After watching enough of them from both sides, the failures stop looking varied. The same six causes appear, roughly in the same order, in companies that had nothing else in common. None of them is exotic; all of them are visible early to someone who knows what to look at.
1. The project has no owner
China is added to someone's existing job — export manager, founder, head of purchasing — as an extra. It then receives whatever attention is left after the urgent things, which in a bad quarter is none. A project nobody is measured on drifts for a year and is then judged to have failed.
2. The budget covers the start, not the middle
Money is found for the trade fair, the translation and the first trip. Nothing is provisioned for the eighteen months between « we have contacts » and « we have revenue », which is where the whole project actually lives. The plug is then pulled at month ten, having paid for everything except the part that would have worked.
3. Nobody on the team reads Chinese
Every piece of information — a supplier's registration, a competitor's prices, what customers say about your category — arrives filtered through one person. That person becomes the company's only view of a market, and their incentives quietly become the company's strategy.
4. The first partner was the first one who said yes
Enthusiasm at a trade fair is not qualification. A distributor who signs quickly, asks for exclusivity and orders once is a common and expensive pattern: the territory is locked, nothing moves, and unlocking it takes a year of contract discussion.
5. The offer was never adapted
The European catalogue is translated and presented unchanged — same range, same packaging, same price logic, same arguments. It fails not because Chinese buyers reject foreign products, but because the arguments answer questions they were not asking.
6. The company gave up one step before it worked
This is the most painful one. The relationships that produce orders in China are usually the ones in their second or third year. Companies that leave at eighteen months leave behind the exact asset they spent eighteen months building — and the next European entrant inherits nothing, because it was relational, not documented.
What the surviving projects share is unglamorous: one named person accountable, a budget that covers the middle, and someone in the loop who reads both languages and has no incentive to filter what they report.