Your Factory Sent Defective Products. What to Do Next

Do not pay the balance.

That is the first move, and for most founders it is the only real leverage they will ever have. Everything else follows from whether the money has left your account.

A bad production run splits into two very different situations. Goods still sitting on the factory floor with a balance owing is a negotiation. Goods already landed in your warehouse with the factory paid in full is a recovery job, and a much harder one. Which one you are in decides everything below.

What should you do first when your factory sends defective products?

Three things, in this order, before you send a single message.

1. Freeze the payment. If the balance is unpaid, it stays unpaid. If it is scheduled, cancel it. Do not pay a partial amount as a gesture of good faith. It reads as acceptance.
2. Quantify the problem. "The quality is bad" gets you nowhere. "142 of the 800 units inspected have loose stitching at the side seam, which is 17.75 percent" gets you a response.
3. Say nothing final. Do not tell the factory you are rejecting the order. Do not tell them you will take it anyway. Both close doors. Ask what happened and what they propose.

The instinct is to fire off an angry email at 11pm. Resist it. Once you have said "this is unacceptable, I want a full refund", the factory's next move is to go quiet and wait for you to soften. You want them working, not waiting.

How much leverage do you actually have?

Almost all of it sits in the unpaid balance.

Standard factory payment terms are 30 percent deposit and 70 percent before shipment. That 70 percent is your entire negotiating position. A factory holding your finished stock has a cost problem. The materials and the labour are already spent, and the order is taking up warehouse space and earning nothing. They want it gone almost as much as you want it right.

Once you pay the balance, that flips. The stock is yours, the money is theirs, and your leverage drops to whatever future orders you are willing to dangle. If you are a small buyer, that is not much.

This is the whole argument for a pre-shipment inspection before the balance goes out. An inspection runs roughly USD 250 to 400 per man-day. A bad run costs the order.

What are your real options after a failed inspection?

There are six, and only six. Every negotiation is some combination of them.

• Sort. The factory checks every unit and pulls the defective ones. You end up with fewer good units. Cheapest and fastest. Works when the defect is visible and the rate is low.
• Rework. The factory repairs the defective units. Works for stitching, trims, assembly and labelling. Does not work for a moulding fault or the wrong material.
• Remake. The factory produces the defective portion again. Slow. Adds weeks. Only realistic if you have time in hand.
• Discount and accept. You take the stock at a reduced price and sell it, or sell it as seconds. Fast, and it gets you selling again. You now own the problem.
• Partial shipment. Ship the good units now and deal with the rest after. Keeps a launch alive.
• Reject. Rare, and usually a last resort. Getting deposit money back out of a factory is slow and often unsuccessful.

Who pays matters as much as which option you pick. Sorting and rework labour sits with the factory when the defect is against your approved spec. Re-inspection is the one founders forget. If the factory reworks, the goods need inspecting again, and the factory should carry that second inspection fee. Agree it in writing before they start.

One warning on rework. A factory under time pressure will sort fast and repair roughly, because labour is cheaper than remaking. Rework frequently produces a second, different defect. Never accept reworked goods on a factory's word. Re-inspect.

What counts as evidence a factory will respond to?

Not photographs on your phone.

The evidence pack that actually moves a factory has four parts.

• A third party inspection report showing the AQL level you agreed and the pass or fail result against it. The common standard is AQL 2.5 for major defects and 4.0 for minor.
• Defect counts broken down by type, expressed as a percentage of the sample.
• Photographs of the defect next to the approved sample, in the same frame.
• Your signed approved sample or tech pack, showing the spec the goods have missed.

That last one is where most claims collapse. If you never signed off a golden sample, there is no agreed standard, and the conversation becomes your opinion against theirs. It is also why samples coming back wrong matters long before production starts.

What if the goods have already landed?

Your options narrow. You are not finished.

1. Inspect properly now. Same evidence pack. Do it before you sell a single unit.
2. Stop selling the affected stock. If the defect is safety related, stop immediately and get advice. Australian Consumer Law applies to you as the seller to your customer, whatever your factory did. Your compliance obligations do not transfer up the chain.
3. Ask for a credit, not cash. Take the claim and the evidence to the factory and ask for a credit against your next order. Factories will almost always give a credit before they give a refund. Cash leaving the business is real to them. A credit is a discount on work they were hoping to win anyway.
4. Reprice what you have left. Recalculate your landed cost per unit across the units you can actually sell, not the units you ordered. A 15 percent defect rate lifts your real cost per sellable unit by nearly 18 percent.

If you end up holding stock that is sellable but imperfect, what you do with it is a commercial decision rather than a sourcing one. Kristy covers that side in what to do with stock that isn't selling.

Is it worth suing an overseas factory?

For most orders, no. The reason is not that factories are untouchable. It is that most founders' contracts are unenforceable where it counts.

China does not enforce the majority of foreign court judgments. An English language agreement naming an Australian court is close to worthless against a Chinese factory. A contract written in Chinese, governed by Chinese law, naming a Chinese court with jurisdiction where the factory holds assets, is enforceable. Factories can tell the difference at a glance.

That is the real value of a proper contract. Not the lawsuit at the end. The way it changes how you are treated at the start.

A worked example

Illustrative numbers, but the shape is real.

You order 2,000 units at USD 6.40, so an order value of USD 12,800. You paid a 30 percent deposit of USD 3,840 and owe USD 8,960. Inspection finds a 12 percent major defect rate on the side seam, which is 240 units.

• Remake the 240 units: about four weeks, and you miss your launch window.
• Sort and rework at the factory's cost: about ten days, plus a re-inspection at USD 320.
• Accept with a discount: the factory offers 8 percent off the balance, which is USD 717.

The discount looks like the easy answer. Run the number. Those 240 units at a landed cost of roughly USD 8.10 each are USD 1,944 of dead stock. An 8 percent discount covers about a third of it. Rework plus a re-inspection you pay for yourself is still the better outcome, as long as you have the ten days.

Do that arithmetic every time. Factories offer discounts because discounts are cheap for them.

What to put in place before your next run

• A signed golden sample, held by both sides, dated and photographed.
• AQL levels named in the purchase order. 2.5 major and 4.0 minor is the usual standard.
• Payment terms that put the inspection before the balance, in writing.
• The factory covering re-inspection after any rework, agreed in advance.
• A contract in Chinese, under Chinese law, once the order value justifies it.
• Eyes at the factory during production, not only at the end.

That last point does most of the work. Almost every defect that shows up at pre-shipment was visible on day three. Vetting a factory before you pay a deposit prevents a good share of this, and being in the room during production prevents most of the rest. It is a large part of why we run trips into China each October, around Canton Fair Phases Two and Three.

If the relationship is not worth saving, moving is less painful than founders expect, provided you own your tooling. The sequence for that is in switching factories without losing your tooling.

A bad run is not usually the end of a factory relationship. It is a test of whether you have a standard they can be held to. Most founders discover they do not, at the worst possible moment.