Fjord's production director sent me a spreadsheet last year. Eighteen months of inspection results across 6 orders. The average defect rate: 11.4%.
For context, a well-managed factory runs at 2–4%. At 11.4%, Fjord was losing roughly $34,000 per container — $18,000 in rejected and reworked units, $9,000 in air-freighting replacements to retail partners, and $7,000 in brand damage they couldn't quantify but could feel in their reorder rates.
Their first instinct was the same as almost every buyer's: find a new factory. They had quotes from three competitors, all promising "better quality," all within 5% of their current supplier's price.
I asked them to try one thing before switching. Give me two days on their current factory's production floor, and let me tell you afterward whether the problem is the factory or the process.
This is what I found.
The 11.4% Autopsy
Two days, every station, every jig, every fixture. I interviewed the QC team, the line supervisors, and the operators. Here's what 11.4% actually broke down into:
| Defect Type | Before (% of units) | Root Cause Identified | After (% of units) | Reduction |
| Shell surface scratches / scuffs | 4.2% | Inadequate protective film during assembly. Workers stacking shells directly without interleaving. | 0.3% | −93% |
| Wheel alignment / wobble | 2.8% | Drilling jig tolerance was ±0.5mm. Wheel housing required ±0.2mm. Jig had worn over 18 months without recalibration. | 0.2% | −93% |
| Handle extension mechanism stiffness | 1.9% | Handle tube inner diameter variance from tube supplier. QC was checking OD only, not ID. | 0.4% | −79% |
| Zipper misalignment on expansion gusset | 1.1% | Sewing operator was new (3 weeks on the job). Training checklist didn't include expansion gusset as a separate station qualification. | 0.2% | −82% |
| Logo application inconsistency (heat transfer) | 0.7% | Heat press temperature was fluctuating ±8°C because the thermostat hadn't been calibrated in 12 months. | 0.1% | −86% |
| Other minor (stitching density, lining wrinkle, etc.) | 0.7% | Mixed — general process discipline | 0.3% | −57% |
| TOTAL | 11.4% | 1.5% | −87% |
Here's the finding that changed everything: zero of the six defect categories were caused by "bad factory." Not one.
Worn tooling, an incomplete training checklist, uncalibrated equipment, an incoming QC gap — every single category was a process failure. The factory was perfectly capable of making good luggage. It just wasn't being managed to do it consistently.
And the uncomfortable truth: if Fjord had switched factories without this analysis, they'd have carried the same process gaps to the new factory. Six months later they'd be back at 10–12% defect rates — just with a different factory name on the QC report.
The 5 Fixes That Cost $4,700 Total
Fix 1: Check ID, not just OD — $180
The handle tube problem cost 1.9% of units. The supplier's inner diameter was drifting, and nobody was catching it because the incoming QC spec only listed the outer diameter. We added an ID tolerance of ±0.1mm to the checklist and put a go/no-go gauge at the receiving dock — a steel pin at the minimum acceptable ID. If the pin doesn't slide through, the batch is rejected before it reaches the line. Cost: $180 for the gauge. Impact: handle defects from 1.9% to 0.4%.
Fix 2: A calibration calendar — $500
The wheel-housing drilling jig had worn from ±0.2mm to ±0.5mm tolerance over 18 months. Nobody noticed because nobody was measuring. The heat press thermostat was swinging ±8°C because its last calibration was 12 months ago. We built a calendar: drilling jigs every 6 months, heat presses every 3 months, injection mold cavity dimensions every 12 months or 50,000 shots, whichever comes first. Every calibration gets a dated sticker and a log entry. Cost: $300 recalibration plus $200 in tools. Impact: wheel defects 2.8% → 0.2%, logo defects 0.7% → 0.1%.
Fix 3: Protective film within 5 minutes — $0.12/unit
Shell scratches were 4.2% of all defects — the single largest category. The shells left injection molding pristine and got scratched during assembly because workers stacked them directly on carts with nothing between them. The fix: PE protective film applied within 5 minutes of demolding, left on through assembly, inspection, and packing, removed by the end customer. Material cost: $0.12 per unit — $600 per 5,000-unit order. Impact: shell defects 4.2% → 0.3%.
Fix 4: Station-specific certification — $1,200
The zipper misalignment traced back to one new operator, three weeks on the job, who'd passed general sewing training but had never been qualified for the expansion gusset station — a station with different tension control than standard zipper work. We split the training checklist into station-specific modules. General sewing training qualifies you for 80% of stations; the remaining 20% — expansion gussets, curved seams, reinforced corners — require separate supervisor sign-off before you touch a production unit. Cost: $1,200 to redesign the checklist and run certification sessions. Impact: zipper defects 1.1% → 0.2%.
Fix 5: The weekly process audit — $3,500/year
The most impactful change, and the cheapest per unit. Once a week, for 30 minutes, one QC staff member walks the line with a checklist — not looking at finished products, at the process itself. Are jigs within calibration? Is the film being applied? Are operators certified for their stations? Is incoming QC checking ID as well as OD? QC inspects products. A process audit inspects the system that produces the products. You can catch a scratched shell at QC; you prevent the next 500 scratches with a process audit. Cost: $3,500 per year for 30 minutes a week of QC time. Impact: systemic — it stops regression across every defect category at once.
18 Months Later
Fjord has placed 6 orders since the fixes went in. Average defect rate across those orders: 1.5%, down from 11.4%. Total spend: roughly $4,700 in one-time costs plus $0.12 per unit ongoing for the film.
Now price the alternative they nearly took. Switching factories would have meant mold transfer and requalification at $8,000–12,000, new-factory sampling and testing at $3,000–5,000, a first-order defect risk premium — new factory, unproven process, call it ~15% on order one, which is $25,000+ in rework and replacements — and 6–9 months of relationship rebuilding before reaching a Phase 3 partnership. Total avoided: approximately $36,000–42,000 in direct costs, plus 9–12 months of productive partnership time.
The math isn't close. The scary part is how many buyers never run it.
Switch or Fix? The Framework
Not every factory deserves to be kept. Here's how to decide:
| Signal | Verdict | Why |
| Quality problems are inconsistent — good batches and bad batches from the same factory | FIX | This is a process control problem. The factory can produce quality; the issue is consistency. Fixable with process audits and calibration schedules. |
| Quality problems are consistent — every batch is bad in the same way | FIX or SWITCH | Consistent defects usually mean a capability gap (wrong equipment, wrong skill level) or a willful shortcut (cost-cutting on materials). Capability is fixable with investment. Willful cutting means leave — you can't fix a factory that doesn't want to be fixed. |
| The factory lies about problems or hides them until delivery | SWITCH | Transparency is the foundation of any partnership. If the factory hides problems, you can't fix processes because you don't know which ones are broken. Unfixable. |
| The factory is unwilling to let you audit their process | SWITCH | A factory that won't let you see the production floor is hiding something. Every reputable factory allows buyer visits. Every one. |
| The factory is losing money on your orders and can't invest in improvements | FIX (reprice) or SWITCH | If your price sits below the factory's sustainable cost, quality degrades no matter what. Reprice to a sustainable level, or find a factory whose cost structure fits your volume. |
Reader Question
"How do I bring up quality problems without making my factory defensive?"
It's about when and how, more than what. Don't open with whose fault it is — open with what happened. "What caused the color variance on this batch?" starts a process discussion. "This is unacceptable" starts an argument. Same issue, two conversations, two completely different outcomes. And get it in writing early: the factory that answers questions about its own process is a factory that's learning with you.
That's Case Study File #1. Next up: a US startup that grew from 500 units to 15,000 units in 3 years with the same factory — the scaling playbook, the near-death moments, and the actual numbers behind both.
About the Author
Written by the CLK Manufacturing Team — 16+ years of custom luggage OEM/ODM experience from our factory in Ji'an, Jiangxi, China. Updated: 2026-08-15.
Related Products & Sourcing
CLK Luggage (Jiangxi Chengleke Leather Co., Ltd.) manufactures custom hard-shell luggage, aluminum-magnesium cases and luggage sets for brands, retailers and corporate clients. Relevant resources:
- Browse our luggage sets and custom OEM/ODM luggage manufacturing
- Why choose CLK — factory profile, quality control and certifications
- Get a quote: contact our team or email clkbusiness@clkluggage.com