enterprise systems

Usually not in just one place. ERP, MES, and a data warehouse each support different KPI needs. MES is often best for near real-time execution metrics, ERP for financial and order-level measures, and a data warehouse for cross-system reporting. The right split depends on data definitions, timestamp quality, integration reliability, and governance.

Yes, in many cases you can normalize manufacturing data without replacing legacy systems. The practical path is usually a coexistence architecture that maps source data into a governed model. Results depend on source quality, integration depth, master data discipline, and validation effort. It improves interoperability, but it does not remove process debt or make inconsistent systems agree automatically.

You usually do not reconcile ERP and MES IDs by forcing one system to adopt the other system's identifiers. The practical approach is controlled cross-reference mapping, clear system-of-record rules, and governed event flows. Success depends on master data quality, integration design, change control, and traceability across a mixed brownfield stack.

Start with a small set of supplier KPIs that are operationally meaningful, consistently definable, and traceable to source data. In most regulated manufacturing environments, that means standardizing delivery, quality, and responsiveness first, not building a large scorecard. The limiting factor is usually data definitions, event timing, and system integration quality across ERP, QMS, MES, and supplier portals.