How do I find a manufacturer in China?

The question people mean when they ask this is usually not "how do I find candidates" — candidates are abundant and a search engine will produce a hundred in an afternoon. What they mean is "how do I find a good one", and those are two different problems solved at two different stages. It is worth being clear about which stage you are at, because the tools are not interchangeable.

The four routes, and what each one selects for

Route	What it is good at	What it selects for
B2B marketplaces (Alibaba, Made-in-China, Global Sources)	Breadth, fast first contact, price discovery	Companies that invest in marketplace marketing — which correlates with trading companies, not with production capability
Trade fairs (Canton Fair, sector fairs)	Seeing product in hand, meeting the people, covering a category in days	Exporters with the budget and English capability to staff a booth
Industry associations and directories	Genuine manufacturers, including ones that do no marketing at all	Members — which usually means established firms, and usually means no English
Cluster towns	Depth in one product category, and the sub-tier suppliers behind it	Geography, not capability — the cluster contains the best and the worst in the country

None of these is the right answer in the abstract. A first-time buyer of a simple product is well served by a marketplace. A buyer who needs a specific process, a certification, or a factory that will not resell their design is usually not, because the thing they need is invisible in a listing.

Why the marketplace listing tells you so little

A marketplace profile describes a company as that company chose to describe itself, with badges that mean less than buyers assume. The most consequential ambiguity is the one the listing does not resolve at all: whether the company on the other end operates a factory, or buys from one. Both are legitimate businesses and a good trading company earns its margin. But they fail differently, they price differently, and a quality problem reaches a very different set of hands.

What actually narrows a shortlist

1. The registry record: does the entity exist, how long has it existed, and does its registered business scope include manufacturing the thing you are buying?
2. The gap between what the company claims and what its own paperwork shows — a business licence, an export record and a certificate all name an entity, and those names should match.
3. Whether they will answer a specific technical question about your product without deflecting to price.
4. What they say when you ask which processes are done in-house and which are subcontracted. A candid answer here is worth more than a confident one.

The failure mode to plan for

The common outcome is not fraud. It is a competent trading company that quoted on a specification it did not fully understand, placed the order with a factory you never saw, and cannot fix the problem when the goods are wrong because it does not control the line. That is a sourcing decision made without information, not bad luck, and it is avoidable at the shortlist stage for a fraction of what the shipment costs.