A Chinese Court Enforced This Non-Circumvention Clause. It Was Worth a Third of the Number.

That's my translation of two clauses from a 2020 agreement between a trading company and a factory in Zhongshan, Guangdong. By the standard of most NNNs I review, it's rough. It has no end date, no list of customers, and a damages figure with nothing behind it. A Chinese court enforced it anyway, and the appeal court agreed. Then it awarded a third of the figure.

If your factory has gone around you to your customer, the case is worth reading closely for three things: the argument that the customer made the first move got nowhere, the missing end date nearly mattered, and the damages came down for reasons you can plan around. It's one basic-level judgment upheld on appeal, and a court in another city could see it differently. And the factory never raised the antitrust argument that has since sunk a clause of much the same shape in the Supreme People's Court. I come back to that below.

How the deal came apart

The trading company, Zhongshan Yintai Technology, had customers and no production. The factory, then called Zhongshan Zhengfeng Electronic Technology, had production. The arrangement was simple. Yintai won orders, passed them to the factory, sold the goods on in its own name and kept the difference.

1. 20 November 2020. The two sign the agreement containing the clauses above.
2. 9–29 December 2021. A customer in Dongguan places orders through a WeChat group called "Jiashi Yintai Cooperation." The factory's staff are members of the group. The factory makes the goods and delivers them to Yintai, which delivers them to the customer under its own name. The factory bills Yintai RMB 76,639.80. The customer later pays Yintai RMB 107,608.27.
3. 29 December 2021. Yintai's last order to the factory.
4. 3 January 2022. The customer posts an order in the same WeChat group, addressed this time to the factory. From then on it buys from the factory direct. Yintai stops sending the factory work.
5. March 2023. Yintai sues for the RMB 300,000 and for RMB 30,968.47 of lost margin.
6. 6 June 2023. The Zhongshan No. 2 People's Court awards RMB 100,000 and nothing else.
7. 1 September 2023. The Zhongshan Intermediate People's Court dismisses Yintai's appeal.

One detail deserves more attention than it got in court. Yintai put its customer and its factory in the same group chat. The customer later gave the factory a signed statement explaining what happened next: it had found out Yintai had no production of its own and that the factory was making its goods, so from January 2022 it moved the business to the factory to cut costs and secure supply. That's a commercial problem no clause solves. The clause decides what the breach costs the factory. It can't stop a customer who can see the factory from wanting to buy there.

Was the clause valid?

Yes, with a time limit the court supplied.

The factory's main attack was that the post-termination ban had no end. Read literally, it could never deal with any customer Yintai had ever served, however many years passed, and that was so unfair the clause should be void. It added that it had never approached the customer, and that a contract between it and Yintai couldn't bind the customer's own choices.

The court dealt with the validity point in three steps. The clause wasn't a standard term that Yintai had imposed; it was negotiated, clear and what both sides meant. Its purpose was to stop the factory taking the orders Yintai brought it and then using that opening to trade around Yintai, which the court called an expression of the good-faith principle. The post-termination ban did add to the factory's burden and limit its freedom. But given why the parties signed and what they were trying to do, the court held that the ban was valid "within a reasonable period."

So the court kept the clause alive by reading in a limit the parties never wrote. That worked for Yintai because the breach came within days. It would have been a much harder argument if the factory had waited eighteen months, and the court never said where a reasonable period ends because it didn't need to. A clause with no end date hands that question to whichever judge hears the case.

Did it matter that the customer made the first move?

No, and this is the part I'd show a factory that says "they came to us."

The court held that it made no difference whether the customer contacted the factory or the factory contacted the customer. Dealing with the customer directly was a breach either way. Even if the cooperation had effectively ended once Yintai stopped placing orders, the factory still owed the post-termination duty for a reasonable period, and it had started trading with Yintai's customer only days after Yintai's last order. The court said that went against good faith.

Two phrases in the clause carried that result: "for any reason" and "through its company or any individual." A lot of English-language NNNs I see say only that the supplier "shall not solicit" the buyer's customers. If that's all yours says, the factory's defence is already written for it. It didn't solicit anyone; the customer called. Write the clause about dealing with the customer, and say in terms that it doesn't matter who made contact first.

How much was it worth?

RMB 100,000 of the RMB 330,968.47 claimed.

The factory argued that Yintai's real profit from the whole cooperation was RMB 13,484.90, and asked the court to cut the damages to 30% of that figure. Yintai's own figure for the lost margin was RMB 30,968.47, the gap between what the customer paid and what the factory charged over that one month in December 2021.

The court landed between the two positions but much closer to Yintai's. It set the damages at RMB 100,000, citing Yintai's earnings, the factory's degree of fault, fairness, and the fact that liquidated damages in Chinese law are partly meant to punish. It refused the separate claim for lost margin, for two reasons: one month of trading was no basis for calculating lost profit, and the RMB 100,000 already compensated Yintai. On appeal, Yintai argued that it had paid for all the work of finding the customer, and that the factory had taken a long-term customer from it. The appeal court held the RMB 100,000 wasn't obviously wrong and left it alone.

Put the award next to the numbers. It's about three times the one month of margin Yintai could prove, and about seven times the profit figure the factory put forward. That's a meaningful result for a trader whose clause was badly drafted. It's also well short of the figure in the contract, and the evidence explains most of the gap. The value of a long-term customer is the orders it would have placed next year and the year after. Yintai proved one month.

The rules have tightened since this case. In December 2023 the Supreme People's Court issued an interpretation on contract law that tells courts to start from the actual loss when a party asks for liquidated damages to be reduced, treats a figure more than 30% above the loss as generally excessive, and says that a party in malicious breach generally can't get a reduction at all. The Zhongshan judgments came before that interpretation. I can't tell you whether a court today would call this factory's conduct malicious. But the rule gives you two reasons to keep records: evidence of the loss sets the base, and evidence of what the factory knew goes to whether it can ask for a cut. Here the factory's own staff sat in the group chat where the customer placed Yintai's orders.

What this case doesn't tell you

It's one ordinary judgment from a basic-level court in Guangdong, upheld by the intermediate court above it. Chinese judgments don't bind later courts the way precedent does in common-law systems, and a court in another city could weigh the same facts differently.

Nor did the factory argue antitrust. In December 2024 the Supreme People's Court struck down a clause stopping a producer from selling to its trader's customers, on the basis that the two were competitors at the point of sale and the clause divided customers between them. I've written about that decision and what it means for middlemen here. A factory's lawyer today would probably run both arguments. The Zhongshan clause, which covered every customer Yintai had ever served with no end date, is broad in exactly the ways that argument goes after. What Zhongshan shows is a court enforcing this kind of promise when the only objection was fairness. Whether the same clause would survive an antitrust defence is a separate question, and on the current authority I'd expect it to struggle.

Both companies were Chinese, the contract was Chinese and the court was Chinese. For a foreign buyer, none of this helps unless your NNN is written to be enforced in a Chinese court against the Chinese company that actually signed it.

And the clause bound the factory. It never bound the customer, who was free to buy wherever it liked. Your claim is against the factory, and it's a claim for money.

The clause, rewritten

Here's what I'd change, keeping what worked.

Who is protected. The original: every customer Yintai had ever worked with. What I'd write: customers named in a schedule, or identified in orders, project files or a written notice the factory has actually received, with a promise to confirm on request whether a given company is covered. A factory can't sensibly be held to a promise about customers it had no way of identifying, and a defined list is also far easier to defend against the antitrust argument.

How long. The original: no end date. What I'd write: a fixed period for each protected customer, counted from your last order involving that customer, which a new purchase order doesn't restart. You're no longer relying on a judge to decide what's reasonable.

What counts as a breach. The original: dealing privately, for any reason, through the company or any individual. What I'd write: keep all of that, extend it to the factory's related companies, say expressly that it doesn't matter who made contact first, and carve out anything you instructed, such as a delivery or a site visit you asked for.

What it costs. The original: RMB 300,000, flat. What I'd write: a figure with a stated basis, such as a multiple of the gross margin on that customer's orders over the previous twelve months, with a floor. Then keep the order history, the invoices and the margin records that prove the base, because that's what the court will ask for.

One more thing, and it isn't drafting. If your business depends on the factory never meeting your customer, don't put them in the same chat group.

If your factory has already started dealing with your customer, send me the agreement, the order history with that customer and the chat records, and I'll tell you what the clause is likely to be worth before you spend money chasing it. 48-hour incident support →