Quality Control

AQL Inspection Specs for Custom Pins, Coins and Keychains

Why Approved Samples Still Become Uneven Cartons

Most disputes on custom metal giveaways do not start with a completely wrong product. They start with an approved pre-production sample, followed by cartons that vary: nickel stains in several trays, soft enamel dust under epoxy, loose butterfly clutches, scratched coin backs, jump rings with open gaps, or QR codes that scan in one carton but not another. If the purchase order only says “good quality” or “same as sample,” the buyer and factory have no shared rule for how many defects are acceptable.

AQL, or Acceptable Quality Limit, converts that expectation into an inspection rule. It defines the sample size, defect classes, accept and reject numbers, and the action required when a lot fails. For enamel pins, challenge coins, keychains, brooches, fridge magnets and zipper pulls, AQL must be combined with product-specific tolerances. A 0.3 mm rack mark on the back rim of an antique coin may be acceptable; the same mark across a mascot face on a retail pin is not.

At ZheCraft, AQL language is used on export orders because it reduces subjective arguments at shipment time. It does not replace a golden sample, process control or factory sorting. It gives procurement teams, third-party inspectors and production QC staff the same measurable standard before cartons leave the factory.

Choose AQL Levels by Defect Risk

For most promotional metal products, a practical starting point is critical AQL 0, major AQL 2.5 and minor AQL 4.0 under ISO 2859-1 general inspection level II. This is stricter than many low-price gift programs but still realistic for die-struck brass, stamped iron, zinc alloy casting, plating, enamel filling and individual packing. For paid merchandise, museum shop coins, licensed character pins or event-critical launches, buyers often tighten major defects to AQL 1.5.

Do not assign AQL 0 to every cosmetic issue. Requiring zero minor scratches on a 10,000 piece soft enamel pin order can force 100% screening, increase handling marks, add 2 to 5 production days and raise FOB cost by about USD 0.03 to 0.12 per piece depending on packing complexity. The better rule is to make safety, compliance and functional failures non-negotiable while allowing controlled cosmetic variation outside the primary viewing zone.

Defect class	Typical AQL	Examples for metal giveaways	Usual decision
Critical	0	Sharp exposed pin tip, failed lead or nickel release requirement, missing child-safety lock where specified, detached magnet creating ingestion risk	Reject lot, quarantine, or 100% screen with documented correction
Major	1.5 to 2.5	Wrong Pantone family, loose attachment, plating blister on front, broken keyring, non-scannable QR code, size outside tolerance	Fail if accept number is exceeded
Minor	4.0	Tiny back-side plating speck, slight enamel meniscus variation, light scratch under 8 to 10 mm on non-logo back, small polybag wrinkle	Accept if within AQL limit

Lock the Sampling Plan Before Production

For normal export orders from 500 to 20,000 pieces, general inspection level II is the safest default because it balances risk and inspection cost. Special inspection levels such as S-2 or S-3 should be reserved for destructive or time-consuming tests, including post pull testing, magnet pull testing, salt spray checks, repeated split-ring opening or carton drop testing. Those samples may be damaged and cannot always return to sellable stock.

A 3,000 piece pin order inspected under ISO 2859-1 general level II uses code letter K, so 125 pieces are inspected visually. With AQL 2.5 for major defects, the lot passes at 7 major defects and fails at 8. With AQL 4.0 for minor defects, it passes at 10 and fails at 11. A 10,000 piece order under the same level uses code letter L, or 200 inspected pieces; at AQL 2.5 it passes at 10 major defects and fails at 11.

AQL is weaker on very small orders. For sample runs, Kickstarter add-ons or boutique batches under 300 pieces, ZheCraft usually recommends 100% factory visual sorting plus a final buyer-approved photo record rather than relying only on statistical sampling. For orders over 50,000 pieces, carton distribution matters: the inspector should pull from early, middle and late production cartons, not just the most accessible pallets.

Define Defects by Product Zone

AQL only works when the defect list is specific. Words like “scratched,” “bad plating” and “wrong color” are too vague because custom metal items have different viewing zones. Define Zone A as the front logo face, character face, readable text, QR code, enamel color fields and any retail-facing surface. Define Zone B as side walls, rims, backstamps and attachment areas. Define Zone C as hidden hardware contact points, split-ring overlap, clutch interior and the inside of a brooch hinge.

For hard enamel pins, major defects should include sunken enamel more than 0.15 mm below metal lines, polishing burn-through, color contamination visible at 30 cm, and metal line breaks wider than 0.20 mm in Zone A. For soft enamel, lower fill is expected, but missing enamel, bubbles over 0.5 mm in Zone A, dust trapped under epoxy, or enamel bleed that changes the artwork should be major. For printed pins and coins, registration drift over ±0.25 mm on fine logos should normally be major.

Product	Major defect examples	Minor defect examples
Enamel pin	Wrong attachment, bent post over 5 degrees, front plating blister over 0.5 mm, missing enamel, Pantone clearly off-family	Back-side speck under 0.3 mm, slight enamel level variation within approved sample range
Challenge coin	Diameter outside ±0.3 mm, unreadable text, plating peel, enamel bleed into relief details, antique wash hiding artwork	Edge rack mark under 1.0 mm, antique wash variation within signed sample range
Metal keychain	Jump ring gap over 0.3 mm, split ring deformation, broken swivel, logo offset over ±0.5 mm	Light back scratch under 8 mm, tiny casting pit outside logo area
Fridge magnet	Magnet detaches under pull test, wrong polarity for paired set, weak hold on 0.8 mm steel plate	Glue overflow under 0.5 mm not visible from front

Specify Tolerances, Plating and Compliance

Dimensional tolerances should be written on the drawing and repeated in the final inspection standard. For die-struck brass, iron or zinc alloy pins under 40 mm, a realistic finished-size tolerance is ±0.20 mm after plating and polishing. For 40 to 60 mm challenge coins, ±0.30 mm is practical. Cast zinc alloy keychains with 3D relief, open cutouts or moving parts may need ±0.40 mm unless CNC finishing is quoted.

Thickness affects perceived value, packaging fit and hardware strength. A 1.5 mm pin can usually hold ±0.15 mm, a 2.0 mm badge ±0.20 mm, and a 3.0 mm coin ±0.25 mm. Weight tolerance should be ±5% for stamped brass or iron and ±8% for cast zinc alloy unless the order requires tight retail consistency. If a coin is sold as 45 g, write the allowable range, such as 42.8 to 47.3 g, rather than leaving it subjective.

Decorative nickel, gold, copper or black nickel plating for indoor promotional pins is commonly 3 to 5 microns total. Keychains and challenge coins receive more abrasion, so 5 to 8 microns is safer. Premium retail programs may specify 8 to 12 microns plus 24 to 48 hours neutral salt spray for nickel, brass or imitation gold finishes. Standard decorative plating should not be expected to behave like industrial outdoor hardware without upgraded process controls.

Compliance requirements should be quoted before sampling. Nickel-free plating, CPSIA lead limits, EN 71-3 heavy metal migration, REACH SVHC screening or California Proposition 65 documentation can add 3 to 10 days if laboratory testing is required. For children’s products or wearable jewelry, critical defects include failed compliance results, sharp points, detachable small magnets and missing warning labels where specified.

Add Functional Tests for Hardware

Visual inspection is not enough for products that attach to clothing, keys, bags or steel surfaces. A pin can look perfect but fail because the post weld is weak. A keychain can pass color inspection but return from the customer because the jump ring opens after two weeks. Functional tests should be written into the same inspection sheet as cosmetic defects.

For lapel pins, specify a post pull test of 2.0 kgf for standard 8 to 10 mm posts and 3.0 kgf for larger brooch pins. Pin posts should sit within ±0.5 mm of the approved location and lean less than 5 degrees. Butterfly clutches should fit firmly without wobble; rubber clutches should not split after five install-remove cycles. For magnetic badges, check both holding force and fabric suitability because very strong magnets can damage thin scarves or delicate shirts.

For keychains, inspect split-ring hardness, swivel rotation and jump-ring closure. A practical factory test is 20 open-close cycles on the split ring plus a 5 kg static pull for 10 seconds through the assembled chain path. For fridge magnets, test on a clean vertical 0.8 mm painted steel plate; a 50 mm metal or resin magnet should hold its own weight plus light tapping without sliding.

• Confirm all attachments match the approved sample: clutch, safety pin, magnet, split ring, lobster clasp, swivel or ball chain.
• Measure post angle and position; treat more than 5 degrees lean or more than ±0.5 mm location shift as major.
• Check jump-ring gaps; openings over 0.3 mm are major defects for keychains, charms and zipper pulls.
• Run pull tests on a small destructive sample before final packing, not after cartons are sealed.
• Record failed units with photos, carton numbers and defect class so rework targets the true process cause.

Control Color, Readability and Packing

Color disputes often come from inconsistent viewing conditions. Inspect enamel and printed areas under D65 daylight or a light booth when possible, not under yellow warehouse lamps. For most promotional orders, visual match to the approved sample and Pantone reference is more practical than spectrophotometer data. For brand programs, specify Delta E targets before quoting; a practical range is ΔE under 2.0 for printed flat graphics and under 3.0 for enamel, where pigment batches and baking can create small variation.

Inspect plating after polishing, cleaning and final packing because fingerprints, moisture and sulfur-containing paper can cause late staining. Check the front face at 30 cm under normal light and use 5x magnification only for confirming borderline defects. If magnification becomes the main standard, nearly every polished metal product will show micro-scratches and the inspection standard becomes commercially unrealistic.

Logo readability needs its own rule. Text below 1.2 mm high, QR modules below 0.4 mm and laser engraving on dark plating can become unreadable after plating thickness and polishing variation. During inspection, QR codes and barcodes should scan on at least three common phones or scanners. Non-scannable codes should be major defects when they support marketing, authentication, inventory or event entry.

Packing defects should also have tolerances. Individual polybags should match the approved size, header cards should stay within ±1.0 mm registration, and carton quantity variance should be zero unless overage is agreed. Standard export cartons for small metal items are often kept under 15 kg gross weight to reduce crushing and courier handling damage. For Amazon FBA, retail display or event kits, label errors are usually major defects, not minor packaging issues.

Tie AQL to MOQ, Price and Lead Time

Inspection standards affect price and schedule, so they should be included in the RFQ, not introduced after goods are finished. For custom pins and small keychains, typical MOQs are 100 pieces per design for sampling or low-volume orders, 300 pieces for efficient mass production, and 1,000 pieces for better unit pricing. Challenge coins often start at 100 pieces, but 300 to 500 pieces gives a more stable cost spread over tooling.

Indicative FOB China pricing depends on size, metal, plating, enamel count and packing. A 25 to 35 mm soft enamel pin with standard clutch and polybag may range from USD 0.45 to 1.20 at 500 pieces, while hard enamel may run USD 0.65 to 1.60. A 40 to 50 mm zinc alloy keychain commonly ranges from USD 0.85 to 2.20. A 45 mm challenge coin with two-sided relief and antique plating may range from USD 1.60 to 4.50. Tooling is often USD 40 to 120 for pins and USD 80 to 250 for larger coins or complex keychains.

Normal lead time after artwork approval is 5 to 8 days for sampling and 12 to 20 days for mass production, excluding laboratory testing and international freight. Tight AQL 1.5, 100% QR scanning, nickel-free plating, upgraded salt spray, retail carding or individual serial numbering can add 2 to 7 days. These checks are manageable when planned; they become expensive when discovered during final inspection.

Set Rework and Shipment Rules in Advance

The purchase order should state what happens when a lot fails. Without this, the buyer may expect a full remake while the factory expects sorting only. For most custom metal products, the first response is 100% screening for the failed defect category, followed by re-inspection using the same AQL level after rework. The re-inspection sample should include repaired cartons and untouched cartons so the inspector can confirm whether the failure was isolated or systemic.

Some defects are sortable or repairable: loose clutches, wrong polybags, light dust on packaging, mixed carton labels and replaceable jump rings. Other defects are usually not economical to repair, including wrong base metal, wrong mold shape, severe plating peel, baked-in wrong enamel, incorrect glitter mix, or QR codes engraved at the wrong size. Lead time impact should be realistic: sorting and repacking 5,000 pins may take 1 to 3 days; re-plating can take 3 to 7 days if the base surface is usable; remaking metal parts usually takes 10 to 18 days after approval.

Failure type	Best correction	Typical added time	Cost responsibility to define
Packing mix-up	Sort by SKU, re-label cartons, re-count quantities	1 to 2 days	Usually factory if packing instruction was clear
Loose hardware	Replace clutches, rings or clasps, then pull test sample	1 to 4 days	Factory if hardware spec was approved
Front plating peel	Strip and re-plate if possible, otherwise remake	3 to 12 days	Depends on cause and approved sample match
Wrong enamel color	Remake affected color areas or full item	7 to 18 days	Buyer if Pantone changed late; factory if produced against wrong spec
Dimensional mold error	Modify tooling or cut new tooling	7 to 15 days	Factory if drawing was clear; buyer if tolerance was missing

Before the next order, add one page called “Final Inspection Standard” to the RFQ or purchase order. It should include the drawing, golden sample reference, AQL levels, defect zones, dimensional tolerances, plating requirement, functional tests, packing checks, rework method and cost responsibility. Keep it short enough that factory QC can use it at the line, not only in the purchasing file.

• Choose AQL levels by defect risk, not habit.
• Classify defects by Zone A, Zone B and Zone C visibility.
• Write numerical tolerances for size, thickness, weight, plating, hardware gaps and carton quantities.
• Approve one golden sample as the reference for color, texture, finish and packaging.
• Agree on rework rules, re-inspection method, added lead time and cost responsibility before mass production.

Have a project? Send your artwork and target quantity and we’ll reply with a detailed quotation within 12 working hours.