Product Testing Lab
A product testing lab should not be booked until you know what evidence the report must prove, for which product configuration, and who must accept that evidence. The safest sequence is to define the requirement first, then choose the laboratory, then lock the sample and change-control plan. A technically capable laboratory can still produce a report that is unusable for a particular regulator, certification scheme, retailer or target market if the wrong method, edition, sample, accreditation scope or reporting route is used.
This page focuses on the testing evidence path: test scope, laboratory selection, sample planning, report reuse, failures, retesting and product changes. It does not reproduce the complete certification or registration workflow.
Six questions to answer before asking a laboratory for a quotation
- What decision will the report support? Regulatory compliance, a certification scheme, buyer acceptance, engineering validation, a performance claim or internal quality control can require different evidence.
- Which exact product configuration is being tested? Record model, variants, critical components, materials, firmware/software version, power ratings, accessories and intended-use conditions.
- What requirement and test method apply? Identify the regulation, scheme rule, standard or customer specification, including the edition or amendment when relevant.
- Does the receiving authority require a particular laboratory status? Accreditation to ISO/IEC 17025 may be relevant, but some schemes or regulators also require specific recognition, designation or an approved laboratory list.
- How will samples be selected and identified? The tested units must be traceable to the configuration for which you intend to rely on the report.
- What changes would make the evidence stale? Decide in advance which component, material, supplier, firmware, enclosure or process changes require technical review or retesting.
Testing matrix: convert the commercial request into an evidence plan
Before a purchase order is issued, create a testing matrix. It prevents a vague request such as “test this product for Europe and the US” from turning into an expensive but incomplete report set.
| Decision field | What must be fixed before booking | Why it matters |
|---|---|---|
| Evidence purpose | Regulatory, certification, buyer, claim, engineering or quality purpose | Determines who must accept the output and what form of report is needed |
| Target market / scheme | Country, region, certification scheme or customer programme | A technically correct test is not automatically acceptable everywhere |
| Requirement set | Applicable regulation, standard, method, edition and deviations | Wrong editions or missing clauses can invalidate reuse |
| Product family | Base model, variants and critical differences | Determines whether one worst-case sample can support several variants |
| Sample basis | Prototype or production unit, quantity, conditioning and accessories | Changes representativeness and the meaning of the result |
| Laboratory acceptance gate | Required accreditation scope, regulator/scheme recognition and report format | Prevents testing at a lab whose result the recipient cannot use |
| Change trigger | Changes that require review, delta testing or full retest | Controls whether the report remains applicable after design or supply-chain changes |
How to check whether a laboratory is suitable
ISO/IEC 17025 is the international standard used for testing and calibration laboratory competence. Accreditation is valuable evidence of competence, impartiality and consistent operation, but the label “ISO/IEC 17025 accredited” is not enough by itself. Check the laboratory's actual accreditation scope: the site, test field, method or standard, and any limitations relevant to your product.
Global ACI, which assumed the international accreditation roles previously associated with ILAC and IAF on 1 January 2026, operates a multilateral recognition arrangement in which Testing is a main scope and ISO/IEC 17025 is one of the normative documents used within that scope. That framework supports confidence in accredited results across borders, but it does not override a law, regulator, scheme owner or buyer that imposes additional recognition conditions. Acceptance therefore needs two checks: competence/accreditation and recipient-specific recognition.
- Ask for the accreditation certificate and current scope, not only a logo.
- Confirm that the test method or relevant technical field is within scope for the laboratory location that will perform the work.
- Ask whether subcontracting is planned and, if so, which tests will be subcontracted and under what status.
- For a regulated or scheme-controlled route, verify the laboratory against the authority's or scheme owner's current recognition rules before shipment of samples.
- Confirm whether the final report must carry an accreditation symbol, prescribed identifiers, photographs, model lists, uncertainty information or a specific statement of conformity.
Sample planning: the report is only as useful as the sample identity
A pass result on the wrong sample does not establish conformity of the product you later manufacture or sell. The sample plan should connect each tested unit to the bill of materials, drawings, firmware, labels, power supplies, accessories and other features that can affect the tested requirement.
- Prototype versus production: state clearly whether the sample is a prototype, pre-production unit or normal production unit.
- Worst-case selection: when several variants are intended to share evidence, document why the selected configuration represents the highest relevant test stress. The laboratory or scheme may need to agree with that rationale.
- Quantity: ask the laboratory to identify destructive and non-destructive tests, reserve samples and any sequencing dependencies before you ship.
- Conditioning and preparation: record preconditioning, charging, ageing, software settings and installation configuration where they affect the method.
- Traceability: use sample IDs and keep photographs and configuration records so a later reviewer can identify exactly what was tested.
Can an existing test report be reused?
Report reuse is a technical and acceptance decision, not a simple question of report age. A report may remain useful for one purpose and be insufficient for another. Apply the following screen before assuming that prior testing can be carried forward.
| Change or new use | Typical review question | Safe next step |
|---|---|---|
| Same model, same configuration | Is the same requirement, method edition and recipient still applicable? | Verify current acceptance and document no material change |
| New variant | Does the variant change any safety, EMC, chemical, mechanical, thermal or performance characteristic relevant to the test? | Perform a documented family/worst-case assessment; obtain lab or scheme confirmation where required |
| Component or material substitution | Does the substituted item affect a tested parameter or certified critical component? | Technical impact review; delta or full retest if the affected requirement cannot be closed analytically |
| Firmware/software change | Can the change alter operating modes, emissions, protective functions, power, connectivity or measured performance? | Compare functions and affected clauses; retest the impacted modes if needed |
| New standard edition | What changed between editions and is a transition rule in force? | Gap analysis against changed clauses before relying on the old report |
| New country or scheme | Does the new recipient accept the method, laboratory and report format? | Check market/scheme acceptance before claiming reuse |
Failure and retest logic
A failed test should trigger controlled engineering analysis, not an undocumented “fix and rerun”. Freeze the failed sample configuration, identify the exact failing clause or performance criterion, record the modification, and determine whether the modification can affect tests that previously passed.
- Capture the failure condition, measured result and sample configuration.
- Identify root cause or, if root cause is not yet known, define the leading hypotheses.
- Record the corrective design or process change.
- Map the change against every previously tested requirement that it could influence.
- Agree with the laboratory or applicable scheme whether a limited retest, regression set or full test sequence is needed.
- Keep the failed result, engineering decision and retest evidence linked in the technical file.
Do not assume that repeating only the failed clause is sufficient. For example, a component change intended to solve a thermal failure may also affect electrical safety, EMC or performance. The retest boundary follows the impact of the change, not merely the original failure code.
Test-driven change control after a pass
Once a report is issued, maintain a change log that asks a testing question before a design or supplier change is released. This is especially important when several models share one evidence set.
| Change type | Retest risk | What to document |
|---|---|---|
| Safety-critical component, power supply or protective device | Often material | Old/new specifications, approvals, ratings and affected clauses |
| Material formulation, coating, adhesive or restricted-substance source | Can be material for chemical, flammability or durability evidence | Composition/specification change and affected analytical or physical tests |
| Enclosure, layout, antenna, shielding or cable change | Can affect safety, EMC, wireless or mechanical results | Drawings, dimensions, routing and affected test modes |
| Firmware or software | Depends on functions controlled | Version, release notes, affected operating states and protective functions |
| Supplier/process change with nominally identical part | Depends on controlled characteristics and evidence basis | Equivalence evidence, tolerances, material/process differences |
| Purely editorial label/manual change | Usually low for laboratory retesting, but may affect regulatory information | Confirm no technical or required-information impact before closing |
What to send the laboratory in the first technical brief
- target countries, schemes or customer programmes;
- product description, intended use and user environment;
- model/variant matrix and the differences between variants;
- drawings, bill of materials and critical component list appropriate to the test;
- firmware/software version and instructions for entering all relevant modes;
- existing reports and the exact reuse question you want answered;
- requested standard/method and edition, if already determined;
- sample quantity, accessories, consumables and special installation needs;
- required report language, accreditation/recognition status and recipient;
- known upcoming product changes so the laboratory can flag likely retest consequences.
If you need execution support across laboratories and test programmes rather than only the evidence-selection logic, see product testing services. If the key question is the end-to-end approval route rather than the laboratory evidence itself, use product certification services. For classification and gap analysis before testing, see product compliance consulting; for multi-country routing, see global market access services.
Official reference points for laboratory evidence
- ISO/IEC 17025:2017 — General requirements for the competence of testing and calibration laboratories (ISO states that the 2017 edition was reviewed and confirmed in 2023 and remains current).
- Global Accreditation Cooperation Incorporated — About Global ACI and Global ACI MRA scopes for the current international accreditation framework.
- IECEE CB Testing Laboratories as an example of a scheme in which laboratory status and technical scope are scheme-specific.
Always verify the current rule of the receiving regulator, scheme owner or customer. An accreditation or MRA framework can support confidence in results; it does not by itself prove that every recipient must accept every report.
Frequently asked questions
Is an ISO/IEC 17025 accredited laboratory always mandatory?
No universal rule makes accreditation mandatory for every product test. The requirement depends on the law, conformity-assessment route, scheme, buyer or purpose of the test. Where accreditation is required or commercially important, verify the laboratory's actual scope and any additional recognition rule.
Can one test report cover several models?
Sometimes, if the models can be justified as a family and the tested sample represents the relevant worst case. The differences must be documented and assessed against each requirement. Do not assume that a shared brand name or enclosure automatically makes reports transferable.
Does an old passing report remain valid after a component supplier changes?
Not automatically. Compare the controlled characteristics of the old and new component and identify which tested clauses could be affected. A documented equivalence assessment may be enough in some cases; in others, delta or full retesting is necessary.
Should we retest only the clause that failed?
Only if the corrective change cannot affect other results and the applicable laboratory or scheme accepts that retest boundary. The safer method is to perform an impact assessment first and derive the regression set from the change.
Can a report from an accredited laboratory be used in another country?
Accreditation under an internationally recognised arrangement can support cross-border confidence, but acceptance is not automatic. The destination regulator, certification scheme or buyer may impose its own laboratory recognition, standard edition, national deviation or report-format requirements.