The audits stack. The assurance doesn't.

Here is an uncomfortable observation from two decades in automotive and aerospace quality: the number of audits a plant hosts tracks its customers' distrust, not its defect rate. I have seen a supplier shipping single-digit PPM absorb eleven external audits in a year. A mediocre plant with one trusting customer hosted two. So when Vogue Business ran a piece on fashion's audit fatigue – certification visits, brand-specific codes and social-compliance regimes stacked onto the same factories, none accepting another's evidence – I read it with flat recognition. Not fashion's problem. Tuesday in Tier 1 automotive. And with every EV and battery plant launching in Europe right now, a fresh cohort of nervous customers is discovering the same arithmetic.

How the stack gets built

Nobody designs the stack. It accretes. First comes the certification body, because IATF 16949 or ISO 9001 is the entry ticket. Then each OEM bolts on customer-specific requirements and sends supplier quality to verify them in person, every year. German customers add VDA 6.3 process audits – deep, legitimate, and repeated per customer, because a 6.3 run for one OEM is worthless to the next. Now the market-access regimes are arriving on top: due-diligence rules, battery-passport verification, sustainability audits, even forest certification on the packaging. Each new European regime brings its own audit annex and its own evidence format.

Every layer arrives justified as risk. Not one accepts another's findings, and that is the tell. If assurance were the objective, a credible VDA 6.3 report would be gold for every customer of the plant. It isn't, because the audit is leverage as much as assurance: a visit generates findings, findings create escalation rights, and escalation rights feed price negotiations and sourcing decisions.

An audit whose result only one customer will accept is not assurance – it is property.

What fatigue actually does to a plant

Audit fatigue never appears on a spreadsheet labelled fatigue. It appears as theatre. Two weeks before a major customer audit, the best engineers come off the line to build evidence packs, refresh whiteboards and brief operators on what to say. The organisation gets better at audits. Not at parts.

Then the recycling starts. The same root cause – usually organisational: unclear ownership, no reaction plan at the line, a PFMEA that never left the office – gets written up under one code by the certification body, another by the VDA auditor, a third by the customer's supplier-quality engineer. Three findings, three 8Ds, three closure battles. A finding raised three times under three codes is not a finding. It is a subscription.

Count the cost honestly. A two-day process audit, once preparation, walkthroughs and corrective-action writing are in, consumes 15–20 engineer-days at the plant. At a fully loaded €900 a day, a modest stack of ten external audits burns €135,000–180,000 of engineering capacity a year – capacity that was hired to prevent defects. I led quality initiatives across a multi-site, 2,000-plus workforce where certification, VDA 6.3 and customer audits competed for the same handful of engineers. The audit calendar had quietly become our second production system.

Fewer owners, not fewer audits

The cure is not fewer audits. It is fewer audit owners. IATF proved the logic decades ago: one rulebook, one third-party certificate, every OEM accepting it as the baseline instead of each running a private regime. Extend that downstream. One deep, credible process audit a year – a proper VDA 6.3 by a competent third party – with the report and findings shared among every customer of the plant, and closure verified once, jointly, instead of three times separately.

Underneath that, the plant owns its own layered process audits: daily checks by team leaders, weekly by area managers, monthly by site leadership, all tracked to closure. I built exactly that for a 900-plus-person greenfield plant – the QA/QC department and the internal audit programme were in place before the first customer auditor walked through the door. External audits became boring sampling of a living system rather than an annual examination. Defect costs came down 70 per cent, and the reduction tracked the layered audits, not the visitors.

Customers resist for two reasons. A shared audit is a surrendered lever. And the audit-day machinery – supplier-quality headcount, auditor fees, scorecards – justifies itself through frequency. Behind both sits an honest fear: that the shared auditor will miss what we would have caught. The answer is governance, not repetition. Agree the auditor's competence profile once, audit the auditor, then stop re-running the same audit ten times.

Escalations make it worse, and I know that machinery from both sides of the table – auditee under IATF 16949, AS9100/EN 9100 and VDA 6.3, later the quality executive auditing my own suppliers. After an escalation, the stack triples: extra visits, narrower scopes, weekly calls dressed as audits. I have taken a plant through a quarter with zero critical customer escalations, and it never happened by accepting more auditors. It happened by negotiating depth instead of frequency – one jointly owned verification of the real process, closure evidence signed by the customer's own engineers, then a staged return to the normal regime. Trust rebuilt is audits retired.

One audit worth trusting beats ten nobody believes. The plant with the most audits rarely has the worst process – it has the newest customer. Audit fatigue is not a workload problem. It is a trust problem, and nobody ever solved a trust problem with scheduling. As the stack spreads from fashion into every EU market-access regime – batteries, due diligence, sustainability annexes – buyers who treat assurance as private property will pay for it in supplier churn and findings that never close. The suppliers who outlast the next decade will be the ones whose customers learned to own assurance together instead of renting it separately.