How to Audit a Glassware Manufacturer in China: A Buyer’s Checklist

A factory audit is not a certificate collection exercise. It is an evidence-gathering process. The question is not “what documents does this supplier hold?” but “can I verify that the production line running my order actually does what the documents claim?”

Most glassware audits fail because buyers treat them as a tour. They walk the floor, look at the samples, take photos, and leave with a stack of photocopied certificates. They never check whether the annealing lehr temperature logs match the production dates. They never pull a finished carton from the warehouse and trace it back to a raw material batch. They never ask to see a polariscope reading.

That gap — between what a factory presents and what a factory actually does — is where sourcing risk lives.

This checklist is built around verification, not inspection. It assumes you will ask for evidence, compare it against reality, and treat discrepancies as more informative than smooth answers.

For the broader framework covering testing standards, food-contact compliance, and how audits fit alongside product certification, read the pillar guide: The Glassware Importer’s Guide to Factory Audits, Testing Standards and Food Safety Compliance .

What Should You Request Before the Factory Visit?

The audit begins before you book a flight. Pre-visit document requests tell you whether the supplier operates with the kind of administrative discipline that glass manufacturing requires.

Legal and operational identity. Request the business license and confirm that the registered company name matches the name on the quotation and the bank account. A mismatch is a common sign of trading-company intermediation. Check the registered business scope — does it include glass manufacturing, or only trading? A factory that manufactures should show manufacturing in its scope.

Actual production location. Ask for the factory address and cross-check it against the address on the business license. If the supplier operates multiple sites, ask which specific building and line will run your order. An amfori BSCI audit report can cover a limited production unit rather than the entire legal entity; the report should identify the audited site explicitly. If the audit scope does not name the workshop that will produce your glassware, the report does not cover your order.

Production capacity evidence. Request recent production schedules, furnace campaign records, or order history that demonstrates the factory can handle your volume. For glass, this matters more than for many other categories because furnace scheduling is continuous. A factory that cannot show you its actual monthly output against its stated capacity is asking you to take its word.

Certification documents — with scope. Request ISO 9001, BSCI or SMETA reports, and any product test reports. But do not stop at the cover page. For ISO 9001, check whether the certificate scope names your product category and the manufacturing site. For social audits, note the audit date, the audited site, and the validity period. An audit report that is two years old, or that covers a different workshop, is not current evidence for your order.

Subcontracting disclosure. Ask directly whether any part of your order — forming, annealing, decoration, packaging — will be handled by a subcontractor. Glass factories sometimes outsource decoration or secondary processing. If the answer is yes, the subcontractor’s premises and quality controls are part of your supply chain, and they should be audited or at least documented.

What Should You Check on the Factory Floor?

Once you are on site, the audit shifts from documents to physical verification. Glass manufacturing has specific process points where quality is won or lost.

Raw material intake and storage. Ask to see the incoming inspection records for silica sand, soda ash, and cullet. For cullet specifically, check how incoming recycled glass is sorted and whether foreign material is removed before it enters the batch. Cullet contamination is a common source of defects in lower-tier factories. Look at the storage area: is raw material protected from moisture and contamination? Are different material types segregated?

Batch mixing and furnace control. Glass quality begins with batch consistency. Ask to see the batch recipe and the weighing records. Check whether the scales are calibrated and whether the weighing process is automated or manual. A factory that mixes by hand without recorded weights is a factory that will produce variable glass.

Forming and annealing. This is where you separate serious glassmakers from assemblers. Ask to see the annealing lehr and request the temperature logs for the past 30 days. For borosilicate, a typical annealing profile requires approximately 560°C ± 10°C for three to five hours glass pipe supplier audit checklist, Elfglass, 2026. Inconsistent annealing causes stress fractures that may not appear until weeks after delivery. Check whether the factory uses polariscopes to measure residual internal stress. A polariscope reading shows stress patterns in the glass under polarized light. If the factory has no polariscope and no stress-testing records, they are not controlling one of the most important variables in glass quality.

Quality control stations. Walk through the QC process from hot-end to cold-end. Look for:

• Dimensional checks — wall thickness, height, rim diameter, capacity

• Visual inspection — bubbles, stones, scratches, cord

• Functional checks — lid fit, seal integrity, pour performance

• Breakage and defect categorization — are rejects separated and labelled?

Ask how defects are classified and what the acceptable limits are. A factory that says “we inspect everything” without a defined AQL or defect classification is not running a controlled QC process. For glassware, critical defects such as cracks or seal-compromising chips should be held to an AQL of 0.025%–0.065%, major defects to 0.40%–0.65%, and minor defects to 1.50%–2.5% Generic Attributes Requirements Glass Vials, NGVF, 2016.

Lid and closure fit. For food storage jars, spice jars, and oil dispensers, the lid fit is a critical quality attribute. Ask to see the fitting check: does the lid seat evenly? Does the seal compress correctly? Is there a torque or vacuum test? A glass jar with a poorly fitted lid will fail in the field regardless of how good the glass body is.

Packaging and warehouse. The packaging area tells you how the factory thinks about breakage. Look for:

• Carton construction — single-wall or double-wall? Bursting strength specification?

• Internal dividers — are they full-height and corrugated, or thin and loose?

• Cushioning — is there consistent cushioning on all six faces of the carton?

• Palletising — are pallets stretch-wrapped and corner-protected?

In the finished goods warehouse, check whether product is stacked at safe heights and whether there is a designated non-conforming area for rejected batches. A factory that stores rejects alongside good product without clear segregation is a traceability risk.

Batch traceability. This is the single most important cross-check in the audit. Pick a finished carton from the warehouse — ideally from a batch produced for another customer. Ask the factory to trace it back to the raw material batch, the furnace campaign, the annealing run, and the QC records. If they cannot complete the trace within a reasonable time, their traceability system is not functional. A batch traceability check is a standard item in glass factory audit checklists: randomly select a finished product batch and trace it back to the corresponding raw material batch information, verifying the completeness and accuracy of the traceability chain Glass Pipe Supplier Audit, Elfglass, 2026.

How Do You Cross-Check System Documents Against Actual Production?

A quality management system certificate tells you that the factory has documented procedures. It does not tell you that the procedures are followed. The audit technique is to select a document, then verify it against the physical record or the actual operation.

Pick a recent QC record and match it to production. Ask for the inspection record for a batch produced in the last two weeks. Then go to the production floor or the warehouse and find that batch. Does the record match what you see? If the record says “100% inspected” but the warehouse shows no inspection tags, the system is not being executed.

Check the annealing records against the furnace schedule. Ask for the annealing temperature chart for a specific date. Then ask for the production schedule for the same date. Do they align? If the annealing chart shows a five-hour profile but the production schedule shows three shifts running through the same lehr at high throughput, one of those documents is not accurate.

Ask a line operator to explain a procedure. Pick a quality-critical step — say, the visual inspection of jar rims. Ask the operator to describe what they check and what they do when they find a defect. If the answer is vague or contradicts the written procedure, the training and control system has gaps.

Verify calibration records. Ask for the calibration certificates for measuring equipment: thickness gauges, scales, polariscopes, pressure-test equipment. Check that the calibration dates are current and that the equipment on the floor matches the equipment on the certificate. A factory with expired calibration on its measuring tools cannot credibly claim to measure quality.

What Is the Difference Between BSCI and SMETA — and Why It Matters?

This is where buyers frequently make a costly assumption. amfori BSCI and SMETA are not interchangeable. They are different audit systems with different governance, different platforms, and different recognition rules.

amfori BSCI is managed by amfori, a Brussels-based business association. It is an audit framework built around a code of conduct covering labor standards, working hours, health and safety, fair remuneration, and ethical business behavior. The audit is conducted against a set of questions, with results graded from A to E. BSCI audits are conducted by amfori-trained auditors and results are hosted on the amfori platform.

SMETA is the Sedex Members Ethical Trade Audit. It is a methodology hosted on the Sedex platform. A SMETA audit can be conducted as a 2-pillar assessment (labour standards, health and safety) or a 4-pillar assessment, which adds environment and business ethics. SMETA is not a certification — Sedex explicitly states that no “pass” or “fail” is issued. It produces an audit report with findings.

The critical point: amfori has stated explicitly that SMETA audits are not recognized by amfori BSCI. Although there are memoranda of understanding between the two organizations in areas such as capacity building, amfori’s recognition criteria for other systems require coverage of all ILO Core Conventions, amfori’s specific OHS elements, and verification of minimum wage payment. SMETA does not offer amfori BSCI’s elements of governance, integrity, and auditor competence assurance. Therefore, amfori BSCI cannot recommend its participants accept SMETA audits as part of the BSCI due diligence approach.

For buyers, this means: if your customer or compliance program specifies BSCI, a SMETA report does not satisfy that requirement. Conversely, if your program specifies SMETA, a BSCI report does not substitute. The two audits cover overlapping but non-identical areas and use different platforms and assessment approaches.

What ISO 9001 adds — and does not add. ISO 9001 is a quality management system standard. It confirms that the factory has documented procedures for quality control, corrective action, internal audit, and records management. It does not test product safety, does not evaluate labor conditions, and does not address environmental performance. A factory can hold ISO 9001 and still produce glass with inconsistent annealing, or ship a jar with a non-compliant lid seal.

What you should actually request. If you need social compliance evidence, request the specific audit your customer or regulation requires — BSCI or SMETA, with the correct pillar scope. If you need quality system evidence, request ISO 9001 with a scope that names your product category. If you need product safety evidence, request the product test report, not a factory certificate. These are three different questions with three different evidence types.

How Should Importers Without a Customer-Specified Audit Design Their Own?

Many importers — especially mid-sized buyers without a large retailer’s compliance infrastructure — do not have a customer-mandated audit scheme. They need to design their own risk-based approach.

The logic is tiering. Not every supplier warrants the same audit intensity. A practical framework is:

Tier 1 — New or unverified supplier. Full on-site audit before first order. Document verification, factory walkthrough, batch traceability check, QC process review, packaging inspection. If on-site is not feasible, a live video walkthrough with screen-sharing of quality records is better than nothing, but it is not equivalent to physical verification.

Tier 2 — Conditional or recently onboarded supplier. Desktop document review plus a focused on-site audit on the specific risk area identified in the first audit. If the first audit found weak annealing records, the follow-up focuses on annealing. If it found packaging inconsistencies, the follow-up focuses on packaging.

Tier 3 — Established supplier with consistent performance. Reduced audit frequency, with targeted checks on any change — new product, new lid material, new decoration process, or a change in subcontractor. Performance metrics should drive the audit calendar: on-time delivery rate, defect rate at pre-shipment inspection, and claim history.

The tiering criteria should be documented before the audit, not adjusted after seeing the factory. A common failure is relaxing the audit standard because the factory is pleasant, or because the samples are good. The audit is a risk-control tool, not a relationship-building exercise.

How Do You Verify That Corrective Actions Have Actually Been Implemented?

A corrective action plan is not evidence of corrective action. The plan is a document. Verification requires evidence that the change has occurred.

Check the physical evidence, not the paper. If the corrective action was “install additional safety barriers around the annealing furnace,” go to the furnace and look. If the action was “implement batch traceability system,” pick a recent batch and trace it. If the action was “upgrade packaging to double-wall cartons,” inspect a packed carton from current production and verify the carton specification.

Request data, not narratives. For actions related to quality defects, ask for the defect rate data before and after the change. If the factory claims the fix worked, the data should show a measurable improvement. If the action was related to on-time delivery, ask for the OTD rate for the months since the change.

Verify at the operator level. A corrective action that changes a procedure is only effective if the operators follow it. Ask a line operator to demonstrate the new procedure. If they cannot, the training has not been delivered or the procedure has not been implemented.

Check for recurrence. Ask whether the same nonconformity has appeared in subsequent internal audits. A factory that documents a corrective action but records the same finding in the next internal audit has not closed the loop. The most useful question is: “Show me the evidence that this problem has not recurred since the corrective action was implemented.”

A structured approach — sometimes described as verifying the corrective action in four dimensions: the plan, the on-site implementation, the data supporting effectiveness, and the long-term control to prevent recurrence — is more reliable than accepting a signed corrective action report at face value. The document proves intent. The floor proves execution.

Pre-Audit and On-Site Checklist

Before you book the visit

• Business license scope verified against stated manufacturing activities

• Factory address confirmed and matched to license

• BSCI or SMETA report reviewed: audit date, audited site, validity period noted

• ISO 9001 certificate scope checked: does it name your product category and site?

• Subcontracting disclosed: which processes, which sites?

• Production capacity evidence requested

On the factory floor

• Raw material intake records and cullet sorting process inspected

• Batch weighing records reviewed for accuracy and calibration

• Annealing lehr temperature logs requested (minimum 30 days) and cross-checked against production schedule

• Polariscope and stress-testing records verified

• QC stations walked and defect classification reviewed against AQL criteria

• Lid and closure fit checks observed

• Packaging area inspected: carton spec, dividers, cushioning, palletising

• Finished goods warehouse: stacking, non-conforming segregation

• Batch traceability test: pick a finished carton, trace to raw material

Document cross-checks

• Recent QC record matched to physical batch

• Annealing chart matched to production schedule

• Operator interview compared against written procedure

• Calibration certificates verified against equipment on floor

Social audit verification

• Confirm which audit the customer requires: BSCI or SMETA

• Verify the correct pillar scope for SMETA (2-pillar or 4-pillar)

• Confirm that the audited site matches the production site

• Note that BSCI and SMETA are not interchangeable

At KINGSTAR GLASSWARE, we welcome factory audits as a verification exercise, not a formality. Our Qingdao facility maintains documented annealing profiles, batch-level traceability from raw material intake to finished carton, and product-specific test reports that match the configuration on each order. We hold BSCI audit documentation and can provide the scope information buyers need to confirm that the audited site covers their production.

Ready to schedule a factory audit or request pre-audit documentation? Reach out to KINGSTAR GLASSWARE. Tell us your product, your target market, and the audit scope you require. We will provide the evidence you need to make a confident supplier decision.

FAQ

What is the difference between a glassware factory audit and a product test report?

A factory audit examines how a facility operates — its production processes, quality controls, labor conditions, and traceability systems. A product test report examines a specific item’s performance under defined test conditions, such as lead and cadmium migration for food-contact glass. A factory can pass an audit and still produce a product that fails a test, if the tested configuration differs from the production configuration. Buyers need both types of evidence, and they answer different questions.

Can I accept a SMETA audit report if my customer requires BSCI?

No. amfori has stated explicitly that SMETA audits are not recognized by amfori BSCI. The two systems cover overlapping but non-identical areas, use different platforms, and have different governance and integrity mechanisms. If your customer or compliance program requires BSCI, a SMETA report does not satisfy that requirement. Request the specific audit type that is required.

How do I verify that a factory’s annealing process is actually controlled?

Request the annealing lehr temperature logs for the past 30 days. Cross-check the logged temperatures and durations against the production schedule for the same period. Then ask to see polariscope readings or stress-test records for recent batches. A factory that cannot produce either the logs or the stress data is not controlling the annealing process in a verifiable way.

Related Articles

Preparing to audit a glassware manufacturer in China? Here are three practical pieces from our blog that help buyers verify factories more thoroughly:

• How to Vet a Glassware Supplier for Your OEM Project
  The 10-point checklist covering capacity, quality systems, documentation, and communication — essential preparation before any factory visit.

• Golden Sample: Your Most Powerful Tool in Quality Control
  How an approved golden sample becomes the non-negotiable benchmark during audits and ongoing production verification.

• How to Control Batch-to-Batch Quality
  Practical methods for checking process control and consistency that experienced buyers use during factory audits.