MES/ERP/PLM/QMS, no single function sees the full picture. A shared governance model is usually the only way to keep definitions stable, comparable, and auditable.
Key responsibilities to assign explicitly
Regardless of your exact org chart, make sure the following responsibilities are clearly assigned:
- Definition authority: Who approves a new KPI or a change to an existing one (e.g., OEE, availability, performance, quality rate) and ensures it stays consistent with ISO 22400 concepts where applicable.
- Data source and lineage ownership: Who documents and approves which systems and tags feed each KPI, how manual data entry is controlled, and how changes to those sources are validated.
- Calculation & configuration control: Who maintains the official calculation logic, handles versioning, and manages change requests in MES/BI/ERP. This should run under formal change control, not ad-hoc edits.
- Validation & reconciliation: Who signs off that KPI outputs match reality (e.g., cross-check OEE vs shift reports or maintenance logs) when systems, routing rules, or data integrations change.
- Access & usage rules: Who decides where KPIs are exposed (shop floor displays, management reports, supplier scorecards) and ensures that unofficial variants do not proliferate.
Working in brownfield, multi-system environments
In most regulated plants, you will have more than one system touching each KPI (MES, ERP, historians, manual logs). Full replacement with one “KPI system” is rarely realistic due to validation burden, downtime risk, and integration complexity. Instead:
- Keep one definition of each KPI, even if it is instantiated in multiple tools.
- Document for each KPI: system of record, system of display, and any site-specific exceptions that have been formally approved.
- Run changes through the same change control used for other validated systems; re-validate critical KPIs when integration or MES logic changes.
For regulated sectors, this approach supports traceable, consistent KPIs without assuming a greenfield stack or a single-vendor solution.
