Pre-shipment inspection is the last point in a custom waterproof bag program where a problem is still cheap. After the container closes, the same defect becomes a claim, a return wave, a marketplace complaint rate, or a conversation with a retail buyer who will not repeat the order. Yet most buyers treat inspection as a formality — a photograph of an open carton and a sentence saying everything looks good — when the value of the exercise is entirely in its structure: what is sampled, what is measured, what is tested, and what the findings entitle the buyer to do.
This guide is that structure. It walks the seven checks in the order an inspector should run them, from carton verification through waterproof-specific testing to the accept-or-reject decision, with the sampling logic behind each one and the documentation that makes the outcome enforceable. The production baseline at QUANZHOU JUNYUAN BAGS — custom waterproof bag production since 2014, 4,950 m² SGS-verified facility — is MOQ 500 pieces per style, sampling in 6–10 working days and bulk in 35–50 days, FOB Xiamen. Buyers are welcome to appoint their own inspector against this document — a shared checklist is the fastest path through the stage. Sampling levels and acceptance numbers follow the published AQL tables maintained by ISO, so an inspector and a factory reading the same standard reach the same verdict.
What an inspection can and cannot do
A pre-shipment inspection is a sampling exercise on a finished lot, and understanding that boundary prevents both over-trust and false comfort. It can verify quantity, assortment, packaging, labeling, workmanship, dimensions, material identity and — when tests are built into it — waterproof function on a sample. It cannot guarantee that every unit in the carton is good, because sampling inherently passes some defects; it cannot diagnose root cause, which is the job of the corrective-action loop; and it cannot substitute for the in-process controls that determined the lot’s quality in the first place.
This is why a pre-shipment inspection checklist works best as the last link in a chain rather than the only link. The chain runs specification, then in-process control, then factory final inspection, then the buyer’s or third party’s pre-shipment inspection — and each link changes what the next one finds. A program with real in-process control arrives at inspection with few surprises; a program without it arrives with discoveries, and discoveries at that stage are expensive because the lot is already built. Our quality assurance guide builds the upstream half of that chain, and an AQL inspection is the sampling plan the final link runs on.
Before the inspector arrives: the documents that must exist
- The approved specification: materials by supplier and code, dimensions with tolerances, color with a delta-E limit.
- The reference sample identified by ID and revision, with its photographs and measurements on file.
- The agreed defect classification: critical, major and minor, with examples and tolerances for each.
- The AQL level and the acceptance and rejection numbers for each defect class.
- The packing list and assortment breakdown: styles, colorways, sizes, carton count and units per carton.
- The labeling and marking requirements: origin, fiber content, care, barcode, and any retail-specific marks.
- The test methods to be run on the day: leak test method, sample size and acceptance criteria.
An inspector arriving without that file can only report opinions, and opinions do not survive a dispute. An inspector arriving with it can report measurements against an agreed standard, which is the difference between a conversation and a decision. Prepare the file when the order is placed, not when production finishes — third-party inspectors typically need two to four days of scheduling notice, and the file is the reason the visit produces something usable.
One procedural point deserves emphasis because it decides outcomes: inspection should happen after the lot is packed, not before. Packing reveals damages and assortment errors that do not exist when units are loose on a table, and an inspection conducted on unpacked goods cannot verify the carton count, the marking or the packed presentation that the customer will actually receive. Insist on packed-goods inspection unless the program explicitly requires an inline check instead.
Step 1: quantity, assortment and carton verification
Start with arithmetic, because the most expensive inspection findings are usually the simplest. Count the cartons against the packing list, verify the assortment — each style, colorway and size against the ordered breakdown — and confirm units per carton by opening a sample of cartons rather than trusting the label. Check carton numbering and any shipping marks required by the destination or by a retail channel, and record the carton condition: crushed, damp or poorly sealed cartons are a finding in themselves because they predict damage in transit.
Short shipments and assortment swaps are more common than buyers expect, especially on mixed orders where several styles share a container. A program that ordered three colorways at 200, 150 and 150 and receives 180, 170 and 150 has a merchandising problem on arrival, not just a counting discrepancy — the colorway mix was a decision, and the decision has been changed without approval. Record the actual counts, photograph the carton labels, and put the variance in the report before moving to product checks.
Step 2: packaging and labeling checks
Packaging is part of the product, and labeling errors are the category of defect most likely to cause a customs or channel problem rather than a customer complaint. Verify each element against the approved specification: individual polybag type and any print or recycling mark, hangtag content and placement, insert cards, retail box or header card artwork, barcode legibility and number, and the master carton marking. Confirm that the origin, fiber content and care labeling are present, accurate and permanent — a missing or removable origin label is a customs finding, not a cosmetic one.
Check the packing method as well as the materials: whether units are folded or stuffed in a way that creates permanent creasing, whether straps are secured so they do not abrade during transit, whether hardware is protected from scratching adjacent units, and whether the carton is filled enough to avoid collapse but not so tightly that welded seams are held under pressure for weeks. Our shipping damage prevention guide covers the failure modes this step is looking for, and the import and customs guide explains why labeling accuracy matters far beyond the warehouse.
Step 3: visual and workmanship inspection
Workmanship inspection runs against the defect classification, and it is where an AQL inspection earns its keep. Sample per the agreed plan, then examine each unit systematically rather than holistically — the same sequence every time, so nothing is skipped when the inspector is on unit forty. Seams: uniformity of the weld bead, absence of scorching or unbonded sections, complete termination at corners. Tape: full coverage over the stitch line, no lifted ends, no adhesive squeeze-out. Panels: alignment, symmetry, no stress whitening at fold points. Edges: clean trimming, no raw or frayed material. Interior: liner orientation, pocket geometry, finishing of raw edges.
Classify every finding as critical, major or minor per the agreed definitions and count them separately, because the acceptance decision is made per class rather than in total. A unit with one critical defect is a critical unit; a unit with three minor defects is a minor unit, not a major one. That counting discipline is what makes the outcome defensible, and it is why the classification must be written before production rather than negotiated in the factory’s conference room on inspection day. Our common defects guide documents the failure modes with photographs worth attaching to the file.
Step 4: dimensions, weight and material verification
Measure, do not estimate. Overall dimensions against the specification with tolerances, strap lengths, volume by a consistent fill method, and weight per unit — weight being the single most informative cheap measurement in the whole inspection, because a unit that is eight percent lighter than the approved reference almost always has eight percent less material somewhere: thinner film, lighter coating, smaller hardware, fewer reinforcement layers. Compare against the retained reference sample rather than against the drawing alone, since the drawing and the approved unit occasionally disagree and the unit is what was signed.
Material identity is harder to verify on the day and worth doing anyway: check the material lot record against the production order, confirm supplier and code where documentation is available, and — where the factory has in-house capability — ask for a current film thickness or coating weight measurement for the lot. A silent substitution is the highest-impact failure in waterproof sourcing and it is invisible in a finished unit, which is exactly why the assurance system puts material verification at the receiving door rather than at final inspection. The inspector’s job here is confirmation, not detection.
Step 5: waterproof-specific tests on the day
| Test | Method on inspection day | Sample size | Acceptance |
|---|---|---|---|
| Leak test (filled) | Fill with air or water, submerge or spray per agreed method | 3–10 units per lot | Zero failures |
| Leak test (empty, rolled) | Roll top per spec, apply light internal pressure | 5 units per style | Zero failures |
| Weld visual | Full seam length under adequate light | All sampled units | No unbonded sections |
| Tape adhesion | Attempt peel at seam ends by hand or with a gauge | 3–5 units | No lifting |
| Closure function | Cycle zippers and buckles ten times | 5 units | No sticking or failure |
| Seam peel (destructive) | Cut coupons, pull to failure | 2–3 coupons | Above minimum newtons |
Two of those six are non-negotiable in a waterproof program: a leak test on filled units and a visual seam examination. The rest can be scaled to the risk profile of the product. A beach tote may warrant three leak-tested units per lot; a rescue dry bag warrants a hundred percent leak test, because the consequence of failure is not a return but an injury. That scaling decision belongs to the buyer, and it belongs in the specification before production rather than in a discussion after the cartons are sealed.
Method discipline matters more than sample size. A leak test with no agreed immersion depth, duration and inspection criteria is not a test; it is an activity. Write the method into the order — submersion depth in meters, duration in minutes, what constitutes a failure, and how many units — and the inspection produces a number rather than an impression. The seam integrity testing guide defines the methods, and the IPX guide explains what a rating obliges you to prove.
Step 6: function and hardware testing
Function failures are the defects customers actually complain about, and they are rarely visible. Cycle every mechanical element on the sampled units: buckles and ladder locks opened and closed ten times, strap adjusters loaded and released, roll-top closures rolled and unrolled, valves opened and seated, zippers tracked end to end under light load. Load-test the carrying system — fill a sample unit to its rated weight, hang it for a defined period, and check strap attachment points, webbing anchoring and any welded or taped junction carrying that load.
Watch specifically for the interactions that waterproof construction creates: a stiff welded seam near a flex point, a taped seam that bridges a fold and cracks when the bag is rolled, hardware mounted through a reinforced panel whose reinforcement changes the weld geometry. These are design-level interactions and they show up in function testing rather than in visual inspection, which is why the sampling plan should include at least a few units that are exercised rather than merely looked at.
Step 7: branding and retail presentation
Branding defects are usually major rather than critical, and they are the ones a retail buyer notices first. Check placement against the approved drawing with a measurement rather than by eye, color against the approved reference under consistent lighting, edge definition and registration, adhesion after flexing the panel, and durability by a tape-pull or rub test appropriate to the method. Verify the retail presentation as a whole: hangtag straight, barcode scannable, polybag clean and uncreased, box artwork aligned.
Then check the packed unit the way a customer will meet it. Open one retail-ready carton completely and unpack it as a store would — because defects that appear only in packed form, such as creasing from folding, strap marks pressed into face fabric, or scuffing between units, are invisible in loose inspection and universal by the time the product reaches a shelf. This step costs ten minutes and prevents the most common class of "the bags were fine but the shipment looked bad" dispute.
Recording findings: the report that is actually usable
| Report element | Why it matters | Common failure |
|---|---|---|
| Sampling plan stated | Shows the basis of the conclusion | Sample size not recorded |
| Findings by defect class | Drives the acceptance decision | Findings listed without class |
| Photographs with scale | Makes a finding indisputable | Photos without context or reference |
| Measurements vs spec | Converts opinion into data | "Looks fine" instead of a number |
| Test method and results | Makes performance verifiable | Test done, method unstated |
| Clear conclusion | Accept, reject or conditional | Ambiguous "subject to" language |
The report is a legal and commercial document, not a courtesy. It should state the sampling plan actually executed, list every finding by defect class with a count, attach photographs that include a reference for scale and a caption explaining what is wrong, record measurements against the specification rather than impressions, state the test methods and their results, and end with an unambiguous conclusion. A report that says "minor issues found, acceptable overall" creates a dispute later; a report that says "four major defects against an acceptance number of three: reject for rework" creates a decision.
Photographs deserve special attention because they are what makes a remote inspection useful. Each finding should be photographed three ways: a context shot showing which part of which product, a close shot showing the defect, and a comparison shot next to a conforming unit or the reference sample. That trio removes the argument about whether the finding is real, which is the argument that otherwise consumes the next two weeks.
The accept, reject or conditional-release decision
| Outcome | When it applies | What it triggers |
|---|---|---|
| Accept | All classes within acceptance numbers | Balance payment and shipment release |
| Accept with noted minors | Minors above plan, buyer waives | Written waiver plus next-lot action |
| Reject for rework | Criticals or majors above acceptance | Sorting or rework, then re-inspection |
| Reject for replacement | Systemic defect not repairable | New production slot, revised delivery |
| Conditional release | Shipment deadline outweighs findings | Credit note or price adjustment agreed |
The decision should be driven by the numbers rather than by the calendar, and that is easier to say than to do when a vessel is booked and a launch is advertised. The disciplined approach is to agree the remedy in the order — what happens on rejection, who pays for rework, who pays for re-inspection and for the freight delay — so that the decision at inspection is an application of terms rather than a negotiation under time pressure. Conditional release is legitimate when the findings are genuinely minor and the deadline is real, but it should be accompanied by a written credit or price adjustment agreed before the container moves, not promised after.
Rework deserves a specific caution: sorting a lot and shipping the good units is not rework unless the root cause is known and addressed. If four percent of units fail because of parameter drift on one shift, the remaining ninety-six percent carry elevated risk even if they pass. Rework plus re-inspection is the correct sequence, and the re-inspection should sample from the full lot rather than only from the reworked portion.
After inspection: rework, re-inspection and release
The post-inspection sequence closes the loop. Rework with a defined method and owner; re-inspection against the same sampling plan and the same classification, not a relaxed one; root-cause analysis on any critical or systemic finding; and then release — balance payment, document handover, container loading and vessel booking. Each step has an owner and a date, and each is written down, because the program that skips root cause will meet the same defect on the next order with the same launch pressure attached.
Finally, file everything. The inspection report, the photographs, the test results and the corrective action belong in the style’s quality folder alongside the specification and the reference sample record, which is what makes the next reorder faster and the next dispute shorter. Our own programs run this document as a shared standard — buyers may appoint their inspector and we will work to the same checklist — and the process discipline behind it follows ISO 9001 quality management principles. The full production chain from specification to loaded container is documented.
Frequently Asked Questions
Q1. What is a pre-shipment inspection?
A sampling inspection of the finished, packed lot before it ships, run against an agreed specification, defect classification and AQL plan. It is the last point where a defect is still cheap to fix.
Q2. Should inspection happen before or after packing?
After packing. Packing reveals damage, assortment errors, marking and labeling problems that do not exist when units are loose, and the packed unit is what the customer actually receives.
Q3. How many units should be leak-tested?
Scale to consequence: three to ten per lot for general-purpose bags, one hundred percent for products where failure is a safety issue. Fix the number in the specification before production, not at inspection.
Q4. What documents does an inspector need before arriving?
The specification with materials by supplier code, the reference sample by ID and revision, the defect classification, the AQL level and acceptance numbers, the packing list, labeling requirements and the agreed test methods.
Q5. Can inspection guarantee the whole lot is good?
No. AQL is a sampling plan and passes some defects by design. Close the gap with functional testing on the attributes that matter and a written remedy for a rejected lot.
Q6. What is the difference between critical, major and minor defects?
Critical makes the product unsafe or non-functional; major affects saleability; minor is unlikely to be noticed by a customer. Each class has its own acceptance number, and classification is written before production.
Q7. Why is unit weight worth measuring?
Because a unit eight percent lighter than the approved reference almost always has eight percent less material somewhere — thinner film, lighter coating, smaller hardware. It is the cheapest high-signal measurement in the inspection.
Q8. What should an inspection report contain?
The sampling plan executed, findings counted by defect class, photographs with scale and captions, measurements against specification, test methods and results, and an unambiguous accept, reject or conditional conclusion.
Q9. What happens if a lot is rejected?
Rework with a defined method and owner, then re-inspection against the same plan and classification, plus root-cause analysis. Sorting and shipping the rest is not rework if the root cause is unaddressed.
Q10. Should I accept a lot with minor defects above the plan?
It is your call, but record it: a written waiver for this lot plus an action for the next. Undocumented waivers become the factory’s new baseline on the following order.
Q11. Why inspect packed units as a store would?
Because creasing from folding, strap marks pressed into face fabric and scuffing between units appear only in packed form — invisible in loose inspection and universal by the time the product reaches a shelf.
Q12. How much notice does a third-party inspector need?
Typically two to four days for scheduling. Book at order placement rather than at production completion, otherwise inspector availability becomes the delay that pushes the vessel.
Q13. What is conditional release?
Shipping despite findings because the deadline outweighs them. Legitimate for genuinely minor findings, but it should carry a written credit or price adjustment agreed before the container moves.
Q14. Can I use my own inspector at the factory?
Yes, and good factories welcome it. Working to a shared checklist written before production is the fastest path through the stage, because there is nothing to argue about on the day.
Q15. How does inspection fit with in-process quality control?
Inspection confirms; process control determines. A program with real in-process weld and leak control arrives at inspection with few surprises; one without it arrives with expensive discoveries.
Q16. What should I file after inspection?
The report, photographs, test results and any corrective action, in the style’s quality folder with the specification and reference sample record — so the next reorder is faster and the next dispute shorter.
People Also Ask
What is a pre-shipment inspection?
An inspection of the finished, packed lot before shipment, run against an agreed specification and sampling plan. It is the final checkpoint where defects can still be fixed cheaply.
What should be on a pre-shipment inspection checklist?
Carton and assortment count, packaging and labeling, visual workmanship by defect class, dimensions and weight, waterproof leak and seam tests, hardware function, branding, and a clear accept or reject conclusion.
How many bags should be inspected?
Per the AQL plan agreed in the order — sample size scales with lot size and inspection level. Functional tests are additional and scaled to the consequence of failure.
Can pre-shipment inspection catch waterproof failures?
Only if testing is built into it. Visual inspection cannot see a weak weld bond; a defined leak test with agreed depth, duration and acceptance criteria can.
What happens if my inspection fails?
Rework with a defined method, re-inspection against the same plan, and root-cause analysis. The remedy — who pays for rework and delay — should already be written into the order.
Should I inspect before or after packing?
After packing. Packed-goods inspection is the only way to verify carton count, marking, labeling and the packed presentation the customer will receive.