Two factories can make the same SKU and still be impossible to compare.
One report says “minor scratch.” Another says “appearance issue.” One factory measures in millimeters, another only writes “within tolerance.” One sends close-up photos with no carton or lot reference. The other sends a pass result without the inspection checklist. If those records sit in separate folders, you do not have a multi-factory quality system. You have several isolated quality stories.
The goal is not to make every factory run the same process. The goal is to make every factory work from the same released product requirements, report against comparable controls, disclose meaningful site changes, and provide enough evidence for the buyer to see where risk differs.
The American Society for Quality defines supplier quality management as a proactive, collaborative system and notes that it runs from product design and supplier selection through the product lifecycle and supplier relationship. 1 UTS Inspection makes the same practical distinction for multi-site work: use one controlled baseline, but adjust control depth for actual supplier and product risk rather than assuming all factories have identical capability. 2
Build one controlled quality baseline
Every factory producing the same product family should receive the same current baseline pack. This is the version-controlled source set that says what “right” looks like.
| Baseline-pack element | What it controls |
|---|---|
| Product identity | SKU, model, style, color, size, revision, and approved variants |
| Technical requirements | Drawings, dimensions, materials, BOM, critical characteristics, and tolerances |
| Approved reference | Approved sample, visual standard, photo reference, or first-piece reference with a clear revision ID |
| Workmanship rules | Product-specific defect descriptions, severity rules, and examples |
| Function and test rules | Test method, setup, equipment, conditions, pass/fail result, and evidence required |
| Packaging and artwork | Unit pack, carton structure, labels, shipping marks, warnings, barcode/QR data, artwork version, and packing list requirements |
| Inspection plan | Checkpoints, method, lot/sampling instructions, evidence triggers, and report fields |
| Change record | Date, owner, old/new revision, effective order or sample round, and acknowledgement status |
The pack should have a single owner. If drawings, samples, artwork, and checklist rules are sent through separate informal chats, each factory may unknowingly build to a different version.
1factory identifies change control, production-readiness verification, work instructions, and clear quality checkpoints as foundations for stable supplier quality. 3 UTS warns that a factory using an old drawing or label file may produce consistent goods that are consistently wrong, which is why reports should identify the revisions and references used. 2
Give every physical site its own profile
A legal supplier name does not always identify the factory that makes, packs, tests, or subcontracts the product. Build a site profile for every physical production location.
| Site-profile field | What to capture |
|---|---|
| Supplier and site identity | Legal supplier, factory name, physical address, contact, and operating-site identifier |
| Approved scope | Products, processes, and SKUs the site is approved to produce |
| Critical process map | Key materials, components, machines, tools, finishing steps, testing, packing, and storage points |
| Subcontracting | Which processes, parts, testing, finishing, or packing occur elsewhere |
| Measurement and test capability | Available equipment, relevant calibration/verification controls, operators, and known limits |
| Quality controls | Incoming checks, in-process checkpoints, final checks, nonconforming-goods segregation, records, and training approach |
| Traceability | Material/lot, work-in-process, finished-goods, carton, and revision tracking used on the site |
| Known risks and actions | Open findings, prior recurring defects, current changes, and follow-up evidence required |
UTS recommends recording the physical production site, key subcontractors, critical processes, testing/measurement capability, traceability, known limitations, and open corrective actions before deciding which controls apply. 2 1factory likewise describes inventory control, change control, sub-tier management, production readiness, and process control as key supplier-quality areas. 3
A site profile is not a pass certificate. It is a current record of what the buyer knows, what the site is approved to do, and what still needs evidence.
Make inspection reports comparable by design
If report fields vary by factory, the score can look better simply because a site recorded less. Require a core report structure that every factory and inspection team uses for the same product family.
| Core field | Why it must be common across sites |
|---|---|
| Site, product, order, lot, and production status | Connects the report to the actual factory and goods |
| Product and document revisions | Confirms that all sites were checked against the same current baseline |
| Checkpoint, requirement, and method | Shows what was checked and how—not merely a result label |
| Sampling or scope statement | Stops one checked unit or partial lot from being presented as whole-order evidence |
| Actual measurements and test conditions | Makes results reproducible and comparable |
| Defect vocabulary and severity | Prevents equivalent issues being renamed or grouped differently |
| Evidence IDs | Connects photos/video, instrument readings, document review, and samples to specific findings |
| Finding counts and affected scope | Separates a sample observation from a known or suspected broader issue |
| Limitations and open items | Makes missing access, unavailable records, inactive processes, or unperformed checks visible |
| Outcome and next action | Records the result under the buyer-approved criteria and who owns the next evidence |
UTS says reports from different factories are useful only when they contain comparable evidence. Its recommended common fields include site identification, PO/SKU/lot/production status, revisions, sampling information, photos, measurements/test conditions, packaging/label findings, limitations, and a conclusion based on buyer-approved criteria. 2
Use the same measurement units, defect terminology, photo requirements, and result categories where factories produce the same SKU. Site-specific notes still matter. The shared core is what lets you compare them honestly.
For media requirements, use How to Use Photos and Video in a Quality Control Report. A report is stronger when each image is tied to a product, requirement, finding, location, and defined scope.
Standardize the control, not the factory
A common baseline does not mean every site gets the same inspection depth or has the same risk. The control plan should respond to evidence.
| Risk signal | A possible buyer-defined response |
|---|---|
| New factory, new SKU, new material, tooling, or site | Verify capability, sample/reference alignment, early production readiness, and current documents more closely |
| Repeated workmanship, packaging, or labeling issue | Add focused checkpoints, require correction evidence, and check whether the prior remedy worked |
| Material or component substitution risk | Review BOM/traceability evidence and define what needs a physical, document, or specialist check |
| Stable repeat production with no significant changes | Continue the protected core controls; adjust other controls only when current evidence supports it |
| Serious complaint, failed batch, missing traceability, or unapproved change | Escalate the review under the buyer’s documented risk/approval process |
| Complex product with critical sub-tier process | Map the external process and require evidence appropriate to its influence on the finished product |
ASQ says supplier management should consider performance, quality-system capability, current and potential capacity, technical support, delivery reliability, and the total cost of communication and problem resolution—not only purchase price. 1 1factory advises focusing verification resources on higher-risk suppliers and characteristics, especially with new suppliers, changes, and performance monitoring. 3
The word possible matters here. The buyer must define their own risk thresholds, inspection scope, and decision route. There is no universal frequency or control level that fits every product and site.
Publish a change-notice rule before you need it
Most cross-factory quality drift starts with a change nobody compared: a new raw material source, packing supplier, mold repair, software version, production line, factory location, label file, testing device, or subcontracted process.
Require each site to notify the buyer before a change that may affect an approved product or the evidence used to verify it. The rule should state what the supplier must disclose, who reviews it, and what evidence is needed before the change becomes part of the current baseline.
| Change-notice field | What to require |
|---|---|
| Change ID and site | A unique reference plus physical location |
| Product/lot/revision affected | Which SKU, order, material, process, drawing, or packaging record changes |
| Old and proposed condition | Clear description of what was used before and what is proposed now |
| Reason for change | Capacity, availability, improvement, equipment, site, supplier, or process reason |
| Potential impact | Product function, appearance, dimensions, labeling, packaging, traceability, or testing implications |
| Evidence plan | Sample, first-piece result, document update, measurement, test, inspection, or other agreed proof |
| Review and effective date | Buyer decision route and the order/run from which the new condition is valid |
| Record update | What baseline documents, samples, work instructions, and report templates must be revised |
1factory states that inadequate change control is a common cause of defects and gives the practical example of needing to update old drawings, fixtures, programs, and quality-control plans when a new revision is released. 3
See How to Manage Product Revisions Without Losing Control of the Quote for a related buyer-side process for version, scope, and commercial-record control.
Compare evidence, not just pass/fail labels
A passed final inspection is one data point. It is not proof that every uninspected unit is conforming or that the site will perform consistently on the next order.
UTS explicitly notes that a passed or failed lot-inspection result does not prove the condition of every uninspected item. 2 Multi-factory comparison needs a history of evidence, not a column of green results.
Use a shared review table that keeps raw evidence visible.
| Review area | Questions to ask across sites |
|---|---|
| Product conformity | Do the same measurements, visual rules, and test methods produce comparable results? |
| Recurring issues | Which defect types, locations, materials, or packaging errors repeat by site or product? |
| First-pass evidence | Did the product meet the released baseline before rework or repeat checking? |
| Correction quality | Was the current problem contained, corrected, verified, and tracked separately from recurrence prevention? |
| Change control | Were material, process, tooling, site, artwork, or subcontracting changes declared and reviewed? |
| Evidence quality | Are photos, videos, measurements, document copies, and scope limitations equally traceable across sites? |
| Operational controls | Are key records, measurement controls, traceability, and quality checkpoints available and current? |
| Open risk | Which site/product combinations still have unresolved findings, missing evidence, or pending verification? |
1factory recommends supplier performance measurement across multiple dimensions and says scorecards can include quality, delivery, cost, responsiveness, and improvement activity. 3 This article uses the comparison table as an operating review, not a universal supplier-ranking formula. Article 81 will cover a dedicated supplier scorecard.
Keep correction evidence separate from long-term improvement
If one factory finds a defect, a quick correction may be appropriate for the current lot. That does not show that the cause has been addressed across future production.
| Immediate batch work | Longer-term recurrence work |
|---|---|
| Identify affected goods and separate their status | Define the problem and likely cause using process evidence |
| Issue a controlled correction instruction | Identify a change that addresses the cause rather than the symptom |
| Collect rework/sorting/replacement evidence | Assign an owner and implementation date |
| Verify the correction against the current baseline | Check effectiveness in a later run, audit, or relevant checkpoint |
| Record unverified or excluded goods clearly | Update the baseline/control plan if the finding changes future requirements |
1factory lists immediate containment, root-cause analysis, permanent solutions, and effectiveness verification as parts of a structured supplier corrective-action process. 3
For a factory-specific failure response, see What to Do When a Factory Fails an Inspection. Article 39 will go deeper into building a corrective-action plan.
Run a regular cross-factory quality review
The meeting is less important than the shared evidence. Use the same review agenda whenever new reports, site changes, or recurring problems make comparison useful.
- Confirm baseline status: Which product, artwork, drawing, sample, checklist, and defect-standard revisions are current?
- Check site/lot coverage: Which factories, products, lots, and process steps have evidence? What is missing or outside scope?
- Review exceptions: What repeated defects, open items, change notices, or traceability gaps require attention?
- Compare control evidence: Are measurement, test, packaging, labeling, and photo/report standards being applied consistently?
- Assign follow-up: What evidence is due next, from whom, and under which issue/change ID?
- Review effectiveness: Did previous site-specific changes and actions produce the expected evidence on the next relevant run?
ASQ says collaboration, information sharing, and an understanding of both buyer and supplier roles are important elements of supplier quality management. 1 The review should make gaps visible, not use a score to hide them.
A minimum operating set for multi-factory quality
Before you rely on multiple factories for the same product family, confirm that you have:
- One owner-controlled baseline pack and change log.
- A physical-site profile for each factory and known critical subcontractor.
- One core inspection/report template with common terms, units, evidence IDs, and limitation fields.
- A product-specific defect standard and a current reference sample or visual benchmark.
- A change-notice path that records review and effective scope.
- A risk-based control plan that protects important checks even when other scope changes.
- A shared evidence review that compares facts, not only pass/fail labels.
- Separate registers for current-batch correction, recurrence work, and unresolved evidence.
- A buyer-defined approval path for decisions outside the quality report itself.
Multiple factories can increase capacity and reduce dependency. They also create more places for a product definition to drift. One controlled baseline, comparable evidence, and honest site-specific risk records make that drift easier to see before it becomes a customer problem.