Product Testing Services
Product testing services should be commissioned only after the testing question is defined. A laboratory can generate technically sound data and the report can still be unusable for a target market if the wrong standard, edition, method, sample configuration or recognition route was selected.
The safe sequence is therefore: define the product and target markets, build the test matrix, verify evidence-acceptance rules, verify the laboratory for the exact scope, lock the sample strategy, then authorize testing. Testing is evidence; it is not automatically a certificate, registration or market authorization.
Build the testing matrix before asking for a quote
Start with the decision the evidence must support. One product can need several different kinds of testing, and the required laboratory status may differ between them. Do not treat “ISO/IEC 17025 accredited” as a substitute for checking the actual accredited scope and any regulator or scheme recognition.
| Evidence objective | What must be fixed before testing | Typical evidence output | Main acceptance question |
|---|---|---|---|
| Regulatory safety | Target market, product classification, applicable legal requirement and referenced standard/method | Test report with product identification, method, results and verdict where applicable | Does the target regime accept this method, edition, laboratory and report format? |
| EMC / radio | Device configuration, operating modes, accessories, firmware, frequencies and worst-case setup | Measurement report and configuration record | Must the testing come from a regulator-recognized laboratory or named scheme? |
| Chemical / material restrictions | Materials, coatings, homogeneous parts, supplier declarations and substance scope | Analytical results for defined substances/materials | Can component evidence be used, or must the finished product/material be tested? |
| Performance / claim substantiation | Exact marketing claim, test conditions, pass/fail criterion and comparison basis | Performance dataset or verification report | Is the method appropriate to the claim and to the party that will review it? |
| Durability / reliability | Use profile, load cycles, environment, conditioning and failure criterion | Lifecycle, stress or endurance results | Is this voluntary engineering evidence, customer-required evidence, or part of a regulated route? |
Verify that the result will be accepted
Laboratory competence and report acceptance are related but different questions. ISO/IEC 17025 addresses laboratory competence, impartiality and consistent operation. ILAC explains that an accredited laboratory’s certificate includes a scope of accreditation and that users should verify the specific tests or methods covered. ILAC also notes that accreditation improves international acceptance but cannot guarantee that every regulator or other party in a market will accept a result.
- Identify the exact rule, scheme, certification body, authority, retailer or customer that will rely on the report.
- Check whether it requires a specific standard edition, method, national deviation or test location.
- Check the laboratory’s current accreditation scope for every required method—not only the accreditation certificate headline.
- Check whether additional regulator/scheme recognition is required. For example, CPSC accepts a laboratory’s accreditation for specified children’s-product safety test methods on a test-by-test basis.
- Confirm the required report language, product identification, photographs, sample traceability, uncertainty statements, signatures, digital validation or other formalities before testing starts.
Plan samples before the purchase order
A quote is incomplete if it does not say what will be tested. Sample planning should connect the commercial product family to the tested configuration and record why the selected specimens are representative.
| Sample decision | Question to answer | Risk if left unresolved |
|---|---|---|
| Prototype or production sample | Must evidence represent final production, or is pre-compliance/prototype testing acceptable at this stage? | A passing prototype report may not support the final product if construction or materials change. |
| Model family / variants | Which variants share the same safety-critical design, components and worst-case characteristics? | An unjustified “family” can cause unsupported report reuse; testing every cosmetic variant can waste cost. |
| Worst-case configuration | Which voltage, power, radio mode, accessory, load, material, size or operating condition is most demanding for each test? | A convenient sample may not cover the configuration with the highest compliance risk. |
| Sample custody and identity | Who selects, seals, ships, photographs and identifies samples, and must selection be independent? | Weak traceability can make otherwise good data difficult to defend or reuse. |
Decide whether an existing report can be reused
Do not decide report reuse from the logo on the first page. Reuse is an acceptance decision that must be checked against the new product, new market and new route.
- Compare the tested model, bill of materials, critical components, firmware and ratings with the current product.
- Confirm the standard and test-method edition and whether the target route still recognizes them.
- Confirm that the original laboratory had the required accredited/recognized scope when the test was performed.
- Check whether the new market, authority, certification body or customer accepts reports from that laboratory and country.
- Identify differences and map each difference to the tests it could affect; a targeted delta test may be enough in some routes, while others require a new full test.
- Obtain the complete report and supporting evidence. A certificate number, summary page or supplier statement alone may be insufficient for technical reuse.
There is no universal rule that a report issued by an ISO/IEC 17025-accredited laboratory is automatically reusable worldwide. ILAC expressly notes that market-specific requirements can still determine acceptance.
Handle failures and retesting without losing traceability
A failed test is not only a laboratory event; it is a change-control event. The failure record should show what failed, why the corrective change is expected to address it, and what evidence remains valid.
- Freeze the failed sample configuration and preserve the original report, photographs, setup and raw observations available to you.
- Classify the failure: design, component/material, software/firmware, manufacturing, conditioning, setup or an incorrect test scope.
- Define the corrective change and list every requirement the change could influence—not only the requirement that failed.
- Agree the retest scope in writing with the laboratory and, where relevant, the certification body or authority. Record why unaffected tests remain valid.
- Issue the revised product/BOM/firmware revision under change control and link the passing retest evidence to that revision.
Use test-driven change control after a pass
Passing a test does not end the evidence lifecycle. A later change can break the factual link between the tested sample and the product being manufactured. Use a trigger map so engineering, procurement and compliance teams know when evidence must be reviewed.
| Change trigger | Minimum review question | Possible action |
|---|---|---|
| Critical component or supplier change | Does the substitute change safety, chemical, EMC/RF or performance characteristics covered by the report? | Document equivalence, targeted testing or broader retesting depending on the route. |
| Material / coating / formulation change | Does the composition alter restricted-substance, flammability, migration or other material-based evidence? | Reassess affected chemical/material tests. |
| PCB, radio module, antenna or firmware change | Can emissions, immunity, RF exposure, power, protections or operating modes change? | Reassess affected electrical/EMC/RF tests and route conditions. |
| Mechanical design or enclosure change | Can access, strength, stability, spacing, temperature or ingress behavior change? | Reassess the affected safety and durability tests. |
| New market or new claim | Does the existing report match the new market’s acceptance rules and the exact claim being made? | Run a gap review before reusing evidence; commission delta/new testing if needed. |
As a concrete regulatory example—not a universal rule—the U.S. CPSC states that when a children’s product undergoes a material change after initial certification, the changed product or component must be retested by a CPSC-accepted laboratory for the affected requirements, and a new Children’s Product Certificate (CPC) must be issued. This is why every project should define its own retest triggers from the applicable regime rather than copying a generic interval.
Send the laboratory enough information for a defensible quotation
Before booking, ask the laboratory to return a test plan or itemized scope that you can audit against the matrix. A price without a method list, sample count and acceptance assumptions is difficult to compare.
- Product name, model list, intended use, user group and target markets.
- Photos, manuals, specifications, ratings and critical construction information.
- Bill of materials or at least critical components/materials relevant to the tests.
- Applicable requirements already identified, plus any open classification questions.
- Requested standards/methods and editions—or a request for the laboratory to state the basis it proposes.
- Variant/family strategy and the proposed worst-case sample for each test group.
- Required laboratory accreditation/recognition and the evidence proving it for the specific methods.
- Deliverables: full report, photographs, raw-data availability where relevant, language, electronic validation, sample disposition and retest terms.
Know where testing stops
This page owns the evidence path: test scope, laboratory acceptance, samples, report reuse, failures, retesting and test-driven change control. If the unresolved question is how to run the whole compliance project, use product compliance services. If the question is a formal approval or certificate route, use product certification services. For operational comparison of laboratories, use product testing laboratory selection. For routing the same product across several countries, use global market access services.
Official references
Use the source that governs the actual product and market. The references below support the general evidence and laboratory-selection principles used on this page; the CPSC links are a jurisdiction-specific example.
Frequently asked questions
Is ISO/IEC 17025 accreditation enough to choose a laboratory?
No. It is a strong competence signal, but you still need to verify the laboratory’s current scope for the exact method and any additional regulator, scheme, certification-body or customer recognition required for your route.
Can I use one test report for several countries?
Sometimes, but never assume it. Compare the product, standard/method edition, national deviations, laboratory recognition and the receiving authority or certification body’s acceptance rules for each market. Reuse may be full, partial with delta testing, or not accepted.
How many samples should I send?
There is no universal number. The method, destructive steps, conditioning, variant strategy, statistical needs and regulator/scheme rules can all change the quantity. Ask the laboratory to state the sample count per test group and why it is sufficient before shipment.
Do I need to retest after changing a component?
Possibly. Map the change to every requirement it could affect, then apply the rules of the relevant regime or scheme. Some changes can be supported by equivalence or targeted testing; others trigger broader retesting. Do not limit the review to the test that previously failed or passed.
What should I do if two laboratories quote different test scopes?
Do not choose by price first. Ask both laboratories to map each proposed test to the same product classification, target market, requirement and standard edition. The difference may reveal a classification assumption, missing method, unnecessary test or a recognition limitation that must be resolved before booking.