System audits and product inspections answer different questions, and sourcing programs fail when one is asked to do the other's job. The audit previews whether factory controls can hold across production; the inspection verifies whether the goods in this lot meet specification. Buyers who schedule each at its natural rhythm spend less and learn more than those who overuse either.
Key takeawaysAudits judge whether the factory can hold a process; inspections judge whether this lot meets specification. · Schedule audits around qualification and change, and inspections around shipments where failure costs exceed check costs. · Sampling verdicts bound residual risk at an agreed level rather than certifying each unit [1]. · Acceptance discipline mirrors factory acceptance practice: demonstrate conformance with evidence before release [2]. · Link the tools by feeding lot defect data into audit scope and audit confidence into inspection intensity.
This is part 3 of a series comparing system audits with product inspections. Earlier parts established the basic split between capability evidence and conformity evidence; here we turn that split into scheduling decisions buyers can apply order by order.
Confusion between the two tools wastes money in both directions. Some buyers audit annually but skip lot checks, then absorb full containers of defects a capable factory produced on a bad week. Others inspect every shipment from a proven supplier at maximum intensity, paying for assurance the audit record already supports at lower cost.
The efficient program uses each tool for what it proves and links them through shared data. The sections below map timing, cost logic, and the feedback loop between factory verdicts and lot verdicts.
Two tools, two questions
A system audit samples months of records, observes processes, and interviews operators to judge whether controls operate repeatably. Its verdict speaks about the factory: capable, conditional, or structurally weak within the visited scope. That verdict ages as staff, lines, and workload change, which is why re-verification follows change rather than the calendar alone.
A product inspection measures the goods themselves against drawings, samples, and test criteria. The inspector draws a defined sample, records defects by class with photos tied to cartons, and issues a lot verdict. Its logic follows acceptance practice: conformance demonstrated with evidence before goods move, the same principle behind factory acceptance testing of equipment [2].
Because the questions differ, the evidence differs too. An audit file holds procedures, record samples, and corrective closures; an inspection file holds defect counts, measurements, and carton-tied photos. Buyers who want to approve a factory commission an independent supplier QMS audit independent supplier QMS audit for the first file, then build the second file lot by lot as shipments move.
Timing and cost logic
When each pays for itself
Audit when the decision is about the partner: initial qualification, new product lines, major process or site changes, and recurring defects that suggest drift rather than bad luck. The cost is justified by the switching economics, since discovering structural weakness before tooling or volume orders saves multiples of the visit price.
Inspect when the decision is about the goods: every shipment where the landed cost of defects, rework, chargebacks, or lost customers exceeds the check. Sampling keeps this affordable, because the formal logic of acceptance sampling lets a defined pull bound lot risk without opening every carton [1]. Skipping checks on high-exposure lots to save a small fee reverses the arithmetic.
Choosing the right tool
| Decision | Right tool | Why |
|---|---|---|
| Approve a new factory | System audit | Judges repeatable capability |
| Release this shipment | Product inspection | Measures goods vs specification |
| Diagnose recurring defects | Audit focused on failing process | Finds control breakdown |
| Settle a lot dispute | Re-inspection to same criteria | Evidence both sides accept |
| Reduce check intensity | Audit history plus clean lots | Confidence earns lighter sampling |
| After line or site change | Focused audit then lot checks | New conditions need both layers |
Linking audits and inspections
- Calibrate intensityLet audit confidence set sampling strictness: hold tighter checks on new or weak suppliers, standard levels on proven ones with clean histories.
- Feed defects backSend the defect breakdown from failed lots into the next audit scope, so the visit examines the processes behind real escapes.
- File both togetherKeep audit files and lot files side by side in the supplier record, making trends across orders visible at approval reviews.
- Re-audit on signalTrigger a focused audit when lot data drifts even if the calendar says the factory was approved, because conditions change faster than schedules.
The cheapest program is not the one with fewer checks but the one where each check answers the question it is suited to. Audits qualify partners; inspections release goods, and mixing those jobs wastes both budgets.
Sources
- Wikipedia: Acceptance sampling —— Overview of acceptance sampling including AQL concepts used in product inspection.
- Wikipedia: Factory acceptance test —— Overview of factory acceptance testing performed before shipment or installation.
Frequently asked questions
When should I order an audit versus an inspection?
Order the audit when qualifying or re-qualifying a factory, after major changes, or when recurring defects suggest process drift. Order inspection lot by lot whenever the commercial consequence of a bad shipment exceeds the inspection cost. The audit answers whether the factory can hold a process; the inspection answers whether this lot meets specification.
What is the fundamental difference between them?
Audits examine the management behind production across months of records, while inspections measure the goods in front of the inspector today. One judges capability and the other judges conformity, so neither substitutes for the other. Mature programs run both on different rhythms: audits periodically, inspections per shipment.
Can a sampled inspection really judge a full container?
Yes. Sampling lets a limited pull represent the lot at an agreed risk level, which is why inspection stays affordable at volume. The statistical reasoning behind inferring lot quality from samples is formalized in acceptance sampling theory. The verdict bounds residual risk; it never certifies every unit individually.
How do I combine both without doubling cost?
Keep both. Use audit evidence to set inspection intensity: strong systems with clean histories justify standard sampling, while weak systems or new suppliers justify tighter checks. Feed inspection defect data back into the next audit scope, so the system visit examines the processes behind real failures.