How Should Buyers Document Differences Between an Approved Sample and Production Product?

Answer first

Buyers should document an observed difference by comparing a traceable approved sample with a traceable production unit, describing the difference precisely, attaching evidence, assigning an owner, and recording a disposition before the affected SKU is treated as accepted. The record should answer five practical questions: what was approved, what was observed, which units are affected, why does the difference matter, and who decided the next action?

The goal is not to prove that every visible variation is a defect. It is to prevent a team from silently treating a changed product as the one it originally approved. A controlled comparison separates harmless variation from a change that needs clarification, rework, revised product information, further review, or a new approval. It also gives procurement, quality, product, packaging, and content colleagues one version of the facts.

Start with an approved sample difference record that identifies the baseline by SKU, revision, approval date, sample identifier, and the specific packaging, instructions, components, and configuration that were approved. Compare like with like. If the production item is from a defined batch or production date, capture that identity. If the reference sample cannot be linked to an approval decision, use it for context but do not call it the approved baseline.

A record is evidence for an internal decision; it is not, by itself, a declaration that a product is compliant, safe, saleable, or suitable for every market. Exact SKU details remain subject to the approved sample, current instructions and packaging, supplier records, retailer rules, importer obligations, and destination-market review.

Define the baseline before examining production

An approved sample is useful only when the team can tell exactly what it represents. Create a short baseline sheet at approval, rather than relying on a photo folder or a person’s memory later. Link the physical sample or controlled digital record to the commercial SKU, supplier item code where applicable, approval date, approver role, intended destination or channel if relevant, and the documents reviewed with it.

The baseline should distinguish the object from its surrounding materials. In practice, that means recording the product version, visible construction, materials or components where supplied, colourway or artwork revision, packaging revision, inserts, instructions, warnings, label layout, and any accessories that were actually included in the submitted sample. It should also note unknowns. “Not verified in this sample” is more useful than filling a field with an assumption.

Use dated, sample-ID-labelled photographs from repeatable views, but do not treat images as a substitute for an approved written baseline. They may not show composition, internal components, process settings, or an instruction revision.

ISO guidance explains that documented information can communicate information, provide evidence that planned work occurred, and preserve knowledge. It also recognises paper, electronic records, photographs, and a master sample as possible media. [1] That is a practical reason to keep the signed or otherwise controlled approval, photographs, specifications, and the retained reference sample connected rather than scattered across separate inboxes.

For EU toys, the Toy Safety Directive requires manufacturers to maintain series-production conformity and take design or characteristic changes adequately into account. Its technical documentation includes design-and-manufacturing information, components and materials, safety assessment, and evidence of production conformity. [3] This supports a clear buyer baseline but does not transfer manufacturer duties to every buyer.

Use stable identifiers, not descriptive names alone

Use controlled identifiers: internal and supplier SKU where available, sample ID, batch or lot identifier, and artwork, instruction, or document revision. Identify variants separately; an approval for one colour, language, pack configuration, or component source does not automatically cover another.

If no approved sample remains, gather the approval, dated images, approved files, and specifications, then state the gaps. Do not reconstruct certainty from a later unit.

Classify the difference before deciding what to do

First confirm the fact; then ask whether it affects an agreed requirement, identity, user-facing information, quality expectation, or destination-market evidence. A difference is not automatically unacceptable, but a small change can matter when it affects components, instructions, labelling, traceability, or existing records.

For US children’s products, the CPSC rule includes components, composition, and assembled interaction/function within product design. It treats certain design, process, and component-sourcing changes as material when they could affect compliance. [2] It is not a universal test; escalate changes in composition, source, construction, or process to the roles responsible for the applicable market review.

Avoid “minor” or “acceptable” at intake. First state the observation: “leaflet is revision B; baseline links revision A.” Then keep the supplier explanation separate from the reviewer’s conclusion.

A production product deviation should therefore be recorded as a condition against a named baseline, not as a vague message that production “looks different.” A difference may be accepted only under defined conditions, rejected, placed on hold for evidence, or escalated. The disposition must make the condition visible.

Need help setting up a practical comparison record for a Toywink B2B project? Email Toywink’s B2B team with the SKU, baseline reference, and the specific difference you need to review. Product details remain subject to the exact SKU, approved sample, current records, and destination-market review.

Build a record that can survive handover

Use one record per distinct change or tightly connected group of changes. A long email thread often mixes several facts, units, and decisions. A controlled record gives each issue an ID and lets people link it to evidence without rewriting the story every time.

At minimum, include the following fields:

• Record identity: record number, creation date, reporter, status, SKU and variant, supplier reference, production batch or unit identifiers when available.
• Approved baseline: sample ID, approval date, approval reference, approved specification or drawing revision, approved packaging/artwork/instruction revisions, and retained-sample location where one exists.
• Observed production state: date and place of observation, quantity or sampling context, factual description, affected part or document, and photos or files showing the difference.
• Impact questions: whether product identity, components, manufacture, pack contents, instructions, labels, claims, traceability, or destination-market evidence may be affected. Mark unanswered questions as open.
• Evidence and correspondence: supplier explanation, revised bill of materials or specifications if supplied, inspection observations, file checks, and links to controlled evidence. Record the document issue date and source rather than copying an unverified assertion into the conclusion.
• Decision control: required reviewers, decision date, decision and conditions, authorising role, corrective or follow-up action, due date where agreed, and closure evidence.

A controlled spreadsheet, quality record, or shared system can work. Prioritise access control, stable links, and current status. ISO guidance lists records of design/development changes and reviewed production changes, including authorisation and actions, as information that may need retention. [1]

Do not replace source evidence with narrative. Link the underlying file, photo, sample location, or correspondence, and name each attachment by SKU, revision, and date. Retain originals alongside annotated images.

Separate evidence, assessment, and decision

A durable record has three distinct layers. Evidence says what was seen or received. Assessment explains the question it raises against the baseline. Decision says who accepted, rejected, held, or escalated it, with conditions. Keeping these layers separate stops an early opinion from becoming a false fact.

For example, evidence can note a different packaging-file revision; assessment can say the approved artwork has not been matched to it; and the decision can hold pack imagery until the revision and importer review are confirmed. Printing does not itself prove approval.

This is where sample to production comparison becomes more than a visual sign-off. The record connects a physical observation to approved documents, evidence boundaries, and accountable next steps.

Run the comparison in a repeatable order

Confirm reference and production identities first. Capture IDs before opening packs or separating components. Record date, batch, line, or supplier reference only when available; never infer a batch from a carton style or file name.

Compare in a fixed order: SKU/configuration, physical product, included items, packaging, labels, instructions, and supplied documents. Check stated against observed revisions. Use repeatable views and a prepared field list.

Measure or test only against agreed methods and criteria. If the baseline lacks a dimension, tolerance, behaviour, or method, record the gap; do not create a pass criterion after seeing the result. Route required evidence or review to the right role.

For US children’s products, do not assume an old test report or certificate covers a new configuration. Under 16 CFR 1107.23, further testing depends on the nature of a material change and may use component-level results in specified circumstances. [2] The manufacturer and responsible experts must determine action for the exact product and market.

Use the difference ID on images, hold tags, and messages. For repeated differences, record sampling scope and known affected units without extrapolating beyond evidence.

Ask the supplier focused questions

Ask what changed, when, which variants or lots are affected, whether components or manufacturing steps changed, and whether labels, packaging, instructions, or documents changed. Request revision IDs and dates, not only screenshots.

A supplier answer is input, not automatic approval. Compare it against the baseline and requirements. If it reveals a new inconsistency, open or link a separate record rather than overwrite the first.

Make the disposition explicit and conditional

A decision should be legible to someone who was absent from the review. “OK” is inadequate. Use a controlled status, name the decision maker by role, and state the conditions that make the decision valid.

The product deviation documentation must state whether a decision is limited to a unit, batch, SKU variant, packaging revision, market, or channel. Do not copy an acceptance to another scope without review, or treat it as replacing any party’s obligation.

The EU Toy Safety Directive requires product identification through type, batch, serial, model, or another element, and requires manufacturers to retain technical documentation and the declaration for 10 years after market placement. [3] Exact obligations vary, but identifiers and retained evidence make a buyer record more useful later.

Comparison record checklist

Use this checklist at review and closure. Adapt it to the exact SKU and the roles involved; it is an internal control tool, not a universal approval template.

• The approved reference is identified by SKU, variant, sample ID, approval reference, and relevant document revisions.
• The production unit or production evidence is identified by the available batch, lot, date, supplier, or other traceability information.
• The comparison scope states which product, components, included items, packaging, labels, instructions, and files were checked.
• Every observation is factual, dated, and linked to clear evidence.
• Each difference has its own ID or an explicit link to the record that owns it.
• Unknown information and unverified supplier statements are labelled as open, rather than treated as facts.
• Potential effects on product identity, documents, packaging, instructions, traceability, and destination-market obligations have been routed to the appropriate roles.
• The decision says accepted, rejected, held, or conditionally accepted, and defines the precise scope and conditions.
• The required reviewers and authorising roles are recorded; a reviewer is not assumed from an email copy list.
• Corrective actions, required record updates, and evidence of completion are linked before the issue is closed.
• Content, sales, and retailer teams receive only the currently approved information for the exact SKU and scope.

Great Britain’s toy-safety guidance says manufacturers must keep continuing production conforming, retain technical documentation and the declaration for 10 years, and record complaints and non-conforming toys. [4] A buyer record supports communication and control; legal responsibility remains with the relevant economic operators.

Keep the change history usable after closure

Retain the original approval, observation, decision, evidence links, and later revisions. If the baseline changes, issue a new revision and cross-reference the prior record. Do not rewrite history so an old photo appears current.

A simple register can show record ID, SKU, change category, status, decision date, owner, and linked baseline revision. That is the operational quality review change log. It helps teams see repeated differences, open holds, and records that affect the same SKU without publishing sensitive evidence broadly.

For each resolved issue, define the current source of truth: which product specification, artwork file, instruction version, or SKU record may now be used by procurement, quality, content, and retail teams. This creates SKU change evidence without assuming that an internal file is sufficient for every marketplace, importer, or destination-market requirement.

FAQ

Is an approved sample the same as a production specification?

No. A sample shows what was reviewed; a controlled specification sets requirements in drawings, measurements, revision-controlled files, or other criteria. Use both. If they conflict, record and route the conflict to the authorised owner.

Should buyers record cosmetic differences that do not affect use?

Record any departure from the baseline, then decide proportionately. A cosmetic change may affect retailer imagery, colourway identity, pack matching, or agreed finish. Show scope, evidence, and decision; do not silently normalise it.

Can a supplier email be enough to close a difference?

An email can support evidence when it identifies SKU, revision, scope, sender, and explanation, but rarely closes a record alone. Link it to the baseline, observed unit, assessment, and authorised decision. Request controlled attachments for document or component changes.

Who should approve a change?

Involve roles that own the affected requirement: physical construction may need product and quality input, while packaging, warning, or instruction changes may also require content, importer, or market review. Define and record approval authority; it varies by SKU, agreement, market, and difference.

Does acceptance of a difference approve the item for every market or retailer?

No. Internal acceptance must state its scope. It does not promise compliance approval, marketplace or retailer acceptance, or suitability across markets; requirements vary with the product, importer, destination, retailer, instructions, packaging, and records.

What if production is already moving while a review remains open?

Record the known scope, notify responsible decision makers, and apply the agreed hold or escalation process. Do not mark it accepted to match timing pressure; a transparent pending decision preserves information for corrective action.

For an email-only Toywink B2B SKU-comparison discussion, contact info@toywink.com with the SKU, approved reference, production evidence, and decision needed. Review remains subject to the exact SKU, approved sample, supplier records, retailer rules, importer obligations, and destination-market review.

References