Laboratory interior at GZ Cosmetics Lab in Guangzhou

Quality & Testing

What GZ Cosmetics Lab checks during cosmetic development and production, what a test report should contain, and where third-party testing fits.

Evidence before claims

Testing is useful only when it is attached to an identified formula, batch, method and acceptance criterion. A certificate that says pass without those fields is difficult for a brand, safety assessor or retailer to audit.

This page explains the testing scope we can document during development and production. The final plan still depends on product type, claims, packaging and destination market; there is no single universal panel for every cosmetic.

named quality checkpoints

IQC, IPQC, FQC and OQC

Batch

traceability

finished units traced to raw-material lots

Per batch

batch-specific evidence

a COA belongs to the production run it names

On request

commissioned testing

laboratory and method agreed before quotation

Redacted public examples

Representative testing documents

The two documents below are real examples from our own files, published with customer, report, batch, issuing-body and signatory details masked. They are here so you can see how the paperwork is structured before you ask for it, not as evidence about any product other than the one each document names.

Read them as two different kinds of record. A commissioned test report answers questions about one submitted sample against the parameters listed in that report. A batch Certificate of Analysis records what one production run measured against an agreed specification. Neither is a certificate of certification, and neither describes the factory as a whole.

Redacted public example

Commissioned test report

Rosa Damascena Flower Water (大马士革玫瑰纯露), submitted by the client for commissioned testing.

Report issued: 12 September 2026
Sample received: 20 August 2026
Testing period: 20 Aug – 10 Sep 2026
Sampling method: Client-submitted sample

Scope shown on the document

Microbiological parameters, including total aerobic count, yeast and mould, thermotolerant coliforms, Staphylococcus aureus and Pseudomonas aeruginosa.
Lead, arsenic, mercury, cadmium and methanol.
Characteristic component profile used to check the authenticity of the botanical raw material.

Example values visible in the document

Methanol: 9.6 mg/kg
Phenylethyl alcohol: 0.047%

What it concludes. For the parameters listed in the report, the results met the acceptance requirements stated on that report. The report itself records that its results relate only to the sample submitted and are not an overall assessment of mass-production quality.

Limits of this document. This covers one submitted sample and the parameters printed on the report. It cannot be extended to other raw materials, other formulas, finished products, other batches or other markets. The issuing body is masked in this public copy, so it should not be read as an accredited or independent laboratory report, and it is not a factory certification.

Open the full-size redacted preview →

Redacted public example

Batch Certificate of Analysis (COA)

Rosa Damascena Flower Water (大马士革玫瑰纯露), one identified raw-material production run.

COA issued: 18 September 2026
Production date: 15 August 2026
Expiry / retest date: 14 August 2028
Storage basis: Unopened, dark, below 25°C

Scope shown on the document

Product identity: INCI name, botanical species, stated origin in the Rose Valley of Bulgaria and a low-pressure steam distillation process.
Physicochemical parameters including appearance, odour, pH, relative density, refractive index and characteristic component content.
Microbiological parameters against the specification set for this material.
Contaminants and heavy metals, plus a multi-residue pesticide screen.

Example values visible in the document

pH: 5.2
Relative density: 0.998
Refractive index: 1.3345
Phenylethyl alcohol: 0.048%

What it concludes. For this batch, the physicochemical, microbiological and contaminant parameters met the specification listed on the document, and the pesticide screen row records that none of the 480 parameters tested exceeded the limits referenced on the document. The COA states that its results apply to the batch it names.

Limits of this document. A COA is a batch record, not a certificate or a third-party certification. It describes this production run against an agreed specification and says nothing about future batches. It also does not replace finished-product stability, packaging compatibility, preservative challenge testing, a safety assessment or destination-market documentation, all of which belong to the finished product rather than to the raw material.

Open the full-size redacted preview →

Three kinds of evidence buyers tend to merge

Most document confusion in supplier qualification comes from treating these three as interchangeable. They sit at different levels: one describes a system, one describes a sample, one describes a batch. A complete file usually needs all three, and none of them can fill a gap left by another.

ISO 22716 certificate

Proves. That a named legal entity and production site were audited against the cosmetics GMP standard, within the scope and validity period printed on the certificate.

Does not prove. Any test result for an individual product or batch. A GMP certificate is a system-level document and cannot stand in for a COA or a test report.

Commissioned test report

Proves. That a named sample was tested by a specified method for the parameters listed, with the results recorded on the report.

Does not prove. That the whole factory, the whole product range or every future batch behaves the same way. It establishes nothing beyond the tested sample and the listed parameters.

Batch COA

Proves. What one identified production run measured against an established specification, with the production date and storage basis recorded.

Does not prove. That a later batch will match, or that the finished product is approved for sale in any market. Market approval is a separate regulatory route.

Both examples are published in redacted form. They are intended to help you understand document structure and ask better questions during qualification. For your own project we still need to confirm the actual product, batch, test methods, destination market and document permissions before agreeing what can be issued and shared.

Development evidence is product-specific

A cleanser, emulsion, active serum, fragrance and sunscreen do not need identical evidence. We scope physical, chemical, microbiological, packaging and claim-related work from the formula and market route, then identify what is standard, what needs an external laboratory and what belongs to the brand or local responsible party.

Inside the testing workflow

The laboratory environment supports development and routine quality work; project-specific accredited testing is scoped separately.

Sampling only becomes evidence when the batch, time, condition and responsible record stay connected to the result.

The core checks we document

Physical and chemical

Appearance, colour, odour, pH and viscosity where relevant, recorded against an approved specification rather than a one-word conclusion.

Microbiological

Product-specific microbiological quality and preservation evidence. High-water and repeated-use formats require different attention from low-water systems.

Stability

Conditions, intervals, packaging arms and acceptance limits are defined in the protocol. Accelerated work supports risk assessment and is not presented as an automatic conversion into shelf life.

Packaging compatibility

The final formula is assessed with the intended component assembly, including pumps, droppers, liners and airless mechanisms where used.

Claims support

The evidence plan follows the exact wording and target market. A formula specification alone does not substantiate every efficacy claim.

What a usable report contains

Ask for the formula and batch reference, component identification, method or protocol, dates, conditions, interval results, acceptance criteria, deviations and responsible sign-off. Our packaging compatibility and stability guides show how to read those fields before accepting a result.

Before testing

Agree the protocol and acceptance criteria before results are visible.

During testing

Preserve interval data and observations so a trend can be reviewed.

After testing

Keep the conclusion narrow: it covers the tested formula, batch and packaging configuration, not every future substitution.

Two reports buyers should know how to read

A stability conclusion is usable only when the sample, packaging, conditions, criteria and sign-off are visible.

Packaging evidence attaches to the final formula and complete component assembly, not to a generic material name.

Where external laboratories fit

External or commissioned testing can be arranged when a buyer, retailer or market file requires evidence from outside our own quality system. The laboratory, method, sample and the report you receive should be agreed before quotation, because those choices depend on the product, the claims and the destination market rather than on a standing arrangement with one named provider. We do not describe our internal laboratory as CMA- or CNAS-accredited, and an external report does not replace the manufacturer's own batch controls.

How the test plan changes by product format

Water-based cleansers and toners

Preservation, microbiological quality, pH and packaging closure deserve early attention. A rinse-off claim does not remove the need to establish safe use and a stable finished product.

Emulsions and creams

Viscosity, appearance, odour and phase behaviour are reviewed over time and under selected conditions. The jar, tube or pump can change exposure and dispensing performance.

Active serums

The plan follows the active's known sensitivities and the claims being made. Light, oxygen, pH, packaging contact and any relevant assay are considered rather than assuming that a high ingredient percentage proves performance.

Fragrance and essential-oil formats

Volatility, odour change, discolouration, component interaction and transport classification can matter more than viscosity. The selected bottle, seal and pump remain part of the tested system.

Masks, patches and sachets

Individual seals are part of the preservation and delivery system. Seal integrity, fill distribution, material compatibility and storage orientation need to be visible in the protocol.

Sunscreen and regulated claims

The intended market determines the regulatory route and accepted evidence. An SPF figure from one method or market must not be assumed to support every destination.

A stage-gated evidence plan

Testing is most efficient when the decision points are set before expensive packaging and production commitments are made. Early screening can remove unsuitable formula or component options, but it should not be presented as final-product evidence.

1. Brief review

Identify markets, claims, formula risks, pack type and customer-specific standards. This determines which questions need answers before sampling.

2. Development screening

Compare candidate formulas and components, document why one route is selected and record limitations that must be confirmed later.

3. Final configuration

Lock the formula code, intended component assembly, specification and protocol before artwork or tooling creates commercial pressure to accept marginal data.

4. Representative confirmation

Use representative product and the intended manufacturing and filling process. Confirm that earlier laboratory assumptions still hold at scale.

5. Release and ongoing evidence

Connect batch release documentation, retained samples, real-time observations and change control so evidence remains useful after launch.

What happens when a result is marginal or fails

A failed result is not automatically a failed formula. The investigation compares the control, market pack, orientation, interval trend, process history and component specification. If glass and the market pack both change, the formula or process moves higher on the investigation list. If only the market pack changes, component interaction, permeation or sealing deserves attention. If the formula remains within specification but the dispenser stops working, the pump, gasket, dip tube and dose data become the priority.

Confirm

Check sample identity, method, equipment, condition and whether the stated component was actually used.

Contain

Prevent an unresolved configuration from moving into tooling, routine production or shipment.

Investigate

Use comparison data and batch history to separate formula, component, process and method causes.

Change and repeat

Document the chosen correction and define which parts of the evidence package must be repeated before approval.

The evidence package a brand should retain

A usable handover normally links the approved formula and specification, component specification, study protocol, interval results, final report, relevant third-party reports, batch documentation and approved claims. The commercial agreement should say what is supplied, in what language, under what confidentiality terms and whether the brand can provide it to a safety assessor, retailer or regulator. Agreeing those rights after a market authority asks for the file is too late.

Frequently asked questions

Do you use one standard test panel for every cosmetic?

No. The panel follows the product, claims, packaging and destination market. The protocol should explain why each measurement and condition is included.

Can I receive the underlying data?

The project scope should state whether you receive interval data, the final report and component specifications. Set that expectation before testing begins.

Does a test report mean the product is approved?

No. A test report covers the sample and the parameters named in that report. It is not a market approval, it does not certify the factory or the wider product range, and it does not replace local filing responsibilities or a safety assessment.

Scope the evidence before you approve the sample

Send the formula type, claims, packaging and destination markets. We will separate standard factory documentation from project-specific and external testing.

Request a Quote WhatsApp Us

Response within 24 hours · WhatsApp available · Factory visits welcome