W. Edwards Deming articulated it in 1950: you cannot inspect quality into a product. The statement has been quoted at every quality conference for 75 years. It is cited in the opening paragraphs of quality management textbooks, referenced in ISO 9001 training materials, and repeated by quality professionals who then return to their organizations and invest primarily in inspection infrastructure.
The gap between knowing this principle and acting on it is one of the most persistent contradictions in industrial quality management. After 26 years working in aerospace, automotive, and composite manufacturing quality, I want to examine why the gap persists — and what the organizations that have actually closed it did differently.
1. Why Inspection Persists as the Primary Quality Response
The persistence of inspection-dominated quality systems is not irrational. It is a response to real organizational pressures that consistently outweigh the theoretical argument for prevention.
Inspection is visible and attributable. You can see the inspection department. You can count the inspectors. You can measure the defect detection rate and present it in a management review. When something goes wrong, you can point to the inspection system and ask why it did not catch the problem — and that question has a defensible answer. In organizations where quality budget decisions are made by leadership without manufacturing process experience, the inspection investment looks like quality assurance. The prevention investment looks like overhead.
Process control is invisible when it works. The investment in SPC, process FMEA, control plan maintenance, and operator competency produces an absence of defects — which is difficult to budget for, harder to defend in a cost reduction review, and essentially invisible to anyone who is not tracking the baseline from which improvement was achieved. The 1factory quality academy documented this challenge explicitly: in many organizations, Cpk and Ppk data exists at the manufacturing level but is not systematically used for process control decision-making at the management level.
Detection addresses yesterday’s problem immediately. When a customer calls with a quality failure, the first response is containment — find the affected parts, sort them, prevent further shipment. This is correct and necessary. The problem is that the urgency of the containment response consumes the resources and attention that should go to root cause analysis and process correction. The detection infrastructure deployed for containment remains in place because removing it creates visible risk before the process improvement that should replace it has been implemented and validated.
2. The Cost Structure Evidence
Armand Feigenbaum’s quality cost model — developed in the 1950s and validated across hundreds of industry applications since — establishes that defect prevention costs are typically 10-50 times lower than defect correction costs at the point of field failure, and 3-10 times lower than internal detection and rework costs. Feigenbaum identified four cost categories: prevention (process control, training, FMEA), appraisal (inspection, testing), internal failure (scrap, rework), and external failure (warranty, recall, liability).
In aerospace, the consequence structure makes the asymmetry more extreme than in most industries. The Datalyzer published analysis noted that „a Cpk of 4.00 — using only 25% of the spec width — resulted in 75% fewer warranty complaints“ for one automotive transmission manufacturer. The 1factory quality academy referenced the same case study. In aerospace, the consequence of a nonconformance that reaches an assembled aircraft — a fleet-wide inspection, a potential airworthiness directive, the reputational consequence of a safety event — makes the prevention investment argument overwhelming. Yet inspection-dominated quality systems persist throughout the aerospace supply base.
Paradygm’s analysis of IA9100 Clause 9.2 is relevant here: the new requirement elevates reviewing performance indicators from a note to a mandatory requirement during internal audits, „ensuring objective evaluation of QMS effectiveness.“ The shift from inspection-dominated to prevention-based quality is a Clause 9.2 auditable requirement under IA9100 — not just a management philosophy.
3. What Organizations That Have Made the Shift Did Differently
They quantified the cost of their current detection-based approach. Not in theory — in actual numbers. Total quality cost as a percentage of revenue: internal failure costs (rework, scrap, re-inspection), external failure costs (warranty, field service, customer concessions), and appraisal costs (inspection labour, testing equipment, incoming inspection). Industry benchmarks suggest that organizations with inspection-dominated quality systems typically have total quality costs of 10-20% of revenue. Organizations with mature prevention-based systems typically have total quality costs below 5%. The difference between those numbers, applied to your revenue, is the financial case for prevention investment.
They linked process control investment to specific failure modes. Rather than making a general argument for SPC and FMEA, they identified the three to five failure modes that were producing the most quality cost — not the most frequent failures, but the most costly — developed process control interventions targeted specifically at those failure modes, and measured the cost reduction achieved. The initial process control investment was narrow and measurable. It created organizational confidence in the approach before broader investment was requested.
They changed what first-line supervisors were accountable for. Process control requires supervisors to respond to control chart signals, stop processes drifting toward specification limits, and escalate conditions the process control system is detecting. That behavior requires supervisors measured on process stability indicators, not only on output volume and on-time delivery. Without changing the supervisor accountability structure — which is a management design decision, not a training outcome — process control documentation exists but is not acted on. This is Deming’s Point 12 („Remove barriers that rob people of pride of workmanship“) and Point 7 („Institute leadership“) in operational form.
3 Concrete Actions
1. Calculate your organization’s total quality cost — internal failure, external failure, and appraisal — as a percentage of revenue. If you have not done this calculation, it is worth doing before your next quality budget cycle. The calculation does not require specialist software — it requires cost accounting access and a half-day of time. The number it produces is the financial argument for prevention investment that resonates with leadership who think in financial terms rather than quality management terms.
2. Identify your top three failure modes by cost — not by frequency — and design process control interventions specifically for those failure modes. Frequency and cost are not the same measure. A failure mode that occurs rarely but produces significant scrap, rework, or customer impact may be a better target for process control investment than a failure mode that is frequent but cheap to address. The Feigenbaum model predicts that shifting investment from appraisal to prevention for your highest-cost failure modes will produce a measurable reduction in total quality cost within 12-18 months.
3. Review your first-line supervisor performance metrics. If supervisors are measured on output volume, on-time delivery, and labor efficiency — but not on process stability indicators, first-time quality rate, or SPC out-of-control response time — your process control system will never be fully implemented regardless of how good the technical documentation is. The accountability structure drives behavior. The behavior determines quality outcomes. Changing the accountability structure is the highest-leverage action available to organizations that want to shift from inspection-dominated to prevention-based quality systems.
👉 Request a Technical Consultation → www.skqc.org
Sources
- Deming, W.E. — Out of the Crisis (1986), MIT Press. 14 Points for Management: foundational reference for prevention-based quality thinking.
- Feigenbaum, A.V. — Total Quality Control (1951/1983), McGraw-Hill. Quality cost model: prevention, appraisal, internal failure, external failure.
- 1factory — A Guide to Process Capability (Cp, Cpk) and Process Performance (Pp, Ppk): 75% Fewer Warranty Complaints Case Study, 1factory.com
- Datalyzer — Process Capabilities Explained: Cp, Cpk, Pp and Ppk (August 20, 2025), datalyzer.com
- Paradygm — IA9100 Clause 9.2: Performance Indicators Now Mandatory in Internal Audits, paradygm.co
- AIAG — Statistical Process Control (SPC) Reference Manual, 2nd Edition. Automotive Industry Action Group.
- ISO 9001:2015 — Section 9.1: Monitoring, Measurement, Analysis and Evaluation
- EN 9100:2018 / AS9100 Rev D — Section 9.1: Monitoring, Measurement, Analysis and Evaluation
#QualityManagement #ProcessControl #Manufacturing #SPC #InspectionVsPrevention #Deming #Feigenbaum #SupplierQuality #SKQC