Quality Control

Final Inspection Specs for Custom Pins, Coins and Keychains

Shipment Photos Are Not Final Inspection

A familiar failure pattern looks harmless until cartons arrive: the approved pre-production sample was correct, the supplier sent five clean shipment photos, and the delivered lot still has 4% loose butterfly clutches, cloudy epoxy, mixed plating shades, short cartons, or backstamps rotated 30 degrees off position. For distributors and brand teams, the loss is rarely limited to the unit price. The real cost is missed event dates, manual sorting, replacement freight, client chargebacks, and damaged confidence in the buying process.

Final random inspection is the last practical control point before balance payment and export release. It does not replace correct artwork, mold control, approved samples, or in-process checks, but it does verify whether the finished batch matches the purchase order across cosmetics, dimensions, hardware function, packing, carton marks, and count accuracy. For custom metal products, a useful inspection report must be based on defined standards, not a general request to “check quality.”

At ZheCraft in Yiwu, we write inspection requirements before production starts. The specification names the inspected lot, AQL level, defect classes, sampling size, test tools, measurement tolerances, and the action required if the lot fails. Without those details, two experienced people can inspect the same pin or coin and disagree about whether the defect is acceptable variation or a shipment-blocking problem.

Define the Lot, Sample Size and AQL

Inspect by SKU and production batch. If an order contains 1,000 hard enamel pins across three designs, each design should be treated as a separate lot when the artwork, size, plating, attachment, or packing differs. If the metal item is identical and only the backing card artwork changes, the metal inspection may be combined, but card counts and barcode checks still need to be verified by design.

AQL means acceptable quality limit. It is not a target defect percentage or permission for the factory to ship bad units. It is a statistical sampling method that sets accept and reject numbers for a defined lot. For most B2B promotional pins, coins, medals, brooches, and keychains, a practical default is ANSI/ASQ Z1.4 or ISO 2859-1, General Inspection Level II, with AQL 0 for critical defects, AQL 2.5 for major defects, and AQL 4.0 for minor defects. For licensed retail, museum shops, VIP gifts, or products handled by children, tighten to AQL 1.0 major and 2.5 minor.

Inspection should occur when 100% of production is complete and at least 80% is individually packed. Inspecting too early allows inspected and uninspected goods to be mixed later. Inspecting after export cartons are fully sealed creates avoidable carton damage and repacking labor. The best window for most custom metal orders is after individual packing and inner box completion, before final master carton sealing.

Order quantity per SKU	Suggested sample size	Critical AQL	Major AQL	Minor AQL	Typical inspection time
300-500 pcs	50 pcs	0	2.5	4.0	1.5-2.5 hours
501-1,200 pcs	80 pcs	0	2.5	4.0	2-3 hours
1,201-3,200 pcs	125 pcs	0	2.5	4.0	3-4 hours
3,201-10,000 pcs	200 pcs	0	2.5	4.0	4-6 hours
10,001-35,000 pcs	315 pcs	0	1.5-2.5	2.5-4.0	1 working day

Classify Defects Before the Factory Sorts

Defect classification is what turns an inspection from opinion into a decision. A sharp burr on a pin post, a weak keychain jump ring, and a faint polishing line on the back of a coin are not equal. The purchase order should define critical, major, and minor defects before mass production, so the inspector does not negotiate standards on the packing floor.

Critical defects create safety, legal, compliance, or severe brand risk. Examples include exposed sharp metal edges above 0.2 mm, detachable magnets or decorations that create a small-part hazard, wrong trademark ownership, incorrect QR code destination, missing required country-of-origin marking, or nonconforming paint or plating where a restricted-material requirement was specified. One confirmed critical defect should fail the lot until the factory completes root-cause sorting and re-inspection.

Major defects affect function, sellability, or obvious brand appearance. For enamel pins, major defects include wrong Pantone family, missing enamel area over 0.5 mm, plating blisters, loose posts, wrong attachment, front-face scratches longer than 5 mm, backstamp errors, and size outside agreed tolerance. For challenge coins, major defects include warped edges, incorrect antique finish, uneven 3D relief, misaligned color fill, or capsules that scratch the coin face. Minor defects are small cosmetic issues visible only under close inspection, such as a polishing hairline under 10 mm on the reverse side, a tiny enamel speck away from the logo, or slight plating shade variation within the approved sample range.

• Critical: sharp burrs, detached magnets, wrong trademark, wrong QR or serial sequence, restricted-material nonconformity, missing mandatory origin mark.
• Major: loose hardware, wrong plating, missing enamel, front-face scratches, warped coin, incorrect backing card, wrong polybag or mixed SKU.
• Minor: slight reverse-side polishing marks, tiny dust under epoxy, small color speck outside the logo area, light carton scuffing.
• Escalation rule: multiple minor defects on one piece become major if the combined appearance makes the item unsellable.
• Viewing condition: inspect at 30-40 cm under normal daylight or 600-800 lux white light, without magnification unless the PO requires it.

Measure Size, Weight and Plating Thickness

Final inspection should confirm production matches the approved sample and purchase order, not redesign the product. For die-struck iron or zinc alloy enamel pins, normal size tolerance is ±0.2 mm for items under 35 mm and ±0.3 mm for 35-60 mm items. For cast challenge coins, use ±0.3 mm on diameter and ±0.2 mm on thickness for common 2D or 3D coins, because casting, polishing, and antique finishing remove material unevenly.

Weight is a simple way to catch substitution or thinning. If a 50 mm zinc alloy coin approved at 38 g averages 32 g in production, the supplier may have reduced thickness, changed alloy, or hollowed the design. Use a 0.01 g scale and check at least 10 random units per SKU. A practical tolerance is ±5% for small pins and keychains, and ±3% for coins above 30 g.

Plating thickness should be agreed before sampling. Final inspection can verify it only if XRF equipment is available and calibrated. For promotional pins and keychains, gold flash of 0.03-0.08 microns is common and economical. A heavier 0.10-0.30 micron precious-metal layer improves handling resistance but increases cost. Nickel, copper, or brass underlayers are typically several microns, often 3-8 microns combined, depending on substrate and finish. If the order requires nickel-free, outdoor use, salt-spray resistance, or children’s product compliance, visual inspection alone is not sufficient; specify XRF, lab testing, or a defined salt-spray standard before production.

Checkpoint	Tool or method	Practical tolerance	Fail example
Pin width or height	Digital caliper, 0.01 mm resolution	±0.2 mm under 35 mm; ±0.3 mm at 35-60 mm	35 mm pin measures 34.4 mm across sampled pieces
Coin thickness	Digital caliper at 3 points	±0.2 mm for most 2D or 3D coins	3.0 mm coin averages 2.6 mm
Unit weight	0.01 g scale	±5% for pins and keychains; ±3% for coins above 30 g	Keychain is 12% lighter than approved sample
Plating shade	Compare with golden sample under 600-800 lux	Within approved sample range	Rose gold shipment split into yellow and pink batches
Plating thickness	XRF test when specified	Per PO, often 0.03-0.30 microns for precious flash	Gold flash below agreed minimum on wear surface

Inspect Enamel, Epoxy and Printing Separately

Enamel, epoxy, and printed surfaces fail in different ways, so they need different checks. Enamel should be inspected with side lighting as well as overhead lighting, because flat light hides low fill, bubbles, and contamination near metal lines. For soft enamel, recessed color below raised metal is part of the design, but reject exposed base metal inside color areas, mixed colors, or shrinkage gaps above 0.2 mm along a logo edge. For hard enamel, the surface should be polished close to flush with the metal, with no obvious pits, drag marks, or bleed across raised lines.

Epoxy dome inspection focuses on coverage, clarity, edge control, and contamination. A good dome has even flow, clean edges, no trapped dust on the logo face, and no yellow cast unless a tinted finish was approved. Reject epoxy that stops more than 0.3 mm short of a visible edge, spills onto attachment hardware, leaves fish-eye gaps, or contains bubbles larger than 0.5 mm on the front face. For retail programs, define a tighter cosmetic limit because epoxy defects are highly visible under store lighting.

Printed metal keychains, badges, and coins need registration and abrasion checks. Offset, screen, or UV printing on metal should normally hold registration within ±0.2 mm for logos under 50 mm and ±0.3 mm for larger artwork. For fast promotional orders, a 10-cycle dry rub with clean cotton cloth can catch poorly cured ink. For retail, luggage tags, or keychains expected to face repeated handling, specify 50-100 rub cycles and confirm the added curing time in the schedule.

Test Hardware Function, Not Just Appearance

Hardware failures are the defects end users notice first. A pin with perfect enamel still fails if it drops from a jacket. A coin keychain fails if the jump ring opens after one week. Inspectors should pull, twist, open, close, and fit attachments on a defined number of pieces, because weak solder, thin rings, soft clasps, and misaligned brooch bars often look acceptable in photos.

For butterfly clutch pins, test at least 20 pieces per SKU, or 10% of the inspection sample for larger lots. The post should be straight within about 5 degrees, the solder should fully wet the base, and the clutch should not slide off under light pull. For standard lapel pins, a 1.5-2.0 kgf post pull test is practical; for larger badges, 2.5-3.0 kgf may be appropriate if it will not deform the product or create fabric-damage risk. For brooch bars, open and close the catch three times and reject sticky catches, misaligned needles, loose rivets, or tips extending beyond the safety cover.

For keychains, check ring hardness, jump ring closure, plating cracks at bends, and clasp spring action. A 25-30 mm iron split ring with 1.8-2.2 mm wire is common for standard promotional use. Heavier coins or bottle-opener keychains may need 2.5 mm wire, a stronger swivel hook, or a welded connector. Jump ring gaps above 0.2 mm, wire that deforms easily by hand, or incomplete lobster-clasp closure should be treated as major defects.

• Pull-test pin posts at 1.5-2.0 kgf for standard pins and 2.5-3.0 kgf for larger badges where appropriate.
• Check magnetic backs for polarity, magnet seating, and sliding resistance on a clean steel plate, not only direct pull strength.
• Open lobster clasps, swivel hooks, and brooch catches 5 times on sampled units and reject sticking springs or incomplete closure.
• Verify capsules, velvet boxes, backing cards, and inserts fit without bending posts or rubbing plated faces.
• Record hardware size in the report, such as 8 mm butterfly clutch, 30 mm split ring, 35 mm brooch bar, or 2.5 mm jump ring wire.

Control Packing, Counts and Carton Marks

Many expensive shipment problems are packing failures, not manufacturing failures. Mixed designs, short cartons, wrong backing cards, weak export boxes, and missing carton marks create the same buyer-side disruption as poor enamel. Final inspection should verify counts at individual pack, inner box, and master carton level, especially for multi-SKU distributor orders.

For pins on backing cards, check orientation, card print, barcode, polybag seal, and attachment position. A common specification is one pin on a 250-300 gsm coated card, one 35-45 micron OPP bag, 50 or 100 pieces per inner box, and 500-1,000 pieces per export carton depending on weight. For challenge coins, keep master cartons below about 15 kg gross weight unless the warehouse accepts heavier cartons. Heavy coin cartons should use tighter inner packing to stop capsules and velvet boxes from rubbing during transit.

Export cartons should normally be 5-ply corrugated with clean shipping marks, SKU, quantity, gross weight, net weight, carton number, and country of origin when required. Drop testing is not necessary for every low-cost order, but for retail packaging, heavy coins, air shipment, or long LCL routes, a basic corner-edge-face drop sequence from 60-80 cm on one packed carton can reveal weak inner protection before export.

Packing item	Common spec	Inspection risk	When to upgrade
OPP bag	35-45 microns, self-seal or heat-seal	Open seals, trapped dust, wrong SKU	Retail presentation or long sea transit
Backing card	250-300 gsm coated card	Wrong artwork, bent corners, off-center pin	Hang-tab retail or barcode scanning
Inner box	50-100 pins or 20-50 coins	Mixed designs, no inner label	Multi-SKU distributor orders
Export carton	5-ply corrugated, under 15 kg preferred	Crushing, missing carton marks	Heavy coins, Amazon-style receiving, long LCL route
Desiccant	1-2 g per small inner bag or carton as needed	Moisture marks on plating or paper	Sea freight, humid season, paper gift boxes

Price, Lead Time and PO Clause

Factory internal inspection is usually included in the unit price, but it is not the same as an independent third-party final random inspection. In Zhejiang and nearby provinces, a third-party inspection commonly costs USD 180-350 per man-day, depending on location, product complexity, travel, report format, and whether functional tests or XRF checks are required. For a USD 400 reorder, that may be disproportionate. For a USD 8,000 event shipment with no replacement window, it is often cheap insurance.

Inspection affects lead time. Internal final inspection for 500-3,000 pieces usually adds 1 working day if scheduled in advance. Third-party booking normally needs 2-4 working days of notice plus 1 day for the report. If the lot fails, add 1-3 days for counting or repacking issues, 3-7 days for hardware replacement, and 7-15 days or more for replating, recasting, or remaking defective metal parts.

Cost should be proportional to order value and risk. Typical FOB ranges vary by size, finish, and packing, but 300-piece soft enamel pin orders often fall around USD 0.65-1.20 per piece, 5,000-piece metal keychains around USD 1.10-2.80, and 2,000-piece challenge coins around USD 2.50-6.50. MOQ is commonly 100-300 pieces for simple pins, 300-500 pieces for custom keychains, and 100-300 pieces for challenge coins, with better pricing usually appearing at 500, 1,000, and 5,000-piece tiers. Keep strict safety, branding, function, and count checks for every order; increase cosmetic, packaging, and third-party inspection depth as client risk rises.

Write the inspection clause into the purchase order before sampling or mold opening. The clause should define the inspected lot, AQL levels, golden sample, defect classes, measurement tolerances, packing requirements, and the buyer’s right to reject, sort, rework, or re-inspect failed goods. If a third-party inspector is used, send the same clause to the factory and inspector, along with approved artwork, sample photos, packing layout, and critical dimensions.

• Before sampling: approve drawing, dimensions, plating, attachment, packing, and defect classification.
• Before mass production: confirm golden sample, AQL level, carton plan, MOQ tier, and inspection timing.
• Before balance payment: inspect finished goods against the PO, not informal chat messages or memory.
• If the lot fails: require written sorting results, re-inspection of repaired goods, and photos of rejected pieces separated from good stock.
• For reorders: reuse the same inspection clause and compare new production against the stored golden sample and previous shipment report.

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