Pre-shipment Inspection for Knife Orders: Timing and Scope

When PSI should happen

Book the pre-shipment inspection for a knife order after production is finished, packing is nearly finished, and the goods are still inside the factory. In our Yangjiang shop, that means 80-100% complete and at least 80% packed, with some cartons still open so QC can pull samples by carton mark. Too early is risky. At 50-60% completion, the inspector is checking goods that can still change during the last 2 days of polishing, washing, and packing. Bad timing bites. We have seen a clean sample pass on Wednesday, then the grinding line rushed the remaining steak knives on Friday and left burrs on 7 cartons after the final master cartons were taped.

For a PSI knife order, inspection should sit after the factory's final QC, but before the warehouse team seals all master cartons or stages the container. That is when rework still makes sense: swap a scratched blade, replace a wrong color handle, or fix a barcode label before 240 cartons are stacked at the dock. For a third party inspection knife assignment, ask the inspector to verify production status against the packing list, sample products by carton number, check pack-out against the PO, and confirm loading readiness with carton count and shipping marks. If the supplier says the goods are urgent, this is the wrong question to ask. Urgent goods often need a later inspection slot, because late defects show up after final polishing, blister sealing, or the last barcode sticker run.

Timing also depends on order risk. A private-label retail order with printed boxes, inserts, barcodes, color sleeves, and customer SKU labels has more places to fail than a plain bulk carton order. We once had QC pull the sample and find the PO wrote "8 inch chef knife" while the insert said "20 cm santoku"; small typo, big retailer complaint. For Europe or North America, use the PSI window to check compliance papers such as REACH-related declarations, LFGB or FDA contact claims, plus the real shipping label content on the carton. A factory in Yangjiang, China can move fast, but speed does not fix bad sequence. Lock the sequence: product complete, packaging complete, documents complete, loading approved.

What to check on the knives

Start with the knife, not the carton. A clean master carton will not rescue a bad blade; we have seen repeat orders fail this way. For kitchen knives, chef knives, outdoor knives, and pocket knives, inspect the blade first: geometry at the bevel, edge line from heel to tip, surface finish under 600-800 lux packing-table light, handle fit, lock action, pivot tension, tip safety, and visible cosmetic defects. If the spec says 56-58 HRC or 58-61 HRC, check the actual hardness band against the approved sample on the Rockwell tester, not just the purchase order. We run into this often. A knife can pass hardness and still fail because the grinding line left the right bevel heavier, or because QC finds a 0.4 mm handle scale gap near the second rivet.

Use the approved sample to judge the defects that matter in your sales channel. On a chef knife, place the blade on a flat plate for straightness, cut copy paper for edge bite, check spine polish with a finger pass, inspect bolster blending, and test for blade wobble where the tang enters the handle. On a pocket knife, open and close it 5-10 times, then check lockup depth, blade centering, pivot feel, and whether the pocket clip sits straight; last month QC pulled a sample where the clip was 1.5 mm off center, and the buyer flagged it before shipment. On a hunting or tactical knife, measure sheath retention with a pull gauge, check edge protection inside the sheath, and inspect the finish points where sweat or salt spray usually starts rust. For Damascus knives, pattern consistency matters, but coating adhesion and grind symmetry decide whether the knife looks cheap after one week in a display case.

• Critical: broken tip, lock failure, loose handle, blade play, rust, or sharp burrs on touch points that can cut the user during normal handling.
• Major: uneven edge, off-center blade, wrong logo placement, poor polish, or handle parts outside the drawing size by more than the agreed tolerance.
• Minor: faint scratch, slight print shift, or cosmetic variation that does not affect cutting, folding, sheath fit, or safe use.

Do not let the inspector stop at appearance. A useful PSI report connects what QC saw to the approved specification. If your product was approved at 2.5 mm spine thickness and 1.2 kg pull-out force on the sheath, the report needs readings from a digital caliper and pull gauge, or a clear note saying why those items were skipped. “Does it look okay?” is the wrong question. Ask whether the bulk shipment still matches the signed sample after heat treatment, grinding, polishing, logo marking, and packing. That is how we keep a China supplier aligned after the first order, not just during sample approval.

Pack-out and label control

About 4 out of 10 knife shipment complaints we see are not blade problems; they are pack-out problems. The carton looks clean outside, then QC opens it with a tape cutter and finds the wrong insert, no 5 g silica gel, mixed SKUs, or a barcode that fails on a Zebra scanner at 300 mm. For importers, packaging is product quality. Put it on the PSI checklist. Do not leave it as an office job after production, because we have seen this go sideways when the buyer flagged one missing warning label and held 1,200 cartons at receiving.

For Amazon or retail chain orders, check FNSKU position, scan result, and carton content label against the PO line by line. One typo in a SKU, such as 8IN-CH instead of 8IN-Chef, can stop a whole pallet. If you sell into the EU, check whether the buyer spec calls for LFGB, REACH, or food-contact declarations in the file set. If the knives are packed in China for export, the PSI sample should include one sealed retail unit and one open sample, so the inspector can check the print and the actual knife, sheath, insert, silica gel, and manual inside. Small packing mistakes cost more than they look. The math does not work when a USD 0.03 label blocks receiving for 12 days. Treat the packing section of PSI like a production control step, the same way we run blade hardness checks on the grinding line.

Third party inspection leverage

A third party inspection knife report changes behavior for one simple reason: the inspector is not on our payroll. Factory QC sees the grinding line backlog, the ETD, the sales relationship, and the 7 cartons already staged near the loading door. The third party inspector sees the spec sheet, caliper readings, carton marks, barcode scans, and the finished goods on the table. That gap gives the buyer leverage. If a supplier thinks the shipment will pass no matter what QC pulled from the sample table, last-mile checks get loose. If they know one outside report controls release, they fix bent tips, loose handles, wrong EAN labels, and thin edge guards before the inspector walks in.

Use the report as a release tool, not a bargaining note. This is the wrong question to ask: “Can you give us a better price because inspection failed?” The better question is: “What must be reworked before release?” The inspector should sample against the agreed AQL, usually AQL 2.5 for major defects and a stricter limit for critical safety issues. For a 3,000-piece knife order, a sampling standard aligned to ISO 2859-1 or ANSI/ASQ Z1.4 is usually enough, but only if the product classification is locked. If the shipment mixes kitchen knives, gift boxes, and spare sheaths, define the lot by SKU and pack style before inspection. We have seen this go sideways: one PO typo changed “black sheath” to “black sleeve,” and the supplier tried to bury 3 pack styles inside one broad lot.

Good leverage starts with the action path written before the visit. If the report finds one major defect on a safety-critical item, the batch is on hold. Stop shipment. If it finds cosmetic minor defects within tolerance, the batch can ship with a documented concession. If it finds a pattern, the math does not work on promises; ask for rework, reinspection, and a corrected CAPA with photos from the packing bench. We run this better when QC pulls the sample under camera, checks blade tip alignment with a 0.5 mm feeler gauge, and records the carton number beside each defect photo. In Yangjiang, China, the factories that ship consistently are the ones that know the buyer will read the report line by line, from blade HRC notes to master carton drop damage. That is the point of a third party inspection knife program. It turns quality talk into evidence.

Container loading check details

A container loading check is not optional if one wrong container eats your margin. PSI tells you what QC saw in opened cartons. Loading check proves what passed through the container doors. Different control. We have had a clean PSI, then found 37 short cartons, two swapped SKU marks, crushed color boxes on the bottom layer, no seal photo, and one 40HQ with a damp floor that softened outer cartons before Ningbo departure. Bad surprise.

The loading check should confirm three things before loading starts: the truck brought the right cartons, the count matches the final packing list, and the container is clean enough for knife goods. Open the doors. Use a flashlight. QC pulled one container last May because the left wall had rust flakes and a 6 mm pinhole near the floor rail. The inspector should photograph the container number, seal number, carton stack pattern, plus the first and last cartons loaded. If the order is mixed by SKU, record the loading sequence by carton mark, such as KCH-8PC-BLK first and KCH-5PC-WD last, so the destination warehouse can trace any shortage without guessing. If pallets are used, check pallet height, stretch-wrap tension, corner boards, and bottom-layer strength. 1,650 mm is fine for some master cartons. The math does not work when thin retail boxes sit under cast iron handles; we have seen gift boxes bow after only 22 minutes on the loading bay.

• Verify container number and seal number before the first carton goes in.
• Count cartons against the final packing list, not the draft version the merchandiser sent last week.
• Check for water, rust, holes, oil smell, or chemical odor inside the container.
• Confirm no unauthorized cartons, samples, spare blades, or marketing inserts are added.
• Record photos of closed doors and the final seal placement, with the seal digits readable.

For FOB shipments from China, the loading check protects your handover point. For DDP shipments, it still matters because the freight chain can move from factory truck to port yard, vessel, bonded warehouse, then last-mile carrier after departure. If a supplier in Yangjiang, China says the container is fine without opening it, push back. This is the wrong question to ask. Open it, inspect it, count it, photograph it, seal it, and record it. We run this with a timestamped phone album and one loading sheet signed before the truck leaves the gate. That turns a shipping event into a documented shipment.

How to decide release or hold

Make the release call on two points: defect risk and whether the line can rework it before the truck backs in. A missing label, wrong insert, or loose carton flap is a packing job; our packing team can clear 300-500 cartons in one afternoon with a tape gun, label roll, and the revised carton mark sheet clipped to the bench. Safety defects are different. Blade play at the pivot, a sharp burr left after the grinding line, incorrect HRC, or a lock that fails under normal hand pressure affects use and compliance. Hold it. QC should pull the sample, tag the failed cartons with red stickers, and release only after rework and reinspection are signed off.

Do not accept “we will fix before loading” on WeChat as proof. Ask for a written corrective action plan showing root cause, repair method, affected lot quantity, and reinspection date. If the supplier asks for partial release, write down the exact quantity, SKU, and carton numbers allowed to ship. Otherwise the warehouse will mix good cartons with suspect cartons; we have seen this go sideways over one wrong PO suffix typed as “-A” instead of “-B”. For knife orders, partial release works when the issue is packaging or print, such as a barcode sticker placed 20 mm off position. It does not work when the defect is blade performance or assembly consistency. The math doesn't work.

Use shipment value in the decision, but do not let it push you into a bad release. A 5,000-piece order with a 35-45 day lead time and a 500-piece MOQ may not deserve endless delay, but shipping known defects costs more than holding the pallets for 2 days of rework. The supplier in China wants the container out; the buyer wants sellable knives. A PSI gives you the right to keep goods on site until the evidence is clean. Check the report, close-up photos, carton marks, and container loading record against the same SKU list, not against memory. QC pulled one sample last month where the carton mark was correct but the inner box showed the old logo. When they match, release. When they do not, hold.

Frequently asked questions