Call(WhatsApp Us)
+8615360806100Email Us
cherry@laeyo.comSee the factory by live video walkthrough
Book an on-site factory visit in GuangzhouA color shift found after customs clearance is rarely a "bad batch" story on its own — it is a records story. By the time the pallet reaches your warehouse, the product has passed…
A color shift found after customs clearance is rarely a “bad batch” story on its own — it is a records story. By the time the pallet reaches your warehouse, the product has passed through the factory gate, a freight forwarder, a port, a customs hold, and a truck. The only way to determine whether the change originated in manufacturing, in the packaging, or in the supply chain is to reconstruct the batch from controlled documents and compare the arrived goods against the retained sample. This guide shows sourcing managers which batch records and shipping records actually decide the outcome, in what order to read them, and what to request from the factory so a claim does not stall for weeks.
“The serum looks more yellow than the approved sample” is a perception. The procurement question is narrower and answerable: did this batch meet its approved appearance standard at release, and is the arrived product materially different from the retained sample of that same batch? Everything else — who pays, whether to rework, whether to re-order — flows from those two answers.
Two details make this harder than it looks:
Naming the category in your first email changes the investigation your supplier runs, and it changes which documents they pull first.
A defensible batch record set is built so that one finished lot number links back through bulk, formula version, process version, line, raw material lots, packaging lots, operators, and QC and micro results. If any link in that chain is missing, no one can say with confidence whether the batch was in control. Request the set below for the specific lot that changed color — not for the SKU in general.
| Record | What it proves | What “acceptable” looks like |
|---|---|---|
| Certificate of Analysis (COA) | That the batch was tested against a specific, versioned release specification for the intended market | Shows batch number, SKU, spec version (e.g. SPEC-V1), test item, result, document version, market, and an authorized signature — all matching the lot you received |
| Finished batch record (BPR) | That the actual process followed the approved process, with in-process checks signed off | A SKU-specific record, not a generic template; process version stated; IPC results recorded during the run |
| Bulk batch and line link | Which bulk lot and which line produced the fill | Finished lot traceable to a bulk lot and a named line, with no gaps |
| Formula version | Which formula was actually used, and whether it matches the approved version | Formula version on the record equals the approved version; no unauthorized raw material substitution |
| Raw material lots | Which supplier lots went into the batch, and whether their incoming COAs were accepted | Every RM lot linked, incoming COA on file, supplier on the approved list |
| Packaging material lots | Which bottle, pump, cap, and label versions were used — often the real cause | PM lots linked; supplier reference sample on file for material color, especially for post-consumer recycled resin where lot-to-lot tint varies |
| Retained sample | A physical baseline of that exact batch, stored under controlled conditions | Available, same lot number, storage conditions documented, comparable against the arrived goods under fixed lighting and background |
| Microbiological results | Whether preservation held — relevant when a color shift arrives with an odor signal | Result linked to the lot and evaluated against the microbiological specification applicable to that market |
| Stability records | Whether appearance drift was already known and characterized for this formula and pack | Appearance assessed at defined intervals, including accelerated and light-exposure conditions, with photographs or colorimeter readings against an approved color board |
| Deviation and out-of-trend (OOT) records | Whether anything unusual was already flagged at the time of release | No open deviation on the lot; any trend signal (for example a slow downward drift in viscosity still inside spec) documented with a trend review and disposition |
Ask for the batch release decision as a document, not a verbal confirmation. A controlled release decision lists its inputs — raw material and packaging release, batch record, in-process checks, bulk results, finished tests, micro, packaging function, deviations, line clearance — and ends in one status: release, conditional release (only where the procedure permits it), hold, reject, or rework. Quality holds the final release authority; a sales delivery date cannot override a quality hold. Each conclusion should point to a result ID, deviation ID, COA, and retained sample. If you are shown a COA with no release decision behind it, you have a test report, not a release.
Shipping records cannot prove a formula is stable, but they can rule light and heat exposure in or out. Request these for the exact container and the exact window between factory dispatch and your receipt:
A useful sanity check: if only the outer cartons or only the top tier changed color, the records should show why. If every unit changed uniformly, exposure is a weaker explanation and the batch or the pack becomes the lead hypothesis.
Discoloration in a fragranced skincare product has a small set of likely causes: fragrance-driven color shift, oxidation, reaction with the base, light exposure, or migration from packaging. The fastest way to separate them is a controlled comparison the lab can run without a full re-development cycle:
If the blank stays stable and the fragranced sample shifts, the fragrance is the driver, and the conversation moves to fragrance specification and dosage rather than to batch fault. If both shift, look at the base, the antioxidant or chelator system, and the headspace. If the shift only appears where liquid contacts a component, packaging is the lead.
Also check whether an existing trend signal was already visible before shipment. A parameter drifting toward the edge of its specification while still passing — a viscosity reading moving toward the lower limit, for example — should already have triggered a documented trend review. Ask whether any such review was open on this lot: it changes the discussion from “unknown” to “known and managed.”
This guide is informational and does not constitute legal advice; market-specific labeling, claims, and reporting obligations should be confirmed with your regulatory counsel.
Send this as-is, with the lot number inserted. It is scoped tightly enough that a competent quality team can answer it in one pass.
Subject: Appearance deviation — Lot [LOT NUMBER], SKU [SKU NAME], delivered [DATE]
Share the COA for this lot, including specification version and document version.
Confirm the market this lot was released against and the applicable spec version.
Provide the batch release decision for this lot with its status and approver.
Attach the finished batch record index showing bulk lot, line, formula version, and process version.
List all raw material lots used, with incoming COA status for each.
List all packaging lots used for bottle, pump, cap, and label, with supplier names.
Confirm whether the retained sample for this lot is available and where it is stored.
Provide photographs of the retained sample under fixed lighting and a fixed background.
State whether any deviation was open on this lot at release.
State whether any out-of-trend review was open on this lot, and attach the trend review.
Provide stability appearance data for this formula version at each tested interval.
Confirm whether blank-versus-fragranced color comparison data exists for this fragrance.
Provide micro results for this lot against the specification for the destination market.
Provide packaging compatibility and function test results for this pump and bottle combination.
State whether label, coating, or gasket migration was assessed for this pack.
Confirm the retain-sample policy and retention period in our quality agreement.
Provide the dispatch date, container number, and seal number for this shipment.
Provide transit temperature records, including any data gaps.
Provide the customs hold dates and facility, and the storage conditions during hold.
Confirm warehouse storage conditions after clearance and the delivery date.
Open a linked complaint investigation and confirm the lot's quality trend status.
State whether this complaint is being assessed as a safety-related event.
Confirm in writing whether the lot is on hold pending investigation.
Not on that basis alone. A Pass confirms the lot met the approved appearance standard at release, against a specific spec version, for a specific market. A rejection case needs a controlled comparison between the arrived goods and the retained sample of the same lot, plus confirmation that the spec version and label artwork version were correct for your market. Get both before you escalate.
Localized staining is a packaging signal far more often than a formula signal. Ask for the pump and gasket material specification, the packaging compatibility report, and the packaging lot numbers for this batch. A “pump not dispensing” or residue complaint on the same lot strengthens that direction and should be investigated together, not separately.
That is a decision made with your quality lead, your regulatory advisor, and the market requirements — not a decision the factory makes for you. What you should insist on now: the lot status in writing, a documented investigation, a similar-complaint search, and a clear statement of whether any safety-related event has been reported. A sales delivery schedule should never be allowed to override a quality hold.
Against an approved color board or with a colorimeter, under fixed lighting and a fixed background, comparing the arrived unit to the retained sample of the same lot. Phone photos taken under warehouse lighting are useful for triage but are weak evidence in a commercial discussion. Ask for instrument readings where they exist.
Lock four things before the next purchase order: the specification version per destination market, the approved material color standard for the bottle and pump (with a physical reference sample), a defined retain-sample policy, and the requirement that each SKU carries its own batch record and process version rather than a shared template. Those four controls cover most of the gaps that turn a color question into a month-long dispute.
If you are moving a color-sensitive skincare line into a new market, the fastest way to reduce this risk is to fix specifications, packaging standards, and documentation deliverables before the first production run — not after a customs hold. Send your lot details, destination market, and pack specification and we will map the records and test plan your SKU needs.