How should aerospace suppliers choose production KPIs?

Aerospace suppliers should choose production KPIs by starting with the operational decisions they need to control, not by copying a generic dashboard. A credible KPI set usually balances delivery performance, flow, quality, capacity, material readiness, and record completeness. In regulated aerospace work, a metric that improves output while hiding rework, escapes, missing evidence, or unstable processes is not a useful production KPI.

Start with the business and program constraint

The right KPIs depend on whether the supplier is managing rate increase, late backlog, high rework, unstable suppliers, constrained inspection capacity, skill shortages, or poor schedule adherence. A build-to-print machine shop, an assembly supplier, and an MRO operation will not need the same KPI mix.

Program and customer requirements also matter. Some customers may expect specific reporting around on-time delivery, quality escapes, corrective actions, first article status, or special process performance. Those expectations should inform the KPI set, but they should not be the only measures used to run the plant.

Use a balanced set, not one headline metric

Aerospace suppliers commonly need a small number of KPIs across these areas:

• Delivery and schedule control: on-time delivery, schedule adherence, planned-versus-actual completion, and recovery plan performance.
• Flow and constraint visibility: queue time, WIP aging, bottleneck utilization, traveler aging, and time waiting for inspection, material, tooling, or disposition.
• Quality performance: first pass yield, scrap, rework, defect recurrence, nonconformance rate, and cost of poor quality.
• Disposition and containment: MRB aging, corrective action aging, open nonconformances, and repeated defect families.
• Capacity and labor readiness: capacity load by work center, certified labor availability, overtime dependency, and constraint recovery.
• Material and kit readiness: shortages, late supplier parts, kit completeness, and material holds affecting released work.
• Traceability and record quality: missing signoffs, incomplete inspection evidence, late data entry, and record corrections requiring review.

This does not mean every supplier should track all of these at executive level. Some should be shop-floor control metrics, some should be program review metrics, and some should be quality system metrics. Mixing all of them into one score often creates noise instead of control.

Be careful with OEE and other generic manufacturing metrics

OEE can be useful in some repeatable, equipment-constrained operations, but it can mislead in high-mix, low-volume aerospace environments. A machine showing high utilization may still be producing the wrong priority work, building ahead of need, waiting on inspection, or generating rework. For many aerospace suppliers, flow, constraint recovery, schedule adherence, and quality burden are more actionable than a plant-wide OEE number.

The same caution applies to labor efficiency, pieces per hour, and utilization metrics. If they are used without context, they can encourage local optimization, excess WIP, hidden rework, or premature movement of incomplete jobs.

Define each KPI tightly

Each KPI should have a written definition before it is put on a dashboard. At minimum, define:

• the operational decision the KPI supports;
• the exact numerator and denominator;
• the data source and system of record;
• the refresh frequency;
• the owner responsible for action;
• the known exclusions or edge cases;
• the escalation path when the metric is out of range.

Without this discipline, two departments can use the same KPI name while measuring different things. That is common in brownfield environments where ERP, MES, PLM, QMS, spreadsheets, and maintenance systems all hold part of the truth.

Validate the data before trusting the dashboard

KPI selection is constrained by data readiness. If operators backfill labor after the shift, inspection results are stored in PDFs, nonconformances are managed outside the MES, or routings do not reflect the real process, the KPI may be directionally useful but not precise enough for management action.

In aerospace supply chains, replacing every legacy system just to improve KPI reporting is usually unrealistic. Qualification burden, validation cost, downtime risk, integration complexity, traceability obligations, change control, and long equipment lifecycles often make full replacement impractical. A better approach is usually to define the KPI model, identify systems of record, close the highest-risk data gaps, and integrate only where the decision value justifies the effort.

Avoid vanity metrics

Good production KPIs make problems visible early enough to act. Weak KPIs look positive while operations degrade. Common warning signs include:

• measuring output without measuring rework or escapes;
• measuring utilization without measuring schedule priority;
• measuring on-time delivery without tracking past-due recovery and customer-impacting misses;
• measuring quality only after final inspection;
• measuring averages that hide aging WIP, stuck jobs, or chronic bottlenecks;
• publishing dashboards that no role is accountable to act on.

The practical test is simple: if a KPI changes, does someone know what decision to make, what process to inspect, and what evidence to review? If not, the metric may still be interesting, but it is not a production control KPI.