Why good plants still fail audits
Most Indian manufacturers who lose points in an IATF 16949 or ISO 9001 surveillance audit are not sloppy operations. They make good parts, they employ capable people, and they genuinely care about quality. What sinks them is rarely a dramatic failure — it is a handful of quiet, systemic gaps that repeat across the industry, each one a workflow that was never made routine. This guide names seven of them, and for each shows the workflow that prevents it.
The pattern behind all seven is the same: quality treated as a set of documents produced for the audit rather than records generated by daily work. When the calibration register, the PPAP package, the NCR and the FMEA are things someone updates before the auditor arrives, they drift out of reality between audits — and the auditor’s job is precisely to find that drift.
1. Uncalibrated gauges still in use
The most common — and most serious — mistake is measuring parts with gauges that are overdue for calibration. It undermines everything downstream: if the gauge cannot be trusted, neither can a single inspection reading it produced, which means your accept/reject decisions rest on nothing. IATF 16949 and ISO 9001 clause 7.1.5 require calibration against traceable standards on a defined frequency, and an auditor will physically pick a gauge off the floor and ask for its calibration record.
The fix: a gauge register that derives a next-due date for every instrument and a calibration follow-up that alerts before a gauge falls due — not a spreadsheet someone remembers to check. When a gauge does fail calibration, every measurement it produced must be flagged for review. In Fast Quality, calibration follow-up drives the due/overdue alerts and the gauge MIS is the audit-ready register; a failed gauge is quarantined and its suspect measurements recalled.
2. Treating PPAP as a one-time event
Many Tier-2 suppliers assemble a PPAP package once, get it approved, and never touch it again — even as the process, tooling, sub-supplier or material quietly change over the years. This is a mistake because PPAP approval is valid for the process that was submitted. A significant change without a re-submission means the parts you are shipping today were never actually approved, and a customer audit that finds the mismatch is a serious finding.
The fix: tie PPAP to change management, so any significant engineering change, process change or deviation automatically raises the question “does this need a re-submission?” A document-backed PPAP package — where every element is a controlled document with a version and approval status — makes re-submission a controlled update rather than rebuilding a folder from scratch. In Fast Quality, an approved change triggers a PPAP re-submission where required, keeping the package current with the process it describes.
3. Closing NCRs without root cause
An NCR closed with “material re-inspected, found OK” or “operator advised to be careful” is not closed — it is a defect on a snooze timer. Re-inspection is containment; counselling is not a corrective action; and neither removes the cause, so the defect returns. Auditors are trained to sample closed corrective actions and probe exactly this: “show me the root cause, and show me the evidence it did not recur.”
The fix: escalate significant or recurring NCRs into a real 8D, where D4 demands an actual root cause and D5 a permanent corrective action, and hold the team to it. A defect that appears a third time on the Pareto has earned an 8D by definition. See the full workflow in NCR and non-conformance management and reducing rejection and rework. In Fast Quality, a major NCR escalates into an 8D with CAPA rather than a quiet close.
4. Never re-scoring the FMEA after CAPA
The FMEA is meant to be a living risk analysis, but in practice most are frozen at launch. A corrective action fixes a real failure on the line, and the FMEA that predicted — or missed — that failure is never updated. The result is a risk document that lies: it shows a low RPN for a failure mode you have just proven happens, or it never listed the failure at all. When an auditor cross-references a closed 8D against the FMEA and finds no change, the finding writes itself.
The fix: make FMEA re-scoring a mandatory output of every CAPA. The failure mode’s occurrence (and sometimes detection) score is revisited after the permanent action, the special characteristic and control plan are checked, and the change is recorded. This is the loop that keeps the FMEA honest. In Fast Quality, CAPA from an 8D feeds an FMEA RPN re-evaluation through change management, so the risk analysis reflects reality rather than the day of launch.
5. Documents outside control
The classic floor finding: an operator working to a drawing, control plan or work instruction that is not the current revision — because the controlled copy lives in one place and the floor copy is a photocopy taken months ago. Uncontrolled documents mean parts made to superseded specifications, and ISO 9001 clause 7.5 exists precisely to prevent it.
The fix: a real document-control system where drawings, control plans, FMEAs and work instructions are versioned, moved through an approval lifecycle, attached to the item record, and where obsolete revisions are withdrawn so the floor can only reach the current one. See the full treatment in the ISO 9001 document control guide. In Fast Quality, quality documents live in the shared document-control subsystem with an approval lifecycle, so the version on the item record is the version in force.
6. Free-text rejection data
A plant can record every rejection diligently and still learn nothing from it, if the rejection reason is free text. “NG”, “bad”, “rejected by QC” and “not ok” cannot be counted, ranked or Pareto-charted, so the single most valuable use of rejection data — knowing which defect to attack first — is lost. This is a self-inflicted blindness that no amount of diligent logging cures.
The fix: a controlled defect-code catalogue and the discipline of tagging every rejection from it, mapped to the work centre where the defect arises. That single change turns a rejection log into a rejection Pareto — see quality reports and KPIs. In Fast Quality, shared defect masters give every rejection one coded language across incoming, in-process and final rejection.
How many of these seven are live in your plant right now?
See calibration alerts, PPAP tied to change, real 8D root cause, living FMEAs and controlled documents — the workflows that prevent all seven — in 30 minutes.
7. Quality that lives in spreadsheets
Underneath the other six sits the root mistake: running the whole quality system on disconnected spreadsheets and paper registers. The FMEA is in one file, the control plan in another, calibration in a third, inspection on paper, rejections in a book. Nothing joins up, so a failure mode in the FMEA cannot be traced to a defect on the line to an 8D to a control-plan change — the very traceability the standard demands. And every audit becomes a frantic reconciliation because the “system” is really a dozen files maintained by hand.
The fix: one connected system where the specification defines “good”, inspection records against it, defect codes link rejection to 8D to FMEA, and controlled documents carry PPAP — so the records reconcile and the history is traceable by design. This is the entire argument for a real QMS, laid out in the pillar guide on what quality management software is. The other six mistakes are all symptoms of this one; fix the disconnection and they become hard to make.
| Mistake | What it costs | The workflow that prevents it |
|---|---|---|
| Uncalibrated gauges in use | Every reading untrustworthy | Gauge register + calibration due-date alerts |
| PPAP as a one-time event | Shipping unapproved parts | PPAP tied to change management & re-submission |
| NCRs closed without cause | Defects that keep returning | Escalation to a real 8D with D4/D5 |
| FMEA never re-scored | A risk analysis that lies | CAPA drives FMEA RPN re-evaluation |
| Documents out of control | Wrong revision on the floor | Versioned document control (clause 7.5) |
| Free-text rejection data | No Pareto, no priorities | Controlled defect-code catalogue |
| Quality in spreadsheets | No traceability, audit scramble | One connected QMS |
Five findings that never had to happen
A Tier-2 supplier prepares for its surveillance audit the usual way — a fortnight of updating registers. The auditor picks a micrometer off the floor: calibration overdue by six weeks. They cross-reference a closed 8D against the PFMEA: no re-scoring. They find an operator using a control plan two revisions old. They ask for the rejection Pareto: there is only a free-text log. And they note the PPAP package predates a tooling change made last year. Five findings, all avoidable, and all symptoms of quality maintained for audits rather than by workflow. Had the calibration alerts, the CAPA-to-FMEA loop, controlled documents, defect codes and change-linked PPAP simply been how the plant worked day to day, there would have been nothing to find.
How Fast Quality prevents each
Fast Quality Software, built by Improsys in Pune on the shared Fast Suite platform, is designed so these seven mistakes are hard to make:
For the connected system that makes all of this the default, see the pillar guide on what quality management software is.
Frequently asked questions
What is the most common quality management mistake?
Using gauges that are overdue for calibration. It is the most serious because it undermines everything downstream — if the gauge cannot be trusted, no inspection reading it produced can be either, so every accept/reject decision rests on nothing. IATF 16949 and ISO 9001 clause 7.1.5 require calibration against traceable standards on a defined frequency, and auditors routinely pick a gauge off the floor and ask for its record. The fix is a gauge register with next-due dates and alerts that fire before a gauge falls due, plus recall of suspect measurements when a gauge fails.
Why is treating PPAP as a one-time event a mistake?
Because PPAP approval is valid only for the process that was submitted. If the process, tooling, sub-supplier or material changes and no re-submission is made, the parts you are shipping today were never actually approved — and a customer audit that finds the mismatch treats it as a serious finding. The fix is to tie PPAP to change management so any significant change raises the re-submission question automatically, and to keep the package as controlled documents with versions and approval status so a re-submission is a controlled update rather than rebuilding a folder.
Why isn’t “re-inspected, found OK” a valid NCR closure?
Because re-inspection is containment, not corrective action — it sorts the current stock but does nothing about the process that keeps producing the defect, so the defect returns. Auditors specifically sample closed corrective actions and ask for the root cause and evidence of non-recurrence. Significant or recurring NCRs should escalate into a real 8D where D4 demands an actual root cause and D5 a permanent corrective action. A defect that appears a third time on the Pareto has earned an 8D by definition.
Why must the FMEA be re-scored after a corrective action?
Because an FMEA is meant to be a living risk analysis, and a corrective action changes the risk it describes. If a CAPA fixes a real failure but the FMEA is never updated, the document lies — showing a low RPN for a failure you have proven happens, or never listing it at all. When an auditor cross-references a closed 8D against the FMEA and finds no change, that is a finding. Making FMEA re-scoring a mandatory output of every CAPA — revisiting occurrence and detection, checking the special characteristic and control plan — keeps the analysis honest.
What ties all these mistakes together?
Running quality on disconnected spreadsheets and paper registers instead of one connected system. When the FMEA, control plan, calibration, inspection and rejections all live in separate files, nothing joins up, so a failure mode cannot be traced to a defect to an 8D to a control-plan change — the traceability the standard demands — and every audit becomes a reconciliation scramble. The other six mistakes are symptoms of this one. Fix the disconnection with a QMS where the specification defines good, inspection records against it, defect codes link rejection to 8D to FMEA, and controlled documents carry PPAP, and the individual mistakes become hard to make.