Supplier Audit
A supplier audit is not a scorecard ceremony. It is a comparison between what the KPI pack says and what the factory actually shows — on the floor, in WIP, in how problems are escalated, and in whether procedures exist only on paper.
The audit axis: what the KPI says vs what the factory shows
Reported performance can be numerically correct and still operationally misleading.
The useful audit question is not “Do they have KPIs?” It is: Can the plant demonstrate the same story the dashboard tells — and can it show how problems become visible and get owned?
Selection and ramping logic belong on Supplier Evaluation and Selection. Continuity structure belongs on Supplier Risk and Supply Continuity. This page stays on verification and culture.
Audit signals I treat as red flags
These are concrete mismatches I have learned to take seriously. They are diagnostic signals from practice — not a universal audit checklist or a scoring model.
Strong reported OEE with WIP that reveals the bottleneck
A strong reported OEE can look reassuring while work-in-process tells a different story: material piled before a constrained resource, starvation after it, and “efficiency” achieved by running the wrong machine hard. The KPI says availability and performance; the floor shows where the system actually breaks.
Very high on-time delivery maintained by overtime and heroic shipments
Very high reported on-time delivery can look excellent until you ask how it was protected. If the answer is structural overtime, weekend recovery, and individual heroics, the KPI is reporting survival method — not stable process capability. Continuity risk is still present; it is only hidden inside the delivery percentage.
Zero customer complaints with internal rework absorbing the defects
A clean external complaint record can coexist with heavy internal rework, sorting, and quiet containment. The customer did not complain because the defect never left — which is better than shipping bad product, but it is not the same as a process that does not generate the defect. Audit must ask where quality cost and delay actually live.
ISO procedures people do not use
Documented procedures that operators and supervisors do not actually follow are not control. They are theatre. The audit question is whether the written system matches the lived system — who updates it, who trains it, and what happens on the shift when the documented path is slower than the informal one.
Problem-solving culture is the stability test
The suppliers that looked most stable in my work were not the ones who claimed to have no problems.
They were the ones who could show:
- where problems typically appear
- who owns them when they appear
- how the system makes them visible early enough to act
A stable supplier is not the one without problems. A stable supplier knows where the problems are, has people who solve them, and has a system that makes them visible.
That culture test belongs in audit conversations and plant walks. It cannot be fully replaced by a vendor scorecard that only averages monthly percentages.
How audit feeds the rest of the model
Audit findings should change allocation, development priorities and continuity design — not sit in a PDF after the visit.
- If the floor contradicts the KPI story, slow the volume ramp described on the evaluation page.
- If on-time is heroic rather than systemic, raise continuity risk even when deliveries still arrive.
- If problems are invisible inside the supplier’s own system, commercial “partnership” language does not reduce exposure.
Cost comparison across two supply models remains a separate discipline on TCO / Comparative Supply Analysis. Audit answers whether the operating system behind either model is real.