A prototype sample proves that a design can work. A production sample proves that a specific factory can repeat it at scale, with the actual tooling, materials and process. Approving both at the same gate — or against the same criteria — is how brands end up with a market-ready concept that fails first-article inspection. Build two separate gates: one that validates design feasibility, and one that validates manufacturing repeatability. This part covers why the split matters and how to define and run the first gate.
Visual guide · Yoursender Insights
Prototype approval vs. production approval
Use two distinct gates: one for whether the design works, another for whether the factory can reproduce it consistently.
| Decision area | What to verify |
|---|---|
| Primary question | Prototype: does the concept work? | Production: can the process repeat it? |
| Main focus | Prototype: function and design intent | Production: materials, process and workmanship |
| Evidence | Prototype: test notes and design changes | Production: approved spec and process checks |
| Decision | Prototype: revise or proceed to validation | Production: approve or hold the production release |
Use this framework as a starting point; confirm requirements and acceptance criteria for your own product and market.
Why Prototype and Production Samples Need Separate Approval Gates
The two samples answer different questions, come from different processes, and carry different risks:
- Prototype: Does the design meet the user requirement? Made with 3D printing, CNC, off-the-shelf parts or hand assembly — useful for learning, not proof of production capability.
- Production sample: Does the supplier’s line reproduce the approved design consistently? Made from production tooling, real materials and the intended process.
Merging them creates predictable failures: tooling gets cut before the design is frozen, cosmetic details get argued over while function is still unproven, or a hand-finished sample is accepted as evidence of mass production. The tradeoff is time — two gates add review cycles — but re-cutting a mold costs far more than one extra sample round.
Define What Each Sample Must Prove
Write the evidence requirement into your checklist before any sample is built, so the supplier knows what will be measured.
| Question | Prototype gate | Production gate |
|---|---|---|
| Design intent | Validate | Confirm unchanged |
| Material and finish | Approximate | Exact, per specification |
| Tooling and process | Not required | Production tooling required |
| Tolerances | Indicative | Measured against drawing |
| Packaging and labeling | Not assessed | Shippable form |
Prototype Sample Gate: Fit, Function, and Feasibility
- Set pass/fail criteria first. Define the test, the measurement method and the acceptance threshold before the sample ships.
- Fit: check interfaces, mating parts and tolerance stack-up. Record actual measurements rather than a visual impression.
- Function: run the intended-use test and document the result against the threshold.
- Feasibility: ask the supplier how this sample was made and how production will differ. Any delta in process or material is a risk to log.
Decision: approve to tooling, iterate, or reject. Risk to watch: a sample that looks right can still hide process risk, and a hand-built unit may not repeat.
Next step: freeze the design baseline the production sample will be measured against, and record the gate criteria in your project workspace via Yoursender.
Production Sample Gate: Tooling, Materials, and Process Capability
The production sample gate is not a second prototype review. It confirms that final tooling, specified materials, and the intended production process can reproduce the approved design consistently. Before the sample is made, confirm the tooling is production-intent, not a bridge or hand-built tool. Ask for the cavity layout, key process parameters, and the line or cell that will run the order. Verify materials against the approved specification with lot or batch traceability, and close any proposed substitutions in writing before the run.
Decision criteria: tooling is final and stable, materials match the approved specification, the production process and operator are representative, and the supplier has a control plan for critical characteristics. A common risk is accepting a sample built with prototype tooling or hand assembly; it can hide molding, fit, and cycle-time problems that appear at volume. The tradeoff is schedule: waiting for final tooling adds time, but it prevents approving a process that cannot scale.
Build the Sampling Gate Checklist: Criteria, Tests, and Tolerances
Build the checklist before the sample arrives so the review is evidence-based, not opinion-based. Set tolerances from the drawing or approved specification, not from the supplier’s default process capability. If a tolerance is missing, agree on it before the run. Define sample size, test methods, and acceptance criteria in advance, and keep a signed golden sample for visual and fit reference.
| Area | Criterion | Evidence | Disposition |
|---|---|---|---|
| Dimensional | Critical dimensions within drawing tolerance | Measurement report with method and tool | Pass / fail / deviation |
| Materials | Resin, metal, or fabric matches approved spec | Material certificate and lot traceability | Pass / fail / deviation |
| Functional | Assembly, fit, cycle, leak, or load test as applicable | Test record and photos | Pass / fail / deviation |
| Cosmetic | Color, texture, finish, and defects match golden sample | Signed limit sample | Pass / fail / deviation |
| Packaging and labeling | Artwork, barcode, warnings, and pack configuration correct | Pack-out photos and proof | Pass / fail / deviation |
Document Deviations and Conditional Approvals
Every miss should be logged, not negotiated verbally. Record the deviation ID, specification, observed result, impact, root cause, corrective action, owner, and due date. Classify each as critical, major, or minor. Do not grant conditional approval for safety, regulatory, or mandatory labeling deviations. For minor or cosmetic issues, a conditional approval must be written, scoped to a defined quantity or time, and tied to a re-sample or inspection trigger. State the disposition clearly: approved, approved with conditions, rework and re-sample, or rejected.
Next step: review the production sample against the checklist, record the disposition, and update the specification and control plan before bulk release. If you need support turning this gate into a supplier-ready sampling plan, start a project with Yoursender.
Set Sign-Off Roles and Evidence Requirements
Sampling gates stall when approval is verbal, spread across chat threads, or owned by nobody in particular. Before the first sample ships, name one accountable approver per gate and define what they are allowed to approve.
| Role | Signs off on | Evidence required |
|---|---|---|
| Brand product owner | Final approval, fit, finish, packaging | Signed sample with photos by angle, revision number, date |
| Technical / quality reviewer | Dimensions, materials, function, safety testing | Measurement report against spec with tolerances, material confirmations, test reports |
| Sourcing / merchandising | Cost, MOQ, lead-time impact of any change | Written deviation list with cost and schedule effect |
| Factory QA | Their own build record and inspection criteria | Internal checklist, batch traceability, retained sample |
Give each role a clear decision right: approve, approve with conditions, or reject. Reviewers outside their domain can raise objections but should not block sign-off.
Require a written evidence package with every sample submission:
- Dated photos of top, bottom, label, interior, and packaging
- Dimensional report showing measured values against tolerance bands
- Material and component sources, including color or finish references
- Any applicable test reports, plus a list of known deviations
- Confirmation that the sample matches the current specification revision
A conditional approval must state the deviation, the quantity or time it covers, and the trigger for re-approval. The common failure is a factory starting bulk on a thumbs-up emoji or a single photo. Treat any approval not recorded in your project file as not given.
Use Golden Samples and Limit Samples for Reference
A golden sample is the signed, sealed unit that defines the approved standard. A limit sample defines the edge of acceptable variation — the darkest acceptable shade, the largest tolerable stitch irregularity, the dullest allowed gloss. Golden samples answer “what is right”; limit samples answer “how far off is still fine.”
- Build golden samples from a production run, not hand-finished lab units, or the reference cannot be reproduced.
- Produce at least three: buyer copy, factory copy, and a sealed arbitration copy.
- Label each with SKU, revision, date, approver, and material batch.
- For limit samples, create an acceptable and a reject pair for each visual attribute in dispute.
- Store them sealed, away from light and dust, and replace them when the material batch changes.
The tradeoff is time: limit samples add a setup session, but they remove most subjective arguments at inspection. Record each approved revision in your project workspace so the inspection team and factory are working from the same reference.
Next step: before booking any pre-production or during-production inspection, confirm in writing that the golden sample revision on file matches the production lot being built.
Avoid Common Sampling Pitfalls in China Manufacturing
Most sampling failures are process failures, not factory failures. The patterns repeat:
- Treating a prototype as a production approval. A hand-built unit made with machined parts, a different operator, or a temporary fixture proves the design concept — not the tooling, the line, or the material lot.
- Approving “looks good” by photo. Written approval should reference measured dimensions, a cosmetic limit sample, and a signed checklist, not chat messages.
- No sealed golden sample. Keep a signed, dated, sealed reference at the factory and a matching one in your office. Unsigned samples become disputed samples.
- Verbal deviations. Every accepted nonconformity needs a written waiver stating the limit, the quantity affected, and an expiry date.
- Late compliance and packaging testing. Test on production-representative samples, not prototypes, and include retail packaging, labeling, and transit testing in the same gate.
- Undisclosed sub-supplier changes. Confirm material and component sources in writing and make any change trigger a new sample round.
When to Release Tooling and Mass Production
Release tooling only when a production sample, built with production tooling, production fixtures, approved materials, and the intended line, satisfies every criterion below. If a criterion is open, decide explicitly whether to delay or to accept a documented, time-limited conditional release with a capped pilot run.
| Gate criterion | Evidence required |
|---|---|
| Dimensional conformance | Measured report against the drawing, with critical dimensions flagged |
| Cosmetic standard | Signed limit samples plus a written accept/reject definition |
| Function and reliability | Test results on production-representative units |
| Compliance | Applicable test reports or declarations in your brand’s name |
| Packaging and labeling | Approved artwork, drop/transit test result, barcode scan check |
| Golden sample | Sealed, signed, dated, distributed to both parties |
| Change control | Written clause requiring re-sampling after any material or process change |
Tradeoff: releasing tooling early compresses schedule but shifts risk to you, because tooling corrections after approval are usually charged and always delay launch. If you need help sequencing these gates with your supplier, start a project with Yoursender.
Practical Sampling Gate Checklist
- Prototype approved in writing with a defined purpose
- Production sample drawn from a production run, not a bench build
- Dimensional report reviewed and deviations listed
- Cosmetic limit samples signed by both parties
- Function, reliability, and compliance results on file
- Packaging, labeling, and transit test approved
- Golden samples sealed and stored in both locations
- Open issues logged as waivers with owner and expiry date
- Change-control and re-sampling clause in the purchase order
- Written production release signed by a named approver
Frequently Asked Questions
How many production samples should I approve?
Approve a set drawn from one production run: a sealed golden sample per party, one retained for testing, and one for the factory floor. Record the run, date, and serials.
Can I approve with minor cosmetic deviations?
Yes, if the deviation is written, bounded by a limit sample, and time-limited. Undocumented tolerance becomes the new standard.
Is prototype approval a formal gate?
Treat it as one. It confirms design intent and freezes the specification that production samples are measured against.
Who signs off?
One named approver with authority to delay launch, supported by quality, packaging, and compliance inputs.
What happens if the factory changes a material?
The change-control clause should require notification and a fresh sample round before shipment. Verify it at every reorder.
Conclusion
Sampling is a gate, not a milestone. Freeze the specification at prototype, prove it on production tooling, document every accepted deviation, and release tooling only against signed evidence. That discipline is what keeps launch dates and margins intact.
Planning a product or sourcing project? Discuss your sourcing project with Yoursender.