Toy Test Report Variant Coverage: How to Prove Which SKUs Are Covered

A supplier sends one test report for “Model A.” The order contains red and blue versions, a light-up version, and a gift pack. “Same family,” the supplier says. Which of those sellable SKUs can the report actually support?

The answer is not in the family name. It is in the differences between the tested sample and each product being sold, the particular requirement being claimed, and the evidence connecting them. If that chain breaks, a certificate or marketplace model list can quietly grow beyond its test basis.

For a buyer, this is not a paperwork distinction. A purchasing team may agree the base toy, while a sales team later adds a new colour or accessory to the same listing. The factory may continue using the old model name after changing an input. By the time cartons are ready, the report can still be genuine but its link to the goods in the order may no longer be clear. The useful control is to make the sellable configurations and their evidence visible before the purchase order, laboratory booking and certificate are finalized.

This is a practical U.S. children's-product evidence method, not a legal opinion or a universal test plan. Have a qualified compliance professional and a CPSC-accepted laboratory determine the applicable requirements, sample plan and certification route for the actual toy. Our related guides cover CPSIA paperwork, production traceability and material controls. Here we focus on the decision linking each variant to its evidence.

Start with the product being sold, not the name on the report

16 CFR 1107.20 requires manufacturers to submit enough samples of a children's product, or samples identical in all material respects to it, to a third-party conformity assessment body for certification testing. The sample number must provide a high degree of assurance; manufacturing variability can call for more samples. The regulation does not say that one colour, model or family automatically represents all the others.

List every SKU and model number that could appear on the order, carton, retail pack, listing or Children's Product Certificate (CPC). Next to each, record age grading and intended use; physical construction; accessible materials, colours and coatings; included accessories; batteries and electronics; packaging and warnings; factory and key component suppliers; and the revision of the bill of materials (BOM), drawings and firmware. This is a working inventory, not a substitute for determining the applicable rules.

A change can matter to one requirement and not another. An unchanged mechanism with a new coating raises a different evidence question from a new battery door. A new outer box may not change the toy's physical test results, yet its age claim or warning may still need review. Ask the laboratory to assess the applicable rules for the real configurations. The CPSC testing guidance notes that testing depends on product class, age, use and material composition, and that laboratory acceptance is rule-specific.

Give this inventory a version and an owner. Do not let a sales description stand in for the physical specification: photograph the complete toy and its accessible parts, label each accessory, and identify the colour and material supplied for testing. When different SKUs have different component suppliers, say so in the inventory rather than hiding the variation in one family name. The version in the lab request should be the version the factory is asked to build; if production changes, the comparison has to be reopened.

Make the coverage decision one requirement at a time

Use a SKU-to-evidence register, with a separate line for each SKU and applicable requirement. For each line, record the tested product or component, report and result, sample/revision identity, the reason evidence might transfer, any laboratory limitation, and a named certifier-side decision. Mark the cell supported, additional evidence needed or not assessed. “Same series” is not a decision.

Consider an illustrative family—not a claim that these variants comply:

• A-RED: the submitted base toy. Record the actual tests, sample markings and construction. Even this row is not supported until results and applicable requirements are confirmed.

• A-BLUE: same proposed mechanism, new blue coating. Mechanical evidence might transfer if the relevant construction is materially identical; assess the coating and its production traceability separately. Do not presume either result.

• A-LIGHT: added light module and battery door. Review the affected construction and any newly applicable requirements. Hold the changed cells until a product-specific plan and results support them.

• A-GIFT: purportedly unchanged toy with new outer packaging. Check what the pack includes and whether age, warnings, instructions, claims or model identity change before deciding what can be reused.

• A-MAG: new magnet accessory. Hold the variant pending a laboratory assessment and applicable test evidence; the original toy's report is not a shortcut.

There is no universal “worst-case SKU.” The features that challenge one requirement may not challenge another. Small parts, cord configuration, coatings and power options are prompts for a hazard-specific discussion with the laboratory—not pre-approved worst-case choices or statements that a particular test will pass.

The register should distinguish what was tested, what was inferred from a tested sample, and what remains open. A test result may answer a mechanical question for unchanged construction while leaving a new coating unaddressed. Write the reasoning next to the precise row, including the tested sample number and the production revision it is meant to represent. If a source report identifies only a family code, ask for the missing model-and-sample bridge rather than filling it in yourself. Do not turn a blank cell into a pass just because a related model passed.

Ask for a coverage rationale that names the exceptions

Send the laboratory the complete SKU inventory and ask: Which variants and which requirements do the tested samples support? What material-equivalence or sample-selection rationale applies to each shared result? Which rows need other samples, component evidence or finished-product testing? Which limitations apply? Confirm that the laboratory is CPSC-accepted for each applicable testing scope; acceptance for one rule does not establish acceptance for all.

Attach photographs of the sample actually tested, labels, model markings, accessories and power configuration, plus the BOM and drawing revision. Record sample quantity and the laboratory's written response. Then compare the production specification with the tested unit. A report called “Model A” cannot by itself tell you whether the present production Model A still has the same coating, fastener or supplier.

Keep the chain of identity practical. Compare the laboratory intake photographs with the factory-approved specification and with a production sample. Record the model marking, appearance, dimensions relevant to the rule, accessible materials, fastening method, power arrangement, and included accessories. Preserve the dated BOM and drawing revision used for the comparison, plus the name of the person who approved differences. A laboratory letter describing representative models should identify its assumptions and exclusions. If the lab did not examine a later accessory or changed coating, do not silently extend the letter to that version.

The responsible certifier must make the certification decision; a laboratory's opinion supports that decision but does not transfer the certifier's duties. If a report sample failed an applicable certification test, 1107.20(d) requires investigation and correction before certification—not a selective reliance on passing samples.

Treat component reports as conditional evidence

A coating or fabric report can be useful, but it is not a certificate for the assembled toy. Under 16 CFR 1109.5, component testing may support a finished-product certification only where it is sufficient for the applicable requirement and the production component is identical in all material respects to the tested one. Finished-product testing remains necessary for a requirement that must be assessed on the finished product; doubt about sufficiency should favour testing that product.

For every component report relied on, tie the tested material to the production part and traceable lot or other identifier. Keep the applicable rule, method, sampling protocol, test date and values, testing party, and the records and attestations required under the reliance route. Check what happens after testing: substitution, assembly, contamination or degradation can defeat the premise. A supplier PDF without that link is an unresolved cell, not a green tick. See also the CPSC component-testing explainer.

For example, a report on one fabric formulation does not automatically follow a visually identical fabric from another mill or dye lot. Ask the factory how the tested input is identified at receipt, how substitutions are approved, and how the finished toy is linked back to that input. Where the requirement depends on the assembled article—rather than the component by itself—confirm the appropriate finished-product test route with the laboratory. This avoids both unnecessary blanket retesting and the opposite error of treating a generic material certificate as a complete toy certificate.

Reopen the decision when the product changes

Before releasing a new colour, accessory, supplier source, assembly method, firmware or packaging claim, identify the affected rows and ask whether the design, manufacturing process or component sourcing could affect compliance. A description such as “cosmetic only” does not establish that answer.

Where a manufacturer exercising due care knows or should know a design, process or component-source change could affect compliance with an applicable children's-product safety rule, 16 CFR 1107.23 requires sufficient third-party testing of the materially changed product and a new CPC based on passing evidence. The extent of testing depends on the change; in the limited circumstances set out by the rule, changed-component test results may be combined with earlier certification tests. Not every packaging or model-code edit automatically triggers that material-change rule, but every proposed reuse deserves an explicit impact decision. Do not release the changed variant while a required test or new CPC is unresolved.

Use a short change record before the next production lot: what changed, which SKUs and applicable requirements could be affected, what sample and report previously supported them, what the laboratory advised, who made the certifier-side determination, and which production revision can be released. Separate the assessment from the outcome. A change judged immaterial for one requirement still needs a documented basis, not a blanket “same as before” note. If material-change testing is required, the decision record should point to the passing evidence and the new CPC before the affected goods are treated as cleared.

Red flags that should stop an unsupported SKU

Hold the affected variant for review when a CPC lists a model absent from the test file; a report has no identifiable sample photographs or model markings; a new coating or supplier appears only in production records; a component report cannot be tied to the actual input; or a laboratory response says “all colours” without stating which requirements and differences were assessed. A gift pack that adds an accessory is not necessarily the same toy in a different box. A changed battery door should not ride on a prior mechanical result without an impact assessment. These are stop-and-check signals, not automatic conclusions that every product fails.

The person preparing the listing, certificate and shipment release needs the same exclusion list as the laboratory and factory. Mark a SKU blocked in the working register until its missing evidence or scope question is resolved, and keep its model number out of a blanket certificate claim. Otherwise an internal hold may disappear when a sales catalogue is copied to the next order.

Before the certificate or listing, close every row

Keep one controlled file: the actual sellable SKUs, rules assessed, requirement-level matrix, lab reports and applicable acceptance scope, sample identity, production BOM and supplier revisions, component traceability, the laboratory's rationale and limits, and a dated certifier-side decision. Explicitly list exclusions so a blocked model cannot slip onto a CPC or marketplace submission. Continue to monitor manufacturing and new changes after the initial decision; this matrix is not a replacement for ongoing obligations.

Awen Hollek helps buyers make that evidence bridge work on the China side—between factory revisions, component suppliers, the laboratory, and release records. The point is not to retest every superficial difference by reflex, nor to stretch one report over an entire sales range. It is to know exactly which claim each piece of evidence can carry.

Need a clear SKU-by-SKU coverage file before production release? Talk to Awen Hollek about a practical factory-to-lab evidence and change-control process.

FAQ

Can one toy test report cover multiple colours?

Possibly for some requirements, not automatically for all. Confirm that the sample is materially equivalent for each result being reused. Different pigments, coatings or substrates may call for distinct evidence; the laboratory and responsible certifier should document the product-specific rationale and trace production inputs.

Does new gift packaging mean the toy must be retested?

Not necessarily. First confirm that the toy and included parts are unchanged and review age claims, warnings, instructions, use, and model identification. If the change could affect compliance, obtain a requirement-specific assessment; do not label every packaging change a material change without that analysis.

Is a supplier's component test report enough for a CPC?

Only when component testing is sufficient for the relevant requirement and the conditions of 16 CFR 1109.5 are met, including material identity, traceability, due care and required documentation. Some requirements need finished-product testing. The finished-product certifier remains responsible for its CPC.

What should match the tested sample?

Identify model markings, photographed construction, materials, accessories, power setup, age/use assumptions and relevant BOM or drawing revision. Compare these to production, document differences and ask whether they matter to each applicable requirement. A common model name without this bridge is weak evidence.

When does a variant change require a new CPC?

When a design, manufacturing-process or component-source change meets the material-change condition in 16 CFR 1107.23—meaning a manufacturer exercising due care knows or should know it could affect compliance—the rule requires appropriate third-party testing and a new CPC on passing evidence. The changed-component route is limited by the rule; get product-specific advice.

Can a laboratory's “worst-case model” letter cover the whole SKU range?

Only to the extent its documented reasoning, tested construction, applicable test scope and stated limitations support each requirement and SKU. Ask which differences were examined and which rows remain untested. A broad family statement cannot replace the certifier's due-care review.