How to Plan a Factory Check Around One Product-Specific Risk

A useful factory check answers a defined sourcing question. It does not attempt to assess every process, management practice, defect type, or commercial capability at the site. During pre-sourcing research, the task should focus on one product, one risk, and one sourcing decision.

The intended output is a written task card and practical checklist that a competent checker can execute consistently. Its objective, criteria, evidence, sampling approach, and decision rule must all remain tied to the selected product-specific risk.

The core principle is simple: do not turn a focused factory check into a general factory audit. A narrow check may support supplier selection or identify a need for further review, but it cannot establish that every product, process, or management system at the factory is acceptable.

Define the single-risk verification task

Write a narrow risk statement

Use this structure:

For [product/model], verify whether [material, component, process step, or finished-product characteristic] conforms to [documented requirement], because nonconformance could lead to [specific product-safety or performance consequence].

A workable statement should:

• Name one product or clearly defined product family.
• Identify one failure mode rather than a broad concern such as “poor quality.”
• Connect that failure mode to a particular feature, material, component, or manufacturing step.
• Reference an approved drawing, specification, test method, legal requirement, or other controlled criterion.
• State the specific safety or performance consequence.
• Avoid asking the checker to decide what “safe,” “strong,” “durable,” or “good enough” means without a documented criterion.

For example, the task should not ask a checker to “confirm that the enclosure is strong.” It should identify the applicable model, enclosure characteristic, controlled requirement, verification method, and consequence of nonconformance.

Convert the risk into a task card

Complete the task card before contacting factories. Undefined criteria should not be delegated to the checker during the visit.

Task-card field	Required planning entry
Product	Product name, model, revision, and applicable variants
Single risk	One clearly described failure mode
Verification objective	The exact question the check must answer
Check criteria	Controlled specification, drawing, procedure, or test requirement
Process location	Where the relevant material, component, operation, or inspection occurs
Relevant production stage	Incoming material, work in progress, assembly, final inspection, or another defined stage
Evidence required	Records, observations, interviews, samples, measurements, or test results
Sampling basis	Whether attribute sampling is suitable and, if so, the authorized ISO 2859-1 inputs
Decision rule	Conditions for pass, fail, inconclusive, or escalation
Responsible roles	Who performs the check, supplies documents, reviews results, and makes the sourcing decision
Deliverable	Completed checklist, evidence references, nonconformity record, limitations, and conclusion

Keep the task distinct from broader supplier assessment

Include only matters that help answer the verification objective:

• The production operation connected to the selected risk.
• The relevant control procedure and records.
• The equipment, fixture, measuring device, or test method used for that control.
• Samples from a clearly defined product or production lot.
• The factory’s handling of detected nonconforming units, where relevant.

Exclude the following unless they directly affect the selected risk:

• General commercial capability.
• Unrelated production lines.
• Broad social, environmental, or management-system topics.
• Unrelated product defects.
• Certification claims that cannot be verified from controlled evidence.

If an observation indicates a serious concern outside the planned scope, record and escalate it. Do not silently expand the task and then present the result as a complete factory audit.

Design the factory check

Set the objective, scope, and criteria first

The planning sequence below adapts concepts from ISO 19011 auditing guidelines without representing the focused check as an ISO certification audit:

1. Objective: State what the check must determine.
2. Scope: Define the product, process, location, production stage, and time window covered.
3. Criteria: List the requirements against which evidence will be evaluated.
4. Methods: Select document review, observation, interview, measurement, testing, or sampling only when the method supports the objective.
5. Competence: Confirm that the assigned checker understands the product requirement, manufacturing operation, and verification method.
6. Evidence trail: Require every finding to point to identifiable records, observations, measurements, or sample results.

Methods follow the objective; they do not define it. A factory tour, employee interview, or stack of records has limited value if the checker cannot connect that evidence to the selected characteristic and acceptance criterion.

Build an evidence chain around the risk

Organize the check in production order so that the report shows how the requirement is translated into factory control and product results.

1. Requirement

• Confirm that the factory has the correct product specification or drawing revision.
• Identify the exact characteristic and acceptance criterion.
• Record the document title, identifier, and revision.
• Note any difference between the buyer-approved requirement and the factory’s working document.

2. Process control

• Observe the operation that creates or controls the characteristic.
• Compare actual practice with the applicable work instruction where one exists.
• Record the location, time, product status, and objective observations.
• Avoid conclusions based only on an operator’s verbal explanation.

3. Control resources

• Identify the relevant equipment, fixture, template, measuring device, or test setup.
• Record enough identifying information to connect the resource to the observation.
• Check only the status or controls required by the task criteria.

4. Production records

• Review records relevant to the selected product, operation, lot, and time period.
• Determine whether the records contain the information required by the applicable procedure.
• Capture record identifiers, dates, results, product or lot references, and responsible personnel where required.

5. Product verification

• Inspect, measure, or test selected units against the predetermined criterion.
• Use the specified method and conditions.
• Record individual results rather than relying on a verbal or aggregate summary.

6. Nonconformity response

• Determine how failed units associated with the selected risk are identified and controlled.
• Review corrective or preventive controls only where they are relevant to the task.
• Record conflicting evidence, such as acceptable records paired with nonconforming sampled units.

This chain is also consistent with the CPSC’s emphasis on integrating product-safety practices across production, testing, recordkeeping, and corrective action in its manufacturing best-practices guidance.

Decide whether attribute sampling applies

Use ISO 2859-1 sampling procedures only when each examined unit can be classified by attribute, such as conforming or nonconforming.

Before using an ISO 2859-1 plan, a competent person must define:

• The lot and the basis on which units belong to it.
• The product characteristic being examined.
• The method for classifying each unit.
• The inspection level and other required sampling-plan inputs.
• The acceptance quality limit.
• The acceptance and rejection rules.
• The applicable inspection status and switching rules.
• The response if the lot cannot be identified or the required sample cannot be obtained.

Do not:

• Invent a sample size.
• Select convenient units instead of following the planned selection method.
• Change the acceptance rule after seeing results.
• Treat a passed sample as proof that every unit conforms.
• Use acceptance sampling as a substitute for defining the product-safety requirement.
• Derive an operational sampling plan from the ISO webpage summary alone.

The applicable official standard and its tables are needed to establish the operational plan. If the factory cannot identify the lot or make the required units available, apply the predetermined inconclusive or escalation rule rather than improvising a different sample.

Predetermine the decision logic

Define the possible outcomes before the check:

• Pass: Required evidence is available, observations support the defined control, and sample results satisfy the predetermined rule.
• Fail: Evidence demonstrates nonconformance, or the sampling decision rule rejects the lot.
• Inconclusive: Required evidence, access, criteria, lot definition, or sample availability is insufficient.
• Escalate: The finding suggests a safety issue, a control breakdown beyond the planned scope, or the need for specialist review.

A lack of evidence is not evidence of conformity. Record it as a limitation and apply the planned inconclusive or escalation rule.

Practical verification checklist

Planning and execution table

Check	Verification action	Evidence to capture	Complete
Task identity	Record the product, model, revision, factory location, planned date, and task owner.	Completed task header	☐
Single-risk statement	Describe one failure mode and its possible consequence.	Approved risk statement	☐
Verification objective	Write one answerable verification question.	Objective in task card	☐
Requirement source	Identify the controlled specification, drawing, procedure, or test criterion.	Document title, identifier, and revision	☐
Scope boundary	Define the included product, process, location, production stage, and time period.	Written scope	☐
Exclusions	State what the check will not assess.	Written exclusions	☐
Checker competence	Confirm that the checker can understand and apply the criterion and method.	Assignment or competence confirmation	☐
Factory contact	Identify the person responsible for access to the relevant process and records.	Contact and role	☐
Production availability	Confirm that the relevant product and operation are expected to be available.	Written confirmation	☐
Process map	Identify where the risk is introduced, controlled, checked, and recorded.	Short process sequence	☐
Current requirements	Compare the factory’s working documents with the approved revision.	Document references and observed revisions	☐
Work instruction	Determine whether the relevant operation has a defined method.	Procedure reference or noted absence	☐
Process observation	Observe the relevant operation without broadening the task into an unrelated factory tour.	Time, location, operation, and objective notes	☐
Control resources	Identify the relevant equipment, fixtures, templates, or measuring devices.	Identifiers and observed status	☐
Records	Review records connected to the selected product and risk.	Record identifiers, dates, and result summary	☐
Lot definition	Define the population from which samples will be selected.	Lot identifier and quantity source	☐
Attribute suitability	Confirm that every examined unit can be classified against the same criterion.	Written classification rule	☐
Sampling plan	Record the authorized ISO 2859-1 inputs and resulting plan.	Plan reference, sample size, and decision rule	☐
Sample selection	Follow the planned selection method and record where units came from.	Unit or sample references	☐
Verification method	Apply the specified inspection, measurement, or test method consistently.	Method reference and test conditions	☐
Individual results	Record each sampled unit as conforming or nonconforming, with relevant observations.	Result sheet	☐
Sampling decision	Apply the predetermined acceptance or rejection rule without adjustment.	Calculation or decision record	☐
Failed-unit control	Observe how relevant nonconforming units are identified and controlled.	Objective notes and record references	☐
Conflicting evidence	Record differences among documents, interviews, observations, and results.	Conflict log	☐
Access limitations	Note unavailable processes, records, people, lots, or samples.	Limitation statement	☐
Conclusion	Select pass, fail, inconclusive, or escalate using the predetermined logic.	Signed conclusion and rationale	☐
Follow-up owner	Assign any clarification, corrective action, retest, or specialist review.	Owner and required next action	☐

Required report structure

Require a short, reproducible report containing:

1. Task identity and date.
2. Single-risk statement.
3. Objective, scope, criteria, and exclusions.
4. Factory process and records examined.
5. Sampling-plan reference and sample traceability.
6. Individual verification results.
7. Findings and supporting evidence references.
8. Limitations and conflicting evidence.
9. Pass, fail, inconclusive, or escalation conclusion.
10. Required follow-up and decision owner.

The report should distinguish facts from conclusions. Photographs, record references, measurements, and sample results should be traceable to the relevant checklist item rather than included as an unexplained evidence collection.

Sources

• ISO 2859-1 sampling procedures — Source for planning lot-by-lot acceptance sampling by attributes. Use the applicable official standard to establish sample size and acceptance or rejection rules.
• ISO 19011 auditing guidelines — Source for structuring objectives, scope, criteria, methods, evidence, competence, and reporting.
• CPSC manufacturing best practices — Source for placing product-safety controls across manufacturing, testing, records, and corrective action.

Image credit

Title: A worker in a yellow high-visibility shirt operating industrial woodworking machinery
Photographer: PJ Wallace
Source: Unsplash
License: Unsplash License

Editorial use: The image illustrates observation of a specific production operation. It is not evidence about any factory, supplier, product, or control.

Scope and limits

What this plan can establish

A properly executed focused check can establish:

• Whether defined evidence was available within the agreed scope.
• Whether the observed process met the predetermined criteria at the time of the check.
• Whether sampled units satisfied the applicable verification and sampling rules.
• Whether the result supports a pass, fail, inconclusive, or escalation decision for the distinct task.

What this plan cannot establish by itself

This check cannot establish:

• That every unit in a lot conforms.
• That the factory will maintain the same controls after the check.
• That unrelated products, processes, lines, or locations are acceptable.
• That the product complies with requirements omitted from the criteria.
• That the factory or its management system is certified.
• That a focused check replaces broader technical, product-safety, regulatory, commercial, or supplier due diligence.

Final next move

Complete the task card with the actual product, one risk, a controlled acceptance criterion, the relevant factory process, and any authorized sampling inputs. If any of those fields remains undefined, pause the factory-check request and assign the missing decision to the responsible product, quality, safety, or regulatory role.