Mistakes

Supplier Quality and Audit Mistakes That Wreck Your Numbers

The recurring errors that make supplier PPM, audit workload, and escape-cost numbers lie to you, and how to catch each one before it drives a bad decision.

The most common PPM error is mixing the defect count and the shipped quantity from different periods. Symptom: your Supplier Defect PPM swings from 400 to 4,000 month over month with no process change. Root cause is counting Q2 rejects against Q1 receipts, or counting rejected lots as units. Fix: lock both numbers to the same receiving window and count pieces, not lots. A supplier shipping 250,000 pieces with 300 defective units is 1,200 PPM. If you accidentally logged 30 rejected lots as 30 units against 250,000, you report 120 PPM and hide a real problem by a factor of 10.

Scorecards get gamed when weightings do not sum to 100 percent or when a zero-shipment month scores as perfect. Symptom: a supplier who shipped nothing shows a 100 quality score and climbs your ranking. Root cause is dividing by a shipped quantity of zero and defaulting the result to full marks. Fix in the Supplier Scorecard is to null out any metric with no denominator and flag the period as insufficient data. Also verify quality plus delivery plus cost plus responsiveness weights total exactly 1.00; a common slip is 40, 30, 20, and 20 summing to 110 and inflating every score.

Audit workload blows up because people forget travel and report-writing hours. Symptom: your Supplier Audit Workload says 3 auditors cover 120 sites a year, but the team is 40 percent behind by June. Root cause is budgeting only the 8 on-site hours and ignoring 6 to 10 hours of prep, travel, and closure per audit. Fix: load the real total. At 18 hours per audit fully loaded and 1,500 productive audit hours per auditor per year, one auditor delivers about 83 audits, not the 187 an 8-hour assumption implies. Three auditors cover 250 audits, not 560.

Escape cost gets understated by stopping at the scrap value of the part. Symptom: a defective 4 dollar sensor that escaped to final assembly is booked as a 4 dollar loss. Root cause is ignoring the 1-10-100 rule, where a defect caught at incoming costs 1, at assembly costs 10, and in the field costs 100. Fix: the Supplier Escape Cost should carry rework labor, teardown, containment, sort, expedite freight, and warranty. That same sensor pulled from a shipped unit can run 400 dollars once you add field service and a returned assembly, a 100x multiple the raw part price never shows.

Corrective action timing gets measured from the wrong start date. Symptom: your Supplier Corrective Action Cycle Time reports a clean 18 day average while operators complain the same defect keeps arriving. Root cause is starting the clock at 8D acceptance instead of defect detection, hiding 10 to 20 days of triage. Fix: measure from the containment date to verified effectiveness, not from when the supplier acknowledged the form. A CAPA that truly closed in 45 days but is logged as 18 tells you the loop is tight when 40 percent of your escapes are actually repeat failures inside the reporting gap.

Risk scoring fails when every input is self-reported and never revalidated. Symptom: a single-source supplier with a failing on-time record still shows a green Supplier Risk Score. Root cause is static financial and capacity inputs entered at onboarding two years ago and weighted equally with live quality data. Fix: refresh volatile inputs quarterly and weight single-source and sole-region exposure heavily. If 30 percent of spend sits with one supplier scoring 3 out of 5 on financial health, that concentration should push the composite into your top-risk tier, not average out against six low-risk vendors.

PPAP and incoming inspection get treated as fixed when they are demand-driven. Symptom: the incoming dock is a bottleneck every launch quarter and nobody predicted it. Root cause is planning Incoming Inspection Burden and PPAP Review Workload on an annual average instead of the arrival curve. Fix: model peaks. If 40 new part numbers land in one 6 week window at 4 review hours each, that is 160 hours against a team that averages 30 hours a week, a 5 week backlog. Development ROI suffers the same way when you count only prevented scrap and omit the freed inspection and audit hours, understating Supplier Development ROI by 20 to 40 percent.

Bad data upstream corrupts every downstream metric, so validate the source first. Symptom: two dashboards report different PPM for the same supplier in the same month. Root cause is one pulling from receiving inspection and the other from the MRB disposition log, which counts use-as-is and returns differently. Fix: pick one system of record, reconcile the piece counts once, and audit 10 random lots per quarter against physical records. A 5 percent miscount on a 2 million piece annual volume is 100,000 pieces of phantom or missing defects, enough to move a supplier a full tier and trigger an unwarranted audit or, worse, cancel a needed one.

Published 2026-07-01.