What If the Test Report Has the Wrong Model Number?

← All guides

The report arrives, the buyer's compliance team reads it, and the email comes back short: this is not our model. The report says TC-30W. The carton, the rating label and the purchase order all say TC-30B. The product has not changed by a single component, but the document that is supposed to prove it was tested now describes something the buyer cannot trace.

This is one of the most common document problems we see after testing, and it is almost never a testing failure. It is an identity failure. The question that decides everything that follows is not "can the report be fixed" but a narrower one: was the model number wrong on the paper, or was it wrong on the sample?

The short answer

Split the problem in two before you contact anyone.

• The paper is wrong. The sample that was actually tested carried TC-30B on its label, the photos in the report show TC-30B, and the typist entered TC-30W. This is an administrative correction. The issuing laboratory can amend the report without going back to the chamber, and it normally takes days rather than weeks.
• The sample was wrong. What was tested really was TC-30W, or it was an unmarked engineering sample, and TC-30B is the name you intend to sell under. Then the report is not inaccurate - your product identity is unresolved. A corrected model number on the report will not fix a product that was never tested under the name it ships as.

Everything else in this article is detail hanging off that distinction. Get it wrong in either direction and you either pay for a retest you did not need, or you ship with a document set that falls apart the first time a customs officer or a marketplace compliance team looks closely.

Why a single character carries so much weight

A test report is not a standalone document. It is one link in a chain, and every other link in that chain has to name the same product. When one link disagrees, the whole chain is questioned - not just the report.

Document	What it states	What a mismatch causes
Test report	Description and identification of the item tested, as required by the laboratory's reporting rules	Buyer rejects the file; the document no longer describes anything you can ship
Declaration of conformity	Identification of the apparatus that allows traceability - typically model, type or batch reference	The DoC names a model with no supporting report, so the presumption of conformity has nothing behind it
Rating label and packaging	The model the customer physically receives	Label and report describe different products; the label is the first thing a market surveillance officer reads
Technical file	The complete evidence set, kept for the retention period	An auditor comparing documents finds an unexplained gap and asks for the missing test
Commercial documents	Model on the PO, invoice and packing list	Customs or the buyer's goods-in reject the consignment as not matching the declared certification
Marketplace listing	Model field in the seller's compliance portal	Listing suspension during a documentation check, often with the stock already in the fulfilment centre

In the EU, the declaration of conformity has to identify the product in a way that allows traceability, and the technical file has to contain the test reports that support it. A report for TC-30W sitting behind a DoC for TC-30B is not a small filing error. It means the file, as assembled, does not demonstrate conformity for the product being placed on the market. That is the sentence a buyer's compliance manager will use when they reject the shipment, and they are technically correct when they say it.

Five ways a model number ends up wrong

The fix depends entirely on the cause, so identify the cause before drafting an email.

Situation	Whose mistake	Correction path	Evidence the laboratory will want
Transcription error at the laboratory	Laboratory	Administrative amendment, no retest	Photo of the sample label, the original application form, the report page in question
Wrong model typed on your own application form	Applicant	Administrative amendment, sometimes with a fee	Written request from the applicant, corrected application, label photo
Factory internal model on the report, brand model on the label (ODM)	Neither, usually - two naming systems	Amendment to add the brand model, or a model equivalence statement in the file	Relationship between the two names, brand authorisation, label artwork for both
Family report lists only the representative model	Scope decision made at quotation stage	Amendment with a model list, only if the variants were declared and covered	The original family declaration, worst-case justification
Model renamed after the report was issued	Commercial decision	Not a correction - see the next section	Nothing; this is a change of identification, not a typo

The third row deserves a note because it is the one that produces the most arguments. A factory that has sold the same board for years under an internal code will put that code on everything, including the samples it sends to the laboratory. The brand owner then prints a different name on the retail box. Both parties believe they are right. The report is accurate about the item tested; the label is accurate about the item sold. What is missing is a document that ties the two names together, and that document is what the buyer actually needs - sometimes written as a model equivalence statement, sometimes handled by adding the brand model to the existing report.

The case that is not a correction at all

If the model number is wrong and the hardware moved at the same time - a different adapter, a new PCB revision, an added radio mode, a firmware change that alters output power - stop treating this as a paperwork problem. A corrected report would then describe a product that was never tested, which is worse than the original error, because it looks clean.

The same applies when the model on the report is a placeholder that was never the real product name, or when the tested sample was a hand-built prototype with no label at all. In those cases the honest route is a new test or a documented delta evaluation against the closest tested configuration. We have written separately about what counts as a component change that requires retesting, and the reasoning there applies directly: it is the configuration that was tested, not the name it carries.

Only the issuing laboratory can amend the report

This surprises people who assume any accredited laboratory can "correct" a document. It cannot. A report is the issuing laboratory's statement about work it performed. A second laboratory has no access to the raw data, the sample photos or the measurement traceability behind it, so it can only test again and issue its own report - it cannot amend someone else's.

Accredited laboratories operate under rules that make this explicit. In ISO/IEC 17025:2017, clause 7.8.8 (adopted in China as CNAS-CL01:2018 clause 7.8.8), any change to an issued report has to be clearly identified, and where appropriate the reason for the change has to be stated. An amendment is issued as a further document carrying the wording "Amendment to Report, serial number..." or an equivalent statement. If instead a complete new report is issued, it must be uniquely identified and must reference the original it replaces. The clause exists so that two versions of "the same" report cannot circulate without anyone noticing which one is current.

Two practical consequences follow. First, expect the amendment to be a separate numbered document - for example an amendment to report number XXXX-1 - rather than the original PDF quietly reissued with a corrected line. That is normal and buyers accept it. Second, the applicant named on the report controls the request. If the applicant is your factory or your supplier, the laboratory may decline to act on your instruction without written authorisation from the applicant, even though you own the brand.

The accreditation mark matters here too. A report carrying a CNAS or ILAC-MRA mark is amended under the accreditation body's rules, and the amendment carries the same mark. A report without an accreditation mark can also be amended, but it starts from a weaker position when a buyer or an authority questions it.

What to send when you request a correction

Most delays in this process come from incomplete requests, not from laboratory reluctance. A request that arrives with everything attached is usually handled in one pass.

1. The report number, page and line where the wrong model appears, plus the correct string, character by character. "TC-30B, not TC-30W" is unambiguous; "the model is wrong" is not.
2. A photograph of the sample's rating label as it was at the time of testing. This is the single most persuasive piece of evidence, because it shows what the laboratory actually had in front of it.
3. The original application form or test request, so the laboratory can see whether the error originated with them or with you. Nobody minds which - but the answer determines who signs the correction.
4. Where the report is going: a specific buyer, a marketplace compliance portal, a customs broker, a notified body. Laboratories handle a report that is about to be filed with an authority differently from one that is still internal.

A short written request is enough. It does not need to be a legal letter, but it does need to state the error, the correct value, the evidence, and the destination. If the error is yours, say so - laboratories are considerably more cooperative with applicants who do not open by blaming the typist.

After the amendment: the rest of the file has to follow

The amendment is not the end of the job. Once the report changes, everything downstream has to change with it, or you have simply moved the inconsistency.

• Reissue the declaration of conformity with the correct model, a new issue date and a signature. Do not hand-edit a signed PDF; a DoC with visible corrections is treated as unreliable.
• Keep the superseded report in the technical file with a one-line note explaining what was replaced and why. Deleting it looks like concealment if the file is ever examined, and the retention obligation covers the file as it evolved.
• Withdraw the old PDF from circulation. If three buyers and one distributor still hold the previous version, you now have two documents describing different products in the market, and the version someone forwards in eighteen months will not be the one you expect.
• Fix the commercial records - marketplace model field, customs declarations, packing lists, the buyer's vendor master. This is where most corrections quietly fail: the report is right, the label is right, and the invoice still says the old code.
• Check whether a certificate is involved. If the product also holds a certificate - CCC, CB, a notified body opinion - the model on that certificate is changed through the certificate body's own variation procedure, not by amending the test report. Those are two separate processes with two separate owners, and doing one does not do the other.

When correcting the report is the wrong move

We advise against a correction in four situations, and in each of them the client initially wanted one.

1. The report is right and the label is the thing that changed. A model was renamed for marketing reasons six months after testing. Asking the laboratory to amend the report so it matches the new label inverts the evidence: you would be altering a record of a test to fit a later commercial decision. Rename the product properly instead - and check whether the rename needs to be handled as a change of identification in the relevant scheme rather than as a report edit.
2. The hardware changed as well. Covered above. A corrected name on an untested configuration is a document that will not survive the first technical question.
3. You are not the applicant on the report. Without the applicant's authorisation the request usually goes nowhere, and pushing a supplier to "just get it changed" produces either a refusal or, worse, a document that was altered outside the issuing laboratory's control. That last outcome is the one that ends buyer relationships.
4. The wrong model is already in someone else's system. If the model has been used in customs entries, in a marketplace listing, or in a tender submission, a single amendment letter rarely closes the issue. Expect to supply the amended report, a new DoC and a short model cross-reference note together, and expect the buyer to ask why the change happened.

How long it takes and what it costs

For a genuine transcription error, no bench time is involved, so the work is administrative review and re-issue. In our experience that is typically a few working days, and many laboratories charge nothing or a nominal administrative fee - but confirm before you promise a date to a customer. Timelines stretch when the report has already been uploaded to an accreditation or certification system, when the original technical reviewer is unavailable, or when the request comes from someone who is not the applicant.

For anything that touches technical content, the clock resets to testing. There is no version of this where a laboratory legitimately re-dates a report to cover a configuration it did not measure, and any offer to do so should be treated as a warning sign about that laboratory rather than a shortcut.

A ten-minute check before the report leaves your desk

1. Compare the model string on the report with the rating label, the retail box and the manual - character by character, including hyphens and suffixes.
2. Compare the model on the report with the model on the declaration of conformity.
3. Check the trade name or brand on the report; a report in the factory's name against a brand label is the ODM case above.
4. If it is a family report, confirm every variant you intend to ship is named in the report or its annex, not merely covered by a worst-case argument you remember from a meeting.
5. Check the report date against the production date of the units you are shipping.
6. Check the revision or amendment number, and make sure the version quoted in your technical file and in what you sent the buyer is the same one.

What this comes down to

A wrong model number is not a cosmetic defect in a document; it breaks the identification chain that connects the tested sample, the declaration, the label and the shipment. Whether it can be corrected in a few days depends on one question that has nothing to do with paperwork: did the laboratory test the product you are now selling? If it did, an amendment restores the chain and costs almost nothing. If it did not, no correction will, and the sooner that is accepted the less expensive the year becomes.

We review report sets against labels, declarations and technical files before they go to a buyer, and where we performed the EMC, RF and safety testing we issue the amendment ourselves - including when the error originated on the sample label rather than in our typing. Send the report page, the label photo and the market you are shipping to, and we will tell you which of the two situations you are in.