Here is an uncomfortable observation: we have industrialised the blaming of quality managers for failures their org chart forbids them to prevent. A food-safety journal has just run a piece classifying accountability without authority as a structural psychosocial hazard – an occupational risk to the person holding the job. In food manufacturing, apparently, this counts as news. In discrete manufacturing it is called Tuesday.

The timing is not accidental. The 2026 revision of ISO 9001 pushes quality into strategic roles at the exact moment quality talent is walking out of the profession, while the trade press fills up with platforms promising a move from compliance to prediction. Buy whatever software you like. No model predicts its way past an org chart that routes every real decision to somebody else's desk. Role design stopped being an HR footnote the moment the standard asked quality to sit at the strategy table.

The org chart guarantees the escape

Map where authority actually sits. Production owns the schedule and every hour on the line. Procurement owns the supplier, and the award decision is priced in euros per piece. Finance owns headcount, capital and containment spend. Plant management owns, in practice, what ships. And quality? Quality owns the metric – escapes, cost of poor quality, audit findings – and the accountability welded to it.

Now walk one escape through that structure. A supplier ships nonconforming stampings. Your incoming inspector flags them. Containment needs people: finance signs off the overtime. Rework needs line time: production decides whether the schedule can afford it. Sourcing elsewhere: procurement explains the tooling is amortised and the next lot will be fine. Every decision point in the causal chain has a name, a desk and a reporting line, and not one of them reports to quality. Then the part escapes anyway, the customer catches it, and the scorecard asks the quality manager to explain. I have read escape reports where quality's only contribution to the failure was noticing it.

A KPI you are measured on but cannot move is not a target. It is a bill with your name pre-printed on it.

What it costs – in people first, then in euros

The food-sector authors did us a courtesy by printing what the shop floor has known for decades. The mechanics transfer exactly. Dashboards quietly massaged so the red never reaches the plant manager. Quality engineers sandbagging KPIs because the honest number ends careers, theirs included. And the silent resignations: the people who understood how the plant really worked, leaving without an exit interview worth the name, because an honest one would have named their bosses.

Put prices on it. Replacing one senior quality engineer runs north of €100,000 all-in by the time you have paid recruitment, onboarding and six months of half-competence before they close their first 8D. One de-escalated customer kept is worth more than that engineer's annual salary; one escalated customer lost is a contract. An 8D that was theatre – containment signed, root cause invented, the same nonconformance back inside a quarter – means paying for the same failure twice, plus the audit fee. We audit cost of poor quality religiously. We almost never audit the cost of the role design that produced it.

Pair every accountability with a switch

The fix is structural, and none of it is sentimental.

  • Stop-ship criteria that belong to quality. Named, written into the QMS, agreed in daylight – not renegotiated at 02:00 with a truck idling at the dock. Authority that exists only in a job description does not exist at 02:00.
  • Deviation authority that is timeboxed, not negotiated. Quality answers within 24 hours, yes or no, and production plans around the clock. Speed gets designed in instead of extracted under pressure.
  • Decisions pushed to the defect. At Witte Automotive we ran QRQC with stop-and-fix authority in the hands of the people standing next to the machine, backed by A3 and a weekly Q-Wall. That arrangement drove the substantial failure-cost reduction we booked – not because quality got stronger, but because the decision moved to where the information lived.
  • An escalation path built to de-escalate. At SNOP I designed the greenfield QA/QC organisation for a 900-plus-employee plant, and we drew the authority lines before hire number one: who could stop the line, who could release under deviation, who owned the customer call. The first quarter closed with zero critical customer escalations. When authority travels down with the accountability, escalation becomes what it should be – rare.

None of this weakens production, procurement or finance. It removes the fiction that quality can be answerable for outcomes it is not permitted to touch.

The audit question

Run it on yourself before an auditor does.

Key takeaways

  • List your accountabilities, then the decisions you can make alone. The gap between the two lists is your next escape – and your next resignation.
  • Write stop-ship criteria into the QMS with names attached; authority without a named owner is a rumour, not a control.
  • Timebox deviation authority so speed is a design parameter, not a bargaining chip extracted at the dock door.
  • Move decisions to the defect. QRQC-style stop-and-fix on the floor beats escalation upstairs on cost, on speed and on the people doing the work.

Escapes will still happen. Complex systems fail, and I have spent two decades in automotive and aerospace learning how. But the next time one lands on your desk, open the 8D template only after you have asked who in the structure could have said no – and checked where they sit. If the answer is not quality, the finding is not against the quality manager. It is against the org chart. And unlike a person, an org chart can be redrawn.