Supplier Attached Inspection Reports: Five-Step Method for Verifying Incoming Inspection Reports and Data Reliability

1. The "Perfect Data" Report Attached with the Shipment

On the IQC table of a hardware manufacturing company, a batch of stainless steel wire materials was laid out with the accompanying documents: certificate of conformity, material test certificate, dimensional inspection report, and surface quality inspection report, totaling six pages. The report header, batch number, inspection items, specification limits, actual measurement values, and judgment columns were all present, with "conformity" marked in each judgment column. The last page bore the supplier's quality inspection stamp. The inspector, following the "Incoming Inspection Work Instruction," confirmed that "all incoming documents are complete" and clicked "conformity for warehousing" in the ERP system.

Two and a half months later, the springs made from this batch of wire materials experienced batch failures at the client's site. The client's laboratory conducted a metallographic analysis and found that the carbon content in the material exceeded the upper limit specified in the contract, indicating typical high-carbon embrittlement. During the traceability process, the quality manager reviewed the six-page report, and a series of issues emerged: the furnace batch number on the report was off by one digit compared to the number on the physical label; the chemical composition testing method cited in the report was the supplier's internal method, while the technical agreement specified the national standard method; the material test certificate was issued twenty-eight days before the delivery, and it was the same certificate for the previous batch; and the dimensional report, except for the batch number, was identical to a report from another supplier five months earlier.

This report was not forged, but it did not prove anything. The nonconformity identified during the audit was straightforward: the company had not defined and implemented requirements for verifying the inspection/test results provided by external suppliers, and the incoming inspection records did not demonstrate that the planned verification activities were carried out.

Many companies have two versions of incoming inspection: what is written in the documents is "inspection," but what is done on-site is "collecting documents." Collecting documents is not wrong in itself, but the issue is that the collected documents are the supplier's self-declaration. Treating the declaration as the result means accepting the supplier's self-inspection capability, data authenticity, and batch controllability as established facts.

2. Distinguishing the Four Types of Reports: Their Proofing Power Varies Widely

To discuss verification, we need to break down the vague term "report attached with the shipment." The issuing party, legal and system status, and the required depth of verification are entirely different for each type.

A key point at the system level needs to be clarified: the report attached with the shipment, in standard terminology, is considered "inspection/test results from external suppliers," which is one of the acceptable forms of verification evidence. However, it cannot replace the planning and implementation of the incoming inspection process. In other words, a company can certainly use "reviewing supplier reports" as one method of incoming inspection, but it must be clearly planned in advance: which materials will use this method, to what extent they will be verified, who will perform the verification, and how to handle inconsistencies. Writing it as "exempt from inspection" or "materials are conforming if all documents are complete" essentially downgrades the control action from "verification" to "receiving."

To determine whether a report is credible, you only need to check four elements. If any one is missing, the report's proofing power collapses:

• Traceability: The batch number/furnace number/serial number on the report must uniquely correspond to the physical items at hand and be traceable to the supplier's production records. Batch number mismatches are the most common and often overlooked form of report forgery or error—because they look too "complete."
• Method Consistency: The testing methods, sampling plans, and judgment criteria used in the report must align with those specified in the contract, technical agreement, and drawings. Different methods mean incomparable data, and the judgment of conformity loses its meaning (e.g., hardness HRB vs. HRC, salt spray 24h vs. 48h, different tensile sample shapes).
• Institution and Personnel Qualifications: Whether the supplier or third-party laboratory has the capability/accreditation for the project and whether the signatory is authorized. It is not uncommon for third-party reports to bear accreditation marks, but the projects fall outside their scope of accreditation.
• Sampling Representativeness: Whether the samples tested in the report were taken from the batch according to the agreed sampling plan and whether the sample handling and submission chain is traceable. A report that tests only one sample cannot represent an entire furnace batch or a whole batch of coil material.

3. Five-Step Method: Making Reports Truly "Prove" Something

Step One: Create a Report List, Define Verification Depth by Material Risk

The depth of verification should not be determined by the inspector's mood on a given day but by the risk of material failure. First, classify materials into three levels based on their impact on the final product, and then specify which reports must be provided and to what extent they should be verified for each level.

This list should be incorporated into the "Sampling Inspection Work Instruction" (SIP) and the supplier quality agreement, becoming an explicit requirement in the contract. Without prior agreement, it is impossible to hold the supplier accountable for "missing items in the report" after the fact—these "inspection requirements and results" should be clearly communicated during the procurement stage as per the communication requirements in 8.4.2.

Step Two: Create a "Report Verification Checklist" to Turn the Four Elements into Check-Off Items

The value of a verification checklist lies in ensuring that different people reach the same conclusion. It is recommended to include at least the following items, check them off one by one, and leave a trace:

1. Whether the material code, name, and specification version match the purchase order/technical agreement;
2. Whether the report batch number/furnace number matches the physical label and delivery note in three places;
3. Whether the report issuance date covers the current batch (to prevent using old reports for new batches);
4. Whether the inspection items cover all mandatory items agreed upon in the protocol, and if there are any missing items;
5. Whether the standard number and version of the testing method for each item are consistent with the agreement and are current;
6. Whether the sampling plan (sample size, sampling location, sampler) is clearly defined and meets the agreement;
7. Whether the specification limits, units, and judgment columns are complete, and if they are "item-by-item judgments" or just one overall conclusion;
8. Whether the actual measurement data falls within the specification and is of a reasonable magnitude (e.g., no misplaced decimal points, no incorrect units);
9. Whether the issuing institution/laboratory's qualifications and scope of accreditation cover the tested items;
10. Whether the signatures and stamps are complete and if the signatory is an authorized signatory;
11. Whether there are any signs of alteration, splicing, or covering of key fields in the report (for electronic reports, check if they are editable);
12. Compare with historical reports: whether there are any abnormal changes in items, methods, and data levels.

Item 12 is the easiest to overlook but often the most critical. For the same material, if the tensile strength of the first five batches was between 520-560 MPa, and the report for this batch suddenly shows all values between 610-615 MPa, with values clustered in a way that doesn't resemble real measurements, the appropriate action is not to release the batch but to ask, "Has the actual source of this batch changed?"

Step Three: Conduct "Report-Physical Item" Comparison, Use Actual Measurement Data to Test the Credibility of the Declaration

No matter how detailed the paper verification, it is only a review of the declaration. The true way to verify the declaration is to perform a certain percentage of actual measurement comparisons based on risk: re-sample from the current batch and measure key characteristics using the company's own measurement system, then compare the results with the report values.

The frequency of comparisons does not need to be for every batch and every item, but it must exist. For Class A materials, it is recommended to verify the first batch and then sample verify quarterly or based on supplier performance; for Class B materials, verification can be done semi-annually or triggered by anomalies. The core logic is: if you never verify, the numbers on the report will never be wrong, and they will never have any meaning.

For items like appearance, color, and texture that are difficult to describe with data, the approach is to bind the report with physical samples—the supplier's report should note the sample number, retain sealed samples with the shipment, and, if necessary, include sample photos or color card numbers in the report. This practice is often missing in many companies, resulting in reports stating "appearance conforms" while the actual items on-site have three different colors.

Step Four: Tiered Disposition of Verification Results—More Than Just "Release" and "Return"

The output of verification should not be limited to just "conforming" and "nonconforming" conclusions. At least four levels should be defined:

• Direct Release: All four elements are complete, the comparison is consistent, and the supplier's performance is good;
• Conditional Release: The report has non-critical flaws (e.g., unclear sampling plan, method differences in non-critical items), limit use to this batch and record, and simultaneously require the supplier to rectify the report format;
• Suspend Release and Strengthen Inspection: Trust issues arise from batch number mismatches, old reports used for new batches, or abnormal data. This batch should be fully inspected or sent for third-party re-inspection, and the supplier should be placed on an observation list;
• Reject and Initiate Corrective Actions: Nonconformity in actual measurement, clear mismatch between the report and physical items, or signs of forgery or alteration.

The third level is the most challenging test of management maturity. In many companies, if the report has issues but the production line needs the material, the batch is released with a "not to be repeated" warning, and no one records it as a negative point in the supplier's performance. The correct approach is to update the supplier's performance scorecard simultaneously—report credibility is a quantifiable supplier capability indicator, such as "first-time report pass rate," "report issue occurrence rate," and "downtime due to report issues," which should be ranked quarterly and included in supplier grading and allocation.

Step Five: Closing the Loop—Incorporate Verification Requirements into Agreements and Integrate Data into the System

The first four steps address "how to verify this time," while the fifth step ensures "no reliance on human initiative in the future":

• Incorporate into Agreements: Clearly specify the project list, method standards, sampling plans, issuance cycles, electronic report formats, and responsibility for authenticity in the supplier quality agreement. Agree that "losses due to inaccurate reports will be borne by the supplier" to provide a basis for claims.
• Incorporate into SIP: In the incoming inspection work instruction, specify the "report verification method + actual measurement comparison frequency" for each type of material, providing a basis for the inspector and internal audits.
• Systematize: Convert the report verification checklist into a structured input form in the ERP/QMS system. The verification traces (who verified, what items were verified, comparison results) should be archived with the batch records, not just stamped "verified" on the paper report—such a stamp cannot prove anything during a client audit.
• Anti-Forgery for Electronic Reports: When accepting PDF or electronic reports, prioritize those with electronic signatures, verifiable report numbers, or QR code query entries. Pure images or editable documents should be considered low-credibility evidence and supplemented by third-party or actual measurements.
• Reverse Audit: Conduct an on-site or remote audit of the supplier's laboratory for Class A materials annually, to review their testing capabilities, equipment calibration status, report issuance process, and original record management. Auditing only the product and not the laboratory is equivalent to checking only the conclusion and not the method.

4. Five Common Misconceptions

Misconception One: Using COC as COA. A certificate of conformity is just a declaration and does not contain data. Using it to replace an inspection report means deleting the verification action and retaining only a signature.

Misconception Two: One Report for Multiple Batches. Suppliers use the same report for multiple batches, especially in continuous supply. The report date and batch number are the two key points for identifying this issue, and each batch must be verified individually. Do not accept explanations like "we have always been at this level."

Misconception Three: Verifying Only the Conclusion, Not the Method. Seeing "conforming" and passing it. The testing methods, standard versions, and sampling plans are the elements that determine whether the data can be used for judgment. Verifying only the conclusion means accepting the supplier's self-defined judgment rules.

Misconception Four: Considering Reports Complete as Exempt from Inspection, and No Further Monitoring. Exempt from inspection/trust release is a simplified verification based on performance, not a cancellation of verification. When performance declines, there must be an automatic rollback to strengthened inspection, and this rollback rule must be predefined, not decided on the spot when batch issues arise.

Misconception Five: Not Leaving Verification Traces. The items are indeed verified on-site, but the records only state "all documents are complete, conforming." If a client failure occurs, the company cannot provide evidence that "we implemented the planned verification activities," leaving it vulnerable both in the system and legal contexts.

5. One Sentence Summary

The report attached with the shipment is the supplier's declaration, not your inspection result—only verification actions that are traceable, comparable, and accountable can turn someone else's declaration into your own evidence.

The report is a declaration; verification is the evidence.

Knowledge code: 9.2.1

Version: v20260917

Author: QTank QTank is dedicated to providing systematic professional knowledge, methodologies, and practical tools for quality management practitioners, helping companies continuously improve their quality capabilities.