What Japanese Buyers Want in Your Technical Documentation

Your technical documentation package meets every requirement a home-market customer asks for: conformance certificates, dimensional drawings, material data sheets, test reports marked complete. Then a Japanese buyer reviews the same package and asks for something it does not contain, a lot-level inspection certificate, tolerance data at a finer grain, or evidence of how the part behaves near the edge of spec rather than proof it simply passed. The request repeats across calls, and the documentation underneath it has not weakened in the meantime. Your team has not gotten worse at its job between one inquiry and the next.

What is happening is a mismatch between two quality systems, not a gap in your documentation. The Japanese buyer’s own quality management framework requires them to build and retain their own supplier-evaluation record, and a certificate that proves conformance to your standards, however complete, cannot substitute for evidence built against theirs. The stakes are practical rather than academic: a request you cannot answer quickly can read, on the buyer’s side, as a signal about how the relationship will run at volume, not as evidence about your product. If this is your company’s first sustained relationship with a Japanese buyer, the pattern below sits inside a wider set of behaviors covered in a practical guide to doing business in Japan.

Why the Requests Keep Coming Even Though Your Documentation Is Complete

Two mechanisms explain the pattern, and both sit on the buyer’s side of the relationship, not yours. Picture the sequence as it typically plays out. Your team sends the ISO 9001 certificate and standard test report that has closed deals with customers on three continents. The Japanese purchasing contact acknowledges receipt, and the file goes quiet. Then a different question surfaces, this time about lot traceability or a specific tolerance class, often routed through the same contact but clearly written by someone else. The document that closed the last ten deals apparently is not the document this buyer needed.

Their Quality System Requires Their Own Record, Not Yours

ISO 9001:2015, clause 8.4 on external providers, adopted in Japan without modification as JIS Q 9001:2015, requires the buying organization to determine and apply its own criteria for evaluating, selecting, monitoring, and re-evaluating external providers, and to retain documented records of that activity. The obligation belongs to the buyer, not to you.

Your ISO 9001 certificate (or IATF 16949, or any other third-party certification) demonstrates that your quality system conforms to a certification body’s assessment against a common global standard. It does not, and cannot, stand in for the buyer’s own clause 8.4 record, because that record has to be built against the buyer’s own criteria and retained by the buyer itself. A Japanese purchasing or QA department that accepted your certificate as a substitute for its own evaluation would be failing its own audit requirement, not yours. So it asks for the underlying data instead, lot certificates, tolerance data, capability data, and files that as its own supplier-evaluation evidence, independent of how complete your certification already is.

Trust Resets to Full Scrutiny for Any New or Foreign Supplier

Japanese buyer-side quality control commonly runs on a graduated inspection model: full inspection on every lot, moving to sampling (抜取検査) as a supplier earns a track record, moving further to skip-lot or document-only review at the far end, as described in Japanese buyer-side quality control practice. The track record that earns this reduction is specific to that buyer relationship. It does not transfer.

A new or foreign supplier starts back at full inspection regardless of global reputation or how many other customers it already supplies without incident. Many buyers also run an initial flow-control period, roughly the first three lots or three months, of intensified lot-by-lot monitoring before settling into steady-state review. The most likely explanation for why requests can feel repetitive or relayed through an intermediary is that the purchasing contact handling day-to-day communication is often not the party with final technical sign-off; that authority commonly sits with a QA reviewer who is less visible in the relationship.

What a Lot-Level Inspection Certificate Needs to Show

When a Japanese buyer asks for a lot-level inspection certificate, the request is usually for three things at once: actual measured values from that specific production lot, traceability to the lot number, and validation independent of the party who made the product. A statement that the lot “conforms to spec” without the underlying figures typically does not satisfy any of the three.

The document has a name your buyer will use in correspondence: 検査成績書 (kensa seisekisho, an inspection certificate or test report), and for steel specifically, ミルシート (a mill sheet, or mill test certificate). Japan maintains a dedicated standard for what these documents contain in the case of steel and steel products, JIS G 0415, its inspection-document standard for steel, which is closely aligned with the international standard for the same purpose, ISO 10474. Your own product may not be steel, in which case JIS G 0415 is not the certificate you would issue. Treat it as a documented example of the Japanese inspection-document convention: a per-lot record of measured results tied to a specific heat or lot number, rather than a general statement of conformity.

The way these documents are graded by how independently they are validated is captured cleanly in EN 10204, the European standard for metallic-product inspection documents, and the structure generalizes even for readers outside metals. It defines a 2.1 declaration of compliance (a supplier statement with no test data), a 2.2 test report (test data, not independently validated), a 3.1 certificate (validated by the manufacturer’s own authorized inspector, someone independent of the production department), and a 3.2 certificate (the same validation, additionally countersigned by an independent third party or the buyer’s own representative).

Your product will likely never carry a JIS G 0415 or EN 10204 certificate type by name unless it happens to be a metallic product covered by those standards. The principle behind the grades generalizes regardless: actual measured values, lot traceability, and validation independent of the production department. For a machined bracket, this could mean listing the actual measured hole diameter and location relative to datum for the specific lot shipped, not only the drawing’s nominal figure. For an electronic assembly, it could mean logging measured values from a specific test bench run against a specific serial or lot code, rather than a generic pass certificate for the model line.

None of this requires a new inspection system. If your quality team already records actual measured values per lot for internal process control, and simply discards or aggregates them before the file leaves the building, the fix is procedural: retain the lot-level readings, tie them to the lot number already on your packing documentation, and route the review step through someone who did not perform the measurement. That is closer to what a 3.1-level validation asks for than a new testing regime is.

How Tolerances Are Expressed and What Gets Read Into Them

A part can sit well within your own tolerance band and still fail to answer the question a Japanese drawing is actually asking.

General Tolerances and Fits: What “Within Spec” Actually Documents

JIS B 0405 sets general tolerances for linear and angular dimensions that carry no individual tolerance indication on the drawing, grouped into four classes, f (fine), m (medium), c (coarse), and v (very coarse), identical to ISO 2768-1. JIS B 0401-1 defines the limits-and-fits system used for holes and shafts, the H7/g6-style designations, aligned with the international ISO 286 system.

Consider a shaft diameter your own drawing calls out as 20mm plus or minus 0.05mm. That figure alone does not tell a Japanese buyer’s QA reviewer which recognized fit class it corresponds to, or whether it is meant to mate with a bore toleranced to a specific class of its own. Restating the same feature as, for example, an h7 tolerance under the JIS B 0401 fit system, or naming the JIS B 0405 class your general tolerances follow, answers the question in the reviewer’s own reference system rather than asking them to reverse-engineer it from your house convention.

When a Japanese buyer’s drawing calls out a tolerance class or a fit designation, it is asking you to document your part against that specific, named class, not to state “within spec” against a house tolerance band that may be looser, tighter, or built on a different reference system entirely. The practical fix is straightforward: state which recognized tolerance class or fit designation your part actually meets, in the nomenclature the drawing uses, rather than only your own internal tolerance figures.

Why a Calibration Certificate Alone Does Not Answer the Question

A calibration certificate, by convention, does not carry a pass/fail judgment on your product. It documents the measured value produced by a measuring instrument together with its expanded uncertainty. Japanese calibration-industry guidance describes the receiving party, not the calibration lab, as responsible for comparing that measured value and uncertainty against its own tolerance and making its own pass/fail call.

A calibration certificate for your gauge or test equipment proves the equipment is accurate. It does not, by itself, prove your product measurement passed a given tolerance; that comparison is a separate step, and it is the step a Japanese buyer expects to see documented, ideally performed by you at the point of test rather than left for their QA department to reconstruct. Together, the tolerance class and the calibration data answer two different questions a buyer is likely to ask in the same review: what your part is supposed to measure, and what it actually measured against a traceable reference.

Why Buyers Also Want to See Behaviour at the Edge of Spec

Beyond a report that shows the part passed, some Japanese buyers ask for data closer to the edge of the tolerance band, worst-case or near-limit measurements, rather than values clustered around nominal. No single source documents this motivation directly, so the explanation below is informed inference rather than settled policy, built from two practices that are documented.

First, process-capability vocabulary, Cp and Cpk, is defined in JIS Z 8101-2, identical to ISO 3534-2, as statistical terminology, not as a capability-threshold standard; the methods for calculating and applying a capability index sit in a separate methods framework. A Cpk of 1.33 or higher functions in practice as a common contractual expectation in automotive and electronics supply chains, not a figure any JIS document mandates, and higher-reliability sectors such as aerospace or semiconductors often expect 1.67 or above. Where a buyer references a Cpk threshold, treat it as an industry or contractual convention to confirm with them directly, not a regulatory requirement to look up. If your quality system already tracks Cp and Cpk internally for process control, that data is usually enough to satisfy this kind of request as it stands; the change is often only in what you choose to send, not in how you measure.

Second, a practice known in Japanese manufacturing relationships as 4M change management (Man, Machine, Material, Method) commonly obligates a supplier to notify the buyer, often before the change takes effect, of any process change that could affect quality. This is a contractual duty where it appears in an agreement, not a statutory one. In practice, if you already notify customers informally of material or process changes as a courtesy, formalizing that into a documented pre-notification step, tied to the specific part number or drawing revision, moves you from a courtesy to something closer to what a Japanese buyer’s contract is likely already asking for. Together, a buyer working under this kind of change-notification and capability-tracking regime most likely wants edge-of-spec evidence because it wants confidence the process holds up away from the easy middle of the tolerance band, not only that one batch happened to measure well.

When a JIS Reference Belongs Alongside Your ISO or ASTM Documentation

JISC, the Japanese Industrial Standards Committee, is Japan’s member body to ISO and develops JIS standards; the Japan Standards Association publishes and sells them. Many JIS standards adopt their ISO counterpart identical (IDT) or modified (MOD). JIS Q 9001:2015 is identical to ISO 9001:2015, JIS B 0405 is identical to ISO 2768-1, and JIS Z 8101-2 is identical to ISO 3534-2, all cited above.

Where your documentation already cites the ISO equivalent and the corresponding JIS standard is identical to it, you are not required to also generate a JIS-numbered version of the same underlying data; the substance already matches. Adding the JIS reference alongside your existing ISO citation removes the step your buyer’s QA reviewer would otherwise have to take of checking whether the two align on their own. Where the relationship is modified rather than identical, say so directly instead of assuming equivalence, and reference the JIS standard by number and scope so the reviewer can verify it independently. Treat any ASTM comparison the same way: state your ASTM reference plainly and let the buyer confirm equivalence on their side rather than asserting a JIS-ASTM match without a source for it. An ASTM-certified test report, for example, stays useful as is; adding a line noting that no verified JIS equivalent is being claimed, rather than guessing at one, is more defensible than silence on the question.

Inside the Two Departments That Decide Whether You Pass

Two departments handle your documentation on the Japanese side, and they read it for different reasons. 購買 or 調達 (purchasing/procurement) owns the commercial relationship and the supplier record. 品質保証 or 品証 (quality assurance) owns the technical sign-off and, in many organizations, holds shipment-halt authority, meaning QA can stop a shipment at incoming inspection independent of what purchasing has already agreed commercially. A request that arrives through your purchasing contact can originate from a QA reviewer you have never spoken with directly, moving through the internal approval process that can leave a promising inquiry silent for weeks.

Four internal gates typically apply, and each produces its own record. A supplier can clear one gate while still being asked for more evidence at the next:

• New-supplier evaluation: document review plus, often, an on-site audit covering the QMS, equipment, track record, and corrective-action process. Foreign relationships commonly add a three-document set: a basic transaction agreement, a quality assurance agreement, and a non-disclosure agreement.
• First-article inspection: a one-time check against every dimension and spec on the drawing, performed before mass production is authorized, and distinct from routine per-lot inspection.
• Mass-production transfer decision: a PPAP-equivalent gate that decides whether the relationship moves from first-article to steady production. The gate logic is broader than the automotive industry it is best known in.
• Incoming inspection: recurring, matched to lot number, the point where the certificate types and tolerance data discussed above actually get used.

The on-site audit component of new-supplier evaluation typically covers more than a facility walkthrough. Reviewers commonly check how a corrective action from a past nonconformance was documented and closed, and how a supplier manages 4M changes internally, not only whether the finished part meets spec today. A foreign supplier asked for this kind of process evidence for the first time is being asked for something new to the relationship, not something deficient in what it already has.

Neither JETRO nor METI publishes guidance specifically addressed to foreign suppliers entering Japanese supply chains; research across multiple passes for this article found no such material in English, only guidance aimed at Japanese firms diversifying their own outbound sourcing. That gap is a reasonable explanation for why this structure is hard to find described anywhere before a buyer relationship surfaces it directly. The documentation for it largely exists inside Japanese industry practice rather than in a public guide.

What Closing the Documentation Gap Actually Looks Like

None of this changes what your product is capable of. It changes what you can prove about it inside a system that requires the buyer to hold its own evidence, built against its own tolerance references, and reset for a new relationship regardless of your track record elsewhere. The buyer is not asking you to be a different kind of manufacturer. It is asking you to translate what your process already does into evidence built for its own record. Closing the gap means producing lot-level measured data with traceability, stating tolerances and fits in the nomenclature your buyer’s drawings already use, documenting behavior at the edge of spec where your process allows it, and referencing the JIS equivalents alongside the ISO or ASTM standards you already cite.

For manufacturers managing this shift without an in-house Japan team, our Japan market entry consulting works through the specific documentation gap with your existing technical team, translating an institutional expectation into a package your buyer’s QA department can approve on its own timeline.