Audit room 1 — The paper room
Supplier A. I asked for the IATF 16949 certificate scope page, not the cover. A cover page landed in 5 minutes. The scope page arrived an hour later — and it named the holding company but did not name the production line. Because many Chinese groups hold a group-level certificate and run the trim line under a different legal entity, the scope page is where this audit either gives the supplier a pass or sets it aside.
Supplier B. Scope page landed alongside the cover. Same legal entity, same address. Production line at the same address as the cert.
My call. Supplier B survives Audit Room 1 with no observations. Supplier A goes back to the workshop with a finding requiring the scope page to be re-issued under the trim-line legal entity. IATF oversight office records the certification, and on B I confirmed the cert registration matched the supplier’s address.
Field benchmark on the paper room. Three patterns show up. Pattern one: the scope page lands within the hour, names the production line at the same address as the cert, and the cert registration in the IATF database matches the supplier’s business license — this is a managed QMS. Pattern two: the scope page lands within the hour but names the holding company rather than the production line — this is a group certificate that covers a different legal entity than the one running the program. Pattern three: the scope page does not arrive on the day of the visit — this is a supplier whose QMS documentation is held at the holding company level and is not produced at the production line. The paper room is the cheapest of the four to audit and the most predictive of the other three.
Why the cover page is not enough. The cover page is marketing. It says the supplier holds an IATF 16949 certificate. The scope page says what the certificate covers. A group-level certificate covers the holding company; a production-line certificate covers the line. A Class-A interior trim program needs the production line covered, not the holding company. The audit room catches the difference in under an hour, and the buyer’s ability to read the scope page in real time is the test.
Audit room 2 — The machine room
Supplier A. The CMM bay held a single manual bench CMM with no certificate on the wall. The tryout press list was last updated in February. The QA team said “we can schedule a tryout next week”.
Supplier B. The CMM bay held a three-axis CMM with a calibration sticker dated within the last 12 months, a probe-qualification log, and a fixture-recall schedule for the fixture sitting on the table. The tryout press held a T2 sample on the bed when I walked in. A live injection record from the day’s T2 cycle sat on the bench laptop.
My call. On Class-A door and IP trim programs, dimensional control and live tryout history are not negotiable. Supplier A cannot run a Class-A program today; B can. I let B’s QA — run by a 60+ member team at our facility — walk me through the control plan on a real IP trim part.
Field benchmark on the machine room. Three patterns show up. Pattern one: a three-axis CMM with calibration sticker dated within 12 months, a probe-qualification log on the wall, a fixture-recall schedule for the fixture on the table, and a live tryout cycle running on the press with today’s date on the record — this is a Class-A-capable machine room. Pattern two: a single bench CMM with no calibration sticker, no probe log, and a tryout press list last updated several months ago — this is a machine room that runs Class-A work in theory but cannot demonstrate it on the day. Pattern three: a CMM bay that exists on the floor plan but is empty or repurposed — this is a machine room that has been downsized or removed, and the supplier should be asked directly whether the program will run on the documented equipment or on a different setup.
Why live tryout history is the test. A live tryout cycle from the day of the audit tells the buyer that the press is operational and the cycle parameters are stable. A tryout press list last updated in February tells the buyer that the cycle parameters have drifted and the supplier is not running daily cycles on the equipment. For a Class-A program where dimensional control is the primary failure mode, live tryout history is the strongest signal that the equipment is producing Class-A today, not in the past. Ask the QA team to print the day’s cycle record and walk you through the parameter values — the answer reveals the maturity in five minutes.
Audit room 3 — The surface room
Supplier A. Could not produce a recent waviness scan from a Class-A trial. The cavity was described as “hand-polished to A-2.” Steel grade unspecified. No aluminium disclaimer — but the cavity impression suggested softer tool steel than P20.
Supplier B. I asked for a waviness scan printed from a recent Class-A trial and the steel-grade chart. Both came out of a folder within ten minutes: waviness ≤ 0.2 mm over 100 mm; orange peel ≤ 3.0 wave-scan units. Cavity and core in P20 with NaX treatment for high polish. Sliders and lifters in H13. Because Class-A is a functional spec, the supplier has to measure it on the cycle, not just describe it in the quote.
My call. Surface room is where most Class-A failures land during production. Supplier B passed it cleanly; Supplier A would have shipped cosmetic reject rates of 4–7% on a door-panel program, which is the failure mode I would have caught only at PPAP. Because cosmetic reject rates below 2% require waviness control in the cavity, the surface room is also the test of whether the polish discipline survives the program life-cycle.
Field benchmark on the surface room. Three patterns show up. Pattern one: a waviness scan printed from a recent Class-A trial, with values within spec (Wa ≤ 0.2 mm over 100 mm, orange peel ≤ 3.0 wave-scan units), and the steel-grade chart on the wall — this is a surface room that measures cosmetic performance on a cycle. Pattern two: a sample with hand-polished cavity and no waviness scan, with the steel grade described but not documented — this is a surface room that produces cosmetic parts in the quote but cannot measure the cosmetic performance on the cycle. Pattern three: no waviness scan and no steel-grade chart, with the surface described as “to customer spec” — this is a surface room that runs cosmetic work blind, and the buyer should walk away.
Why waviness and orange peel are the two measurements. Waviness (Wa) is the long-wave deviation on the cavity surface that produces read-through under typical paint stack-up on a Class-A face. Orange peel is the short-wave surface texture that produces visible roughness on the painted surface under direct light. Both are measurable on a waviness scan, and both are controllable on a polished P20 cavity with NaX treatment. Suppliers who measure both on a cycle are running Class-A work; suppliers who describe both in the quote are quoting Class-A work but not running it.
Audit room 4 — The trace room
Supplier A. Walked to a single mold; asked to scan its serial and pull its full cavity-insert grade, heat lot, T-cycle history, FAI signatures and heat-treat logs. The supplier produced a paper folder. Printouts from a binder.
Supplier B. Walked to a single mold; its serial brought up heat lot, every cavity-insert grade, every T-cycle, every FAI signature on one screen in 90 seconds. Linked to the OEM customer (BYD tier-1) and the part number.
My call. I treat the paper binder as a fatal finding. Interior trim is safety-adjacent: a failed retainer clip on an airbag-door assembly is a recall risk. Because traceability must be a database for any recall-class part, B passes the trace room and A does not. Because the database is only as good as the data it ingests, the supplier’s traceability discipline has to start at the cavity heat lot, not the surface finish.
Field benchmark on the trace room. Three patterns show up. Pattern one: a mold serial brings up heat lot, every cavity-insert grade, every T-cycle, every FAI signature, and the OEM customer with the part number — all on a single screen in under two minutes. This is a database-driven trace room that can survive a recall in real time. Pattern two: a mold serial brings up the cavity heat lot and the most recent T-cycle, but the older T-cycles are in a binder and the OEM customer mapping is incomplete — this is a trace room that is partially on a database and partially on paper. Pattern three: a paper folder with printouts from a binder — this is a trace room that will not survive a recall in real time, and the recall cost will be measured in weeks of buyer-side effort plus weeks of supplier-side effort to reconstruct the timeline.
Why the trace room is the recall test. Interior trim is safety-adjacent because a failed retainer clip on an airbag-door assembly is a recall class defect. A recall in the EU takes 6-12 months to coordinate between the OEM, the Tier 1, the Tier 2, and the regulatory body; a recall in the US takes 3-9 months. The supplier that can produce the heat lot, the cavity grade, and the T-cycle history in two minutes compresses the recall coordination to days; the supplier with a paper folder compresses it to weeks. The cost difference between days and weeks on a recall coordination is measured in millions of dollars on a typical interior trim program.
Letter close — what I would sign
If I were the Tier 1 buyer, I would shortlist B and walk A back to its financial controller with the paper-room observation only. A Class-A trim program almost never collapses on Audits 1 or 2; it collapses on the surface and trace disciplines that only show up on a real shop floor. Because I have watched supplier A win on price repeatedly and lose the program on T3–T5 delays, I would not sign a PPAP-linked commitment with A based on its current state. I would sign with B, with an independent DFM review at T1 as an external challenge on the 2K trim program where the housing meets the soft-touch armrest.
If you want the industry-side view on Class-A cosmetic reject budgets and PPAP timing, the AIAG quality portal is the place to confirm current thresholds. Our trim program’s audited supplier — B — is the company I work for: 5,000 m² precision tool-room inside the 68,000 m² footprint at No.188 Tanjialing East Road, Yuyao. The audited competitor — A — is anonymized deliberately. The point isn’t to name competitors; it’s to make the rooms legible.
Why the four-room structure works as a buyer-side checklist. The four rooms (paper, machine, surface, trace) cover the four operational disciplines that determine Class-A trim outcome. The paper room is the QMS gate; the machine room is the dimensional control gate; the surface room is the cosmetic control gate; the trace room is the recall gate. Each room has a five-minute audit. The full audit takes 30-45 minutes on the floor, plus 30 minutes reviewing the documents. A buyer who runs all four rooms on every supplier visit builds a comparable dataset across the supply base within a quarter; a buyer who runs only the paper room and the machine room is missing the two rooms where Class-A actually fails.
What the buyer should walk away with. For each supplier, the audit produces four binary outcomes (pass / fail with finding). A supplier who passes all four is a Class-A candidate today. A supplier who passes three and fails one is a Class-A candidate after the finding is closed — the close should happen on a documented 30-day cadence with a follow-up visit. A supplier who passes two or fewer is not a Class-A candidate and should be deprioritised regardless of price. The rooms are not subjective; the outcomes are not negotiable.
Specific questions for the buyer before the second visit
- For the trim line’s IATF 16949 cert, can the supplier email the scope page with the trim-line legal entity highlighted?
- For the CMM bay, can the supplier send me the calibration sticker and probe-qualification log for the CMM that the program will run on?
- For the surface room, what is the supplier’s waviness (Wa) and orange-peel (long-wave) performance on the last three door-panel trial cycles?
- For the trace room, can the supplier give me a 90-second screen-walkthrough on at least one of their active molds?
- What is the supplier’s last PPAP-package timing for a similar Class-A part — T1 to PPAP in working days?
What a follow-up visit should cover
If the first visit clears the four rooms, the second visit should focus on three operational metrics that the first visit cannot test in real time: (1) the dimensional Cp/Cpk trend over the past 12 production runs on a similar Class-A part, (2) the cosmetic reject rate trend over the past 6 months on a similar Class-A program, and (3) the RMA database query showing defect-by-destination over the past 12 months on similar interior trim programs. These three metrics cannot be produced on the day of a first visit; they require the supplier to pull production data from the previous reporting period. A supplier who can produce all three within a working day of the second visit is running a mature QMS; a supplier who cannot produce any of them is running a marketing operation.
The Cp/Cpk trend. Cp/Cpk is the process capability index that measures whether the supplier’s dimensional output is within specification. For a Class-A trim program, the buyer should expect Cp/Cpk ≥ 1.33 on cosmetic dimensions and Cp/Cpk ≥ 1.67 on critical mating surfaces. A trend over 12 production runs shows whether the supplier’s capability is stable (Cp/Cpk holding ≥ 1.33 consistently), improving (Cp/Cpk trending up over the 12 runs), or declining (Cp/Cpk trending down — a warning sign of equipment wear or process drift).
The cosmetic reject rate trend. Cosmetic reject rate is the percentage of parts that fail visual inspection at end-of-line. For a Class-A program, the buyer should expect reject rate ≤ 2% on the trim face. A trend over 6 months shows whether the supplier’s cosmetic discipline is stable (reject rate holding ≤ 2%), improving (rate trending down), or declining (rate trending up — a warning sign of polish discipline loss or steel-grade drift).
The RMA database query. RMA data by defect category and destination port shows the field-failure pattern over the program life. A mature supplier’s RMA database can produce a one-line summary of the last 100 RMA tickets within a working day, broken down by defect (warpage, sink marks, weld lines, paint defects, assembly defects) and by destination (which OEM customer and which port). A supplier whose RMA data is in a mailbox or in a binder cannot produce this summary, and the buyer should walk away.
Site links referenced in this letter
- BYD Tier 1 program product page
- Automotive mold category
- 2K two-shot trim/handle mold
- About Jinlong
- Yuyao site address
The economics of the four-room audit
A four-room audit takes 30-45 minutes on the floor and 30 minutes on documents. The audit cost to the buyer is roughly 2-3 hours of senior procurement time, plus travel if the visit is on-site. The audit cost to the supplier is roughly 2-3 hours of senior QA time, plus any production data the supplier has to compile for the second visit. Total audit cost on both sides is roughly USD 1,500-3,000. The cost of a failed Class-A program on a Tier 1 trim line is measured in millions of dollars (recall cost, line stop cost, warranty reserve), plus the buyer-side procurement time to qualify a replacement supplier (3-9 months). The audit-to-program-cost ratio is roughly 1:1,000 to 1:10,000 — the audit pays for itself many times over on every program it de-risks.
Why publish this letter publicly? Most Tier 1 procurement teams do not publish their supplier audit methodology. The reason is that the methodology, once published, becomes a target for suppliers to optimise against rather than a tool for buyers to use operationally. We are publishing this letter because the four rooms are observable on any supplier floor visit; the rooms are not proprietary; the rooms can be audited by any Tier 1 procurement team with the same outcome. A supplier who can pass the four rooms today is a Class-A candidate; a supplier who cannot pass them today will not pass them tomorrow by hiring a consultant.
Why I am publishing this letter
Because most Tier 1 procurement teams ask the same questions of multiple Chinese mold suppliers and converge on a comparable answer, the letter format makes the difference legible at a single read. Because the two suppliers differ on the four audit rooms, the side-by-side comparison is the fastest way to see why one of them passes a Tier 1 program today and the other does not. Because the operational test is the audit floor map and not the slide deck, the questions at the close are the kind a procurement VP can hand straight to the next visit.
About the author
Wanchen Xuan is a foreign trade specialist at Ningbo Jinlong Electric Appliance Co., Ltd., coordinating international mold and OEM inquiries. LinkedIn · yyjlong@chinajinlong.com.