Defect notifications never arrive at a convenient time. Friday afternoon, five o’clock. An email from your German customer’s SQE. Three photos attached — flash on the same location across the same batch. Subject line: “containment required.” Your first thought is almost always the same: bring it back.
Bring it back to the factory. Your own people, your own floor — sounds like the obvious cost-saving move. Every die-casting plant owner in China has had this reaction to their first complaint notification. It feels natural.
But sorting in a cross-border supply chain is not what you picture. Walk the full reverse-logistics cycle from start to finish, add up every cost — freight, duty, time, customer patience, escalation risk — and the total is almost always higher than sorting on-site in Europe. Not a little higher. Several times higher. The problem is that several of the biggest costs never appear on anyone’s quote.
Plan A: recall to China — walk through every cost
“Recall and sort” sounds like one step. It’s actually a long reverse-logistics chain.
Step one: local collection in Europe. Your customer’s warehouse isn’t next door to your factory gate. It might be outside Stuttgart, in Wolfsburg, or at an industrial park in the Czech Republic. The first move is getting the rejected lot out of the customer’s facility and to the nearest export port. If the defect is safety-related — say, a brake-system bracket — the customer won’t let you use standard freight. They’ll demand expedited pickup, and the price tag goes up accordingly.
Step two: ocean or air freight back to China. Ocean is cheaper, but Hamburg to Shanghai takes three to four weeks on the water. Add trucking and customs at both ends, and you’re looking at five weeks minimum in transit. Air freight is fast — days, not weeks — but die-cast parts are heavy. The air-freight bill for five to ten thousand gearbox housings often exceeds the total profit margin on the entire order. And air freight still needs import clearance and domestic trucking on the Chinese side.
Step three: import customs clearance. This is where many first-timers trip. These parts were originally exported with VAT rebate already claimed. Bringing them back counts as “returned goods,” which means a fresh customs declaration and repayment of the duty and VAT that were refunded on export. Skip this step? Then customs treats the shipment as a new import, and the duty is even heavier. Do it properly? Your finance team spends days processing return-of-goods paperwork through the tax system. Either way, you’re bleeding time and money.
Step four: in-house sorting. The parts finally land in your factory yard. Your troubles are just starting. You pull inspectors off the production line to do the sorting — which means normal production slows down. If the defect is functional — say, a porosity leak — sorting isn’t visual. You need test fixtures. Maybe X-ray. Do you have spare capacity for that right now? Most plants don’t. While your inspectors are sorting, regular orders are slipping. And your customer is already watching you because of the quality problem on the last shipment.
Step five: repack, re-export, re-clear. Sorting is done. Good parts need repacking — the original packaging is likely destroyed from being opened and examined. Then a fresh round: export declaration, ocean freight, European import clearance, last-mile delivery. From the moment the customer notified you to the moment they receive the sorted good parts: normally six to eight weeks.
Two months. Your customer cannot wait two months. Their production line runs on a daily schedule.
Plan B: on-site sorting in Germany — what looks expensive isn’t
The on-site path is much shorter.
Your service partner arrives at the customer facility or nearby warehouse within 24 to 48 hours of notification. They set up a sorting station. They sit down with the customer’s quality engineer and align on the inspection standard face to face. No freight leaves the customer’s site — the sorting happens right there, in the customer’s plant or in a temporary bay next door.
Every part is inspected one by one against the customer’s specification. OK / NG / rework. Data is compiled daily and sent to you and the customer the same day. Good parts are released straight into the customer’s inventory. Production doesn’t stop.
The visible cost of on-site sorting is mainly people: an engineer who understands die-casting defect classification and can communicate in both Chinese and the customer’s language, billed by the day. That daily rate looks high compared to what you pay an inspector on your factory floor. But you’re not paying for his hours. You’re paying to eliminate six to eight weeks of logistics time, round-trip freight, duty, repackaging, re-clearance, production-line stoppage risk, and the three reminder emails per week your customer’s quality manager would otherwise be sending you. All of that, gone in a single decision.
On-site sorting has another benefit that’s easy to overlook: visibility. Your customer sees someone there. They see work happening. They see data arriving every day. They see you are serious about solving the problem. This isn’t just containment — it’s the only thing that rebuilds trust at this stage.
The costs you forgot to count
Plan A’s real weakness isn’t the line items you can put on a spreadsheet. It’s the ones you can’t.
The customer’s production line. If your part is the sole source for that customer, they cannot wait two months. They will demand you air-freight the sorted parts, or demand on-site sorting — and if you push back, they’ll make a note. That note goes on the supplier scorecard under “responsiveness.” Not a line item on anyone’s invoice, but harder to recover from than any line item you’ll ever see.
CS escalation. While recalled parts are still in transit, if even one more defect escapes — from the same lot not fully cleaned, or from the next shipment with a new problem — the customer has grounds to escalate from CS1 to CS2. CS2 costs roughly three times what CS1 does. Once you’re in CS2, your chances of winning a new project from that customer in the next twelve months are near zero. Nobody in any factory has this number in a spreadsheet, but anyone who’s been through it knows the magnitude.
The purchasing manager’s patience. European Tier-1 procurement organisations run annual supplier rationalisation. Your name goes onto the “risk supplier” list after one escalation. Getting off that list takes two consecutive years of zero PPM.
Your engineers. The people sorting in your factory are also the people running new-product trials and dialling in process parameters for the next order. When they spend four hours a morning on the sorting line, your new tool debugging loses a person that afternoon. This isn’t a sorting cost. It’s opportunity cost.
A real worked example
Take a batch of 10,000 aluminium die-cast brackets, roughly 5% defect rate. Export unit price €6. Part weight 0.8 kg each.
Recall to China: inland collection in Europe ~€800, Hamburg–Shanghai ocean freight ~€1,200, Chinese import clearance and duty clawback ~€1,500, domestic trucking ~€300, in-house sorting labour ~€2,000 (two people, one week), repackaging ~€500, re-export freight ~€1,200, European clearance and last-mile delivery ~€600. Direct costs total approximately €8,100. Timeline: six to eight weeks. This excludes any production stoppage cost at the customer, CS escalation risk, and the opportunity cost of your tied-up engineers.
On-site sorting in Germany: engineer travel ~€400, sorting-station setup ~€200, two engineers for four to five working days ~€5,500, daily data reporting and communication included in the service fee. Direct costs total approximately €6,100. Timeline from notification to release: within one week.
The numbers already tell the story. But the real kicker is what the on-site option prevents: no production stoppage, no escalation trigger, and your engineers stay in your factory doing their actual jobs.
When recalling to China actually makes sense
Not every situation calls for on-site sorting. Here’s a simple decision framework.
Recall is fine when: the lot is small — a few hundred parts — the defect is purely cosmetic and doesn’t affect assembly, the customer has no line-down pressure, the part is non-automotive or non-safety, and your factory genuinely has spare capacity. In these cases, bringing it back does save money — because you have no time pressure, no escalation risk, and the freight bill on a few hundred parts is manageable.
On-site is non-negotiable when: the part is the customer’s sole supply source, the defect affects fit or function, the lot is several thousand pieces or more, the customer is in automotive, or you are already in CS1. If any two of those conditions are true, recalling to China almost certainly costs you more.
The simplest test: ask yourself whether, if you don’t finish sorting this week, the customer will call you again. If the answer is yes, sort on-site.
One more thing worth remembering: the customer’s containment notification didn’t come out of nowhere. Something already went wrong in your process. While your service partner handles the containment in Europe, your own engineers should be working the root cause at home. That parallel track — containment on one continent, corrective action on the other — is the fastest path out of controlled shipping. And that’s only possible when you don’t have your entire quality team tied up counting and sorting parts.