Internal Audit in Manufacturing: How to Run a QMS Audit That Actually Finds Problems
Need Custom Components?
Looking for a direct China factory for plastic or hardware parts?
Your Direct China Manufacturer
In-house production for custom plastic and hardware parts.
When we discuss internal audits with overseas buyers, one point comes up repeatedly: a factory that audits itself honestly is far easier to work with than one that only performs well when a customer walks through the door. The internal audit is where a manufacturer's quality system either proves itself or quietly breaks down. Over more than two decades of running plastic injection molding and hardware production in Zhejiang, we have seen audits done well and audits done only on paper, and the difference almost always comes down to three variables.
The Three Variables That Determine Internal Audit Outcomes
Before getting into process details, it helps to separate the factors that actually drive results from the noise. In our experience, the outcome of a QMS internal audit depends primarily on these three things, in this order of priority:

- Auditor independence and competence. An auditor who reports to the department being audited, or who lacks technical understanding of injection molding or hardware processing, will not find real problems. This is the single most common root cause of failed audits.
- Risk-based planning and scope definition. An audit that treats every process with equal depth wastes resources on stable areas while missing volatile ones — new molds, material changes, engineering changes, new operators, or recently shifted production lines.
- Evidence-based findings and closed-loop corrective action. Findings must be traceable to objective records (control charts, inspection logs, tooling maintenance records), and every non-conformity must lead to a corrective action that is verified for effectiveness, not simply marked "completed."
These three variables interact. A competent, independent auditor with a poorly scoped plan will still miss systemic issues. A well-scoped plan with weak evidence collection produces findings that cannot survive a customer's own audit. Below, we expand each variable and then walk through the practical process flow.
Variable One: Auditor Independence and Competence
ISO 9001 and equivalent quality frameworks require that internal auditors do not audit their own work. In a mid-sized factory, this creates a real constraint: the person who knows the molding process best is often the production supervisor. The practical solutions we apply and recommend evaluating in any supplier are:
- Cross-departmental auditing, where the quality team audits production and engineering, while engineering or production leads audit the quality system itself.
- Pairing a process expert with a systems auditor, so technical depth and audit methodology are both present.
- Training internal auditors on both the standard's requirements and the specific failure modes of their processes — for us, that means understanding cavity pressure variation, tooling wear drift, and dimensional capability, not just document control.
From a buyer's perspective, a useful verification question during supplier evaluation is simply: "Who performed your last internal audit, and what is their relationship to the departments audited?" If the answer is vague, the audit records deserve closer scrutiny.
Variable Two: Risk-Based Audit Planning

A manufacturing internal audit should not cover everything equally. Effective audit programs weight frequency and depth toward areas where failure probability and consequence are highest. For an injection molding and hardware operation, typical high-risk zones include:
- New tooling and first article stages, where process parameters are not yet stable and dimensional capability is unproven.
- Material and color changes, including supplier or lot changes for resins and metal stock, where incoming inspection rigor matters most.
- Engineering change implementation, where the risk is a split state — old and new revisions of a part or mold running in parallel.
- Outsourced processes (surface treatment, heat treatment, machining of hardware blanks), which are often audited less frequently than in-house steps despite carrying more risk.
- Calibration status of measurement equipment, especially gauges used for critical dimensions on customer drawings.
The audit schedule should be documented and justified. When we review audit programs, a schedule that audits every process once a year with identical checklists is a warning sign; it suggests the audit exists to satisfy a requirement rather than to control risk.
Variable Three: Evidence, Findings, and Closed-Loop Correction
The difference between a useful internal quality audit and a ceremonial one is almost entirely visible in the findings log. Three verification methods separate them:
- Traceability testing. Pick one shipped lot and trace it backward: from shipment record to final inspection report, to in-process checks, to material certificate, to molding parameter records. Any broken link is a finding. This single exercise exposes more systemic weakness than any document review.
- Horizontal deployment. When a non-conformity is found in one line or mold, check whether the same condition exists elsewhere. A worn locating ring on one tool often indicates a maintenance system problem, not a single-tool problem.
- Effectiveness verification of past corrective actions. Every prior CAPA should be re-tested during the next audit cycle. If the same category of non-conformity recurs, the original root cause analysis was wrong or the fix was never fully implemented.
The Internal Audit Process, Step by Step
For teams building or evaluating an internal audit process, the workflow below reflects standard practice in a manufacturing context. Each stage has documentation that buyers can request as objective evidence.
1. Annual Planning and Scope
The quality function maintains an audit program defining scope, frequency, and auditor assignments for the year, adjusted when triggers occur — customer complaints, major engineering changes, new tooling programs, or a lapse since the last audit of a given process. The plan itself is a controlled document.
2. Preparation
Before the audit, the auditor reviews the process's current documents, the last audit's findings and corrective actions, recent customer feedback, scrap and rework data, and open engineering changes. A checklist is prepared but treated as a tool, not a script — the most valuable findings usually come from following the actual flow of a real order rather than answering checklist questions.
3. Execution
On the floor, the audit combines interviews, observation, and record sampling. In a molding shop, observation means checking that the machine's recorded parameters match what is actually running, that the drying temperature and time for hygroscopic resins match the material data sheet, and that operators can explain what they do when a dimension drifts toward the control limit. In hardware processing, it means verifying that inspection frequency on critical dimensions matches the control plan and that measurement records show real variation, not identical repeated values — a classic sign of pencil-whipped inspection.
4. Findings Classification
Findings are classified by severity, which determines the response timeline. The classification matters because it drives urgency: a major non-conformity affecting shipped product quality requires containment and root cause analysis within days, while a minor documentation gap may be corrected in the next cycle.
5. Corrective Action and Closure
The responsible department performs root cause analysis, implements correction, and the auditor verifies effectiveness before closure. Effective closure requires evidence of the fix in operation, not just an updated document. For example, if the finding was missing torque checks on a hardware assembly, closure evidence should include dated inspection records over subsequent production runs, not a new instruction sheet alone.
6. Management Review Input
Audit results, trends, and unresolved findings feed into management review, which is where resource decisions — additional gauges, training, maintenance investment — get made. An audit program whose findings never influence resourcing is a strong indicator that the system is decorative.
Common Non-Conformities in Injection Molding and Hardware Operations
Certain findings recur across manufacturing internal audits. The table below summarizes the most frequent ones, how they are detected, and what appropriate verification looks like. Buyers can also use this as a line of questioning when assessing a supplier's self-audit credibility.
| Common Non-Conformity | Typical Detection Method | Verification / Acceptance Evidence |
|---|---|---|
| Molding parameters on machine do not match documented process sheet | Compare live machine settings against approved parameter card during audit walk-through | Parameter card matches running values within defined tolerance; deviations logged and authorized |
| First article inspection skipped after tooling repair or engineering change | Trace work orders following mold maintenance events; check FAI records | FAI report with critical dimensions dated after the change, signed by quality |
| Inspection records show no measurement variation (identical repeated values) | Statistical review of recorded measurements; observe actual gauge use | Records show realistic variation; operator demonstrates correct measurement technique |
| Expired or overdue calibration on gauges used for critical dimensions | Match gauge ID on inspection records against calibration register | All referenced gauges within calibration validity; overdue items quarantined |
| Old revision drawings or control plans in use at the line | Compare document revision at workstation against master list | Only current revision present; obsolete copies removed and recorded |
| Corrective actions closed without effectiveness verification | Sample closed CAPAs; check for recurrence of same failure category | No repeat finding within defined observation period; objective evidence attached |
| Outsourced processes (plating, heat treatment) not covered in audit scope | Review audit schedule coverage against process flow including subcontracted steps | Subcontractor controls, certificates, and incoming verification included in program |
How Buyers Should Use Internal Audit Evidence in Supplier Evaluation
For procurement managers and quality engineers assessing Chinese manufacturers, the internal audit is one of the most efficient windows into how a factory behaves when no customer is watching. Three practical suggestions from our side of the table:
- Ask for the last two audit reports, not just the latest. The trend tells you more than any single report. Improving findings depth year over year is a positive signal; identical boilerplate reports with zero major findings are not.
- Ask how findings were closed, and pick one to trace yourself. Choose a finding relevant to your product — tooling maintenance, if you are ordering molded parts; dimension control, if you are buying hardware components — and follow the paper trail from finding to root cause to verification evidence.
- Ask what changed as a result of the audit. A healthy system produces audits that occasionally trigger real investment: a new gauge, revised process limits, retraining, a redesigned fixture. Audits that never change anything are audits that never find anything.
Practical Risks and Easy-to-Miss Issues
Two mistakes we see most often, both in our own early years and in suppliers we have evaluated:
First, treating the audit as the quality department's job. When production and engineering teams view the internal audit as an external inspection to be survived, they prepare superficially and the audit finds little. The audit works when process owners use it as their own diagnostic tool — which is also why auditor selection and framing matter more than checklist length.
Second, over-documenting findings while under-verifying closure. A long findings list with weak closure discipline is worse than a shorter list with rigorous verification, because it creates the appearance of control. If you are building an audit program, resist the temptation to measure audit quality by the number of findings. Measure it by the recurrence rate of the same finding category across cycles — that number, tracked honestly, is the closest thing to a single health indicator a quality system has.
Closing Perspective
For a factory like ours, running general plastic components, tool accessories, and standard hardware parts for OEM and ODM customers, the internal audit is not a compliance ritual — it is the mechanism that catches drift before it reaches a customer's incoming inspection. Mold wear develops gradually, process parameters shift with material lots, and documentation discipline decays quietly. A well-run internal quality audit is the only routine process that checks all of these simultaneously and on record.
Buyers evaluating suppliers in China should weight internal audit maturity heavily, and should treat a factory's willingness to show unflattering audit findings as a trust signal rather than a red flag. A supplier that finds its own problems is a supplier that will find yours before they ship.
Recent Comments (0) 0
Leave a Reply