SAMPLE BASELINE / 01
Make the sample review repeatable
Another reviewer should be able to identify the same sample, repeat the key checks and see exactly what was approved.
Identify the sample and its source
Record sample type, supplier, workshop, model, ID, revision, quantity, dispatch date, receipt date and reviewer. Photograph labels and features that distinguish it.
Give each version its own ID. Mark older versions as superseded; “latest sample” is not a stable reference.
Fix the reference documents
List the specification, drawing, bill of materials, colour, artwork, labels and packing instructions by ID and revision. State which controls if they disagree.
Define the product specification before judging the sample. Save the reviewed files so later changes remain visible.
Use measurable criteria
Record intended use, key functions, dimensions, tolerances, materials, assembly checks and test methods. Note equipment, conditions, units and reviewer where they affect the result.
Use written criteria for measurable requirements. A physical sample helps with appearance and feel; it cannot replace material, performance or destination requirements.
Check appearance and packing
Check finish, colour, fit, moving parts, accessories, artwork, labels and packing against the agreed scope. Photograph the reviewed state and record lighting or other relevant conditions.
List temporary materials, manual finishes and prototype-only features. Record each accepted difference, its owner and the condition for closing it.
Match test reports to this sample
Match report number, sample ID, model, material, dates, method and result. Check that the laboratory’s scope covers the required test and that the report answers the intended product question.
Laboratory competence does not prove who supplied the item or how future lots will perform. Check certificates and reports.
Record differences and the decision
Log each difference, required value, observed result, impact, evidence, owner, due date and action. Choose a clear state: rejected, revision required, conditional or approved reference.
State what conditional approval allows. Keep required hold points open until an authorised buyer records their closure.
Label and keep the reference
Where practical, keep matching buyer and supplier samples. Label them with ID, revision and approval date. Record the custodian, storage, seal condition, photographs and limits caused by ageing or damage.
The written specification still controls characteristics that one physical item cannot preserve, such as composition, hidden construction or long-term performance.
Connect approval to production and inspection
Link the approved sample to the PO, specification, supplier response and inspection brief. Name the revision, site, quantity or stage, remaining holds and person who may release work.
Review changes to the site, materials, tooling, design, packing, test evidence or destination. Decide whether a new sample or test is needed before affected work starts.
One approved sample and a later lot-sampling decision answer different questions. Use ISO 2859-1:2026 only where the agreed lot-sampling plan calls for it.
Next: purchase order, production inspection and pre-shipment inspection.
APPROVAL RECORD / 02
Minimum fields for a reviewable sample decision
| Control area | Record | Evidence | Hold when |
|---|---|---|---|
| Identity | Supplier, site, SKU, sample type, ID, revision and dates | Labels, dispatch record and dated photographs | The reviewed item cannot be distinguished from another version |
| Comparison pack | Controlling specifications, drawings, materials, artwork and packaging revisions | Versioned files and conflict rule | Two references disagree or can be replaced without approval |
| Function and measurement | Intended use, methods, equipment, tolerances and results | Check sheets, measurements and reviewer record | A critical requirement has no objective acceptance method |
| Appearance and pack-out | Finish, assembly, accessories, branding, labels and packaging | Dated photographs and approved exceptions | A prototype-only or temporary feature is not declared |
| Test evidence | Report, tested-item identity, method, result and laboratory scope | Source-checked report and sample match | The report cannot be connected to this sample and decision |
| Exceptions | Requirement, observed difference, impact, owner, due date and action | Issue log and closure evidence | An open exception is hidden inside a general approval |
| Retained reference | Buyer/supplier copies, labels, custodians, storage and condition | Signed labels, seal and custody photographs | The reference is missing, changed or no longer comparable |
| Handoff | Decision state, authorised owner, released work, PO and inspection links | Signed revision and supplier acknowledgement | Production can start from an ambiguous or superseded version |
DECISION STATES / 03
Separate sample acceptance from production release
| State | Meaning | Permitted next action |
|---|---|---|
| Development / rejected | The item does not yet meet the comparison pack or intended-use criteria | Request a corrected, newly identified sample |
| Conditional | Named differences are accepted only for evaluation or a limited purpose | Close the written conditions; do not infer mass-production release |
| Approved reference | The exact sample revision is accepted as one controlled comparison reference | Complete the PO, evidence and release gates |
| Production released | An authorised owner has linked the approved revision to defined work and closed mandatory holds | Start only the named stage, quantity and revision |
| Superseded / reapproval required | A relevant change or later revision has replaced the prior baseline | Hold affected work until impact review and a new decision |
Check where the sample was made
Branch and investment records do not prove where a sample or production lot was made. Ask for both site names and addresses; review any difference before production release.
| Registered structure | Companies | What was returned |
|---|---|---|
| Branch records returned | 8 of 45 | Branch registration records |
| Investment records returned | 18 of 45 | Investee company records |
| Neither type returned | 24 of 45 | Neither record type in this query |
The groups overlap: 5 companies returned both branch and investment records. “Neither” means neither type was returned in these queries; it does not prove one address or no other site.
Source: 45-company NHTSA-linked sample, licensed-platform queries on 21–22 August 2026. This selected group does not represent all suppliers.
METHOD / 04
How this checklist was prepared
The ISO sources below cover version control, clear buyer requirements, laboratory competence and lot sampling. The eight checks and decision states are Currawong’s workflow for recording sample reviews.
We have not inspected your sample or production. This guide is not product testing, a legal or compliance opinion, supplier approval or production release.
PRIMARY SOURCES / 05
Official sources used for this guide
Source notes reviewed on 5 September 2026. Recheck applicable requirements for your product and destination.
- ISO 10007:2017: Guidelines for configuration managementGuidance on identifying product versions and changes.
- ISO: ISO 9001 in the supply chainBuyer guidance on requirements and QMS claims.
- ISO/IEC 17025:2017Requirements for laboratory competence.
- ISO 2859-1:2026AQL-indexed lot-by-lot acceptance sampling.
COMMON QUESTIONS / 06
Common questions
- What is a “golden sample”?
- Teams often use the term for an approved physical comparison reference. The label alone is not a control: give the item a unique ID and revision, connect it to written criteria, record custody and state exactly what decision it supports.
- Can the sample replace the product specification?
- No. It can represent selected visible or tactile characteristics, but written requirements should control dimensions, tolerances, materials, hidden construction, performance and other measurable or destination-specific criteria.
- Does a passing laboratory report approve the sample?
- Only the named tests and results can support the decision, and only when the tested item is matched to the reviewed sample. The buyer must still assess the rest of the specification and destination requirements.
- Can I approve with minor differences?
- You can record a conditional state, but list every accepted difference, impact, owner and closure condition. Do not let conditional acceptance silently become production release.
- What happens after a material, component or factory change?
- Record the impact, supersede the affected baseline and require a new revision or reapproval before affected work proceeds.