Failed Inspection in China: What to Do Before You Pay

A failed inspection is a decision point, not a verdict, and the unpaid balance is your strongest leverage while you make it. Get the defect data first, work out whether the fault is fixable or fatal, then decide. Once the money moves, your options move with it.

The fastest way to lose money here is to react before you have the facts, or to pay before you have a fix. What follows is the order to do it in.

The First Hour: Get Data, Not an Argument

Ask for the full inspection file, not the one-line result. You want defect counts, the sample size, which cartons they came from, timestamps, and photos that show both the fault and how widespread it is. A report that says “failed” without that detail is not something you can act on, or argue from.

Then check the findings against your own standard, not the factory’s opinion. Every fault should map to a defect class you agreed before production, and the pass or fail should follow the AQL standard in your purchase order. A misclassified fault is what turns a fixable batch into a fight.

Say one thing to the supplier the same day: the balance is on hold pending corrective action. You do not need a plan yet. You need a record, and you need the factory to know the money has stopped moving before they start proposing shortcuts.

Can the Batch Be Fixed?

Sort every fault into two piles before you talk about remedies. Wrong labels, missing inserts, loose screws, dirty surfaces, and repackable cartons are reworkable. Weak structure, unsafe materials, a function that does not work, or a dimension that is wrong across the batch usually are not.

Ask the factory to separate the two piles itself, and to explain the cause. A supplier who can tell you which faults are correctable, how the affected units will be identified, and what caused the problem is worth working with. One who answers “we will fix everything” has not looked.

Watch for the fault that reveals the process, not the piece. A defect scattered evenly across every carton is a control problem, and that batch is rarely the last one. A fault in one carton range is an accident; a fault everywhere is how the factory works.

Your Three Options, and What Each Really Costs

Rework is the default when the fault is correctable and you still have time. The factory sorts, fixes, and repacks, then the lot gets checked again. The cost is time, a second visit, and the risk that rework creates new faults such as damaged packaging or mixed cartons. Settle the re-inspection cost and who pays it in the purchase order, long before a batch ever fails.

A discount suits real but survivable faults in the right channel. It is the fastest route and it keeps the schedule. It is also the one buyers regret most, because a price cut on goods that later come back as returns means you paid for the problem twice. Never take a discount on anything a customer would notice.

Rejection is for severe, hidden, or repeated faults, and it is not the disaster it feels like. Yes, you lose the schedule. But shipping goods you cannot sell costs the freight, the duty, the storage, the returns, and the marketplace ranking, and those are the hidden costs of importing that quietly outgrow the order itself.

The Balance Is Your Strongest Leverage

Everything you can do about a bad batch, you can do before the balance is paid, and almost nothing after. That is the whole reason a pre-shipment inspection happens when it does. Once the factory has your money, your remedy is a request, not a condition.

So do not pay against a promise. “Ship now, we will fix it later” and “we will make it up on the next order” are the two sentences that cost buyers the most, and both arrive exactly when your deadline is tightest.

Put the remedy in writing before anything moves. What gets fixed, how, by when, who pays for the rework and the re-inspection, and that the goods do not ship until they pass. A written plan turns a stalled order into a schedule you can manage.

The Four Objections You Will Hear

“The inspector was too strict.” Point back to the standard the factory accepted before production. If they agreed to it then, the issue is compliance, not attitude.

“Most of the units are fine.” Probably true, and irrelevant. A lot fails on the agreed limit, not on the share of good pieces, and the sample exists precisely because nobody checks all of them.

“But the approved sample had the same issue.” Check whether the fault was actually visible in that sample and formally accepted. One defect nobody noticed on one piece does not rewrite the spec for the whole order.

“This defect rate is normal for the industry.” Normal is not the same as agreed. The only limit that counts is the one written into your purchase order before production started, and the factory signed it.

Stopping the Next One

One failed batch is a problem. The same fault twice is a reason to reassess the supplier, and a third is usually the point to move the order. Ask whether the cause sat in the materials, the assembly, the packing, or the factory’s own checks, and whether the answer changes each time you ask.

Move the check earlier so the next fail is cheaper. Catching a drift mid-run with an in-process inspection costs the same day rate but arrives while the fix is still small, instead of when the container is booked.

And if the fails keep coming, the problem is not the inspection. A factory that cannot hold your standard will not be fixed by checking it harder, and a supplier quality audit will usually tell you that faster than a third failed report.

FAQ

Q1: Can I bring in a second inspection company to verify the fail?

Yes, when the first report is unclear, the classification is disputed, or the result sits right on the rejection limit. Use the second firm for an independent read, not to shop for a pass.

Q2: Should the factory sort the whole lot or just the failed cartons?

Usually the whole lot. A narrower sort is fine only when the factory can prove the fault is trapped in a traceable batch or carton range, and proof means production records, not a hunch.

Q3: How do I know the factory actually fixed it?

Photos of repaired units prove nothing on their own. Ask for the corrected quantity, how the affected units were identified, and a re-inspection that confirms it before anything ships.

Q4: My goods are urgent. Can I ship the good units now?

Sometimes, but only if the passing units can be identified reliably and separated under supervision. Write down which faults you accepted and which stay grounds for non-payment, or the concession quietly becomes the new standard.

Q5: Should the re-check look only at the failed defects?

No. It should confirm the fixes and repeat the checks that passed the first time, because rework is where new faults are born, usually in the packing, the assembly, and mixed cartons.

Q6: The factory says my deadline caused the defects. Is that fair?

Rarely. A tight schedule does not transfer quality responsibility unless you approved reduced checks or changed the spec late. If you did, that is a lesson for your side, not a reason to accept the goods.

Q7: I already booked the container. What do I do with it?

Tell the forwarder the day the lot fails, not the day it was meant to sail. Ask what the rollover deadline is, what it costs, and whether you keep the rate, because all three answers get worse the closer you drift to the sailing date.

Q8: What if the factory refuses a re-inspection?

Get the refusal in writing, then check what your purchase order says about access before shipment. A factory that blocks verification after a failed result has already answered the question you were trying to inspect your way to.

Conclusion

A failed inspection only becomes expensive when you rush it: pay too early, argue too fast, or accept a promise instead of a plan. Get the data, split the fixable from the fatal, and decide with the balance still in your account.

The best version of this is the one you never have to run, because someone was watching the line while the fix was still cheap. That is what manufacturing control is for: catching the fault during production rather than deciding what to do about it at the end.