These three channels are usually presented as alternatives, as though you should pick the best one. They are not alternatives. Each one selects for a different kind of company, and understanding that selection tells you more than any feature comparison.
Alibaba.com: suppliers who sell to foreigners
Alibaba's international platform is built for export. Listings are in English, staff answer in English, prices are quoted in dollars, and the whole apparatus assumes an overseas buyer.
That is genuinely valuable, and it comes with a selection effect worth naming: you are seeing the companies that have invested in selling to foreigners. Some are excellent factories with a capable export department. Many are trading companies, because building an English-speaking sales operation is exactly the kind of thing a trading company does well and a small factory does not.
So Alibaba is not a directory of Chinese manufacturing. It is a directory of Chinese companies that want to reach you, which is a smaller and differently-shaped set. That is fine as long as you know it, and it is why establishing what a supplier actually is matters before you commit.
The mechanism behind that shape is worth understanding, because it tells you what searching harder can and cannot fix. Visibility on an export platform is bought. Paid membership, keyword advertising and a staffed account are ongoing annual costs, and a supplier who does not carry them is effectively unfindable no matter how good the factory is. Results are therefore ordered partly by willingness to spend on being found, which is not the same thing as capability and is occasionally the opposite of it.
1688: the domestic wholesale market
1688 is Alibaba's Chinese-language domestic platform, and it is where Chinese businesses buy from each other. Prices tend to be lower because you are seeing the domestic wholesale layer rather than the export layer, and the catalogue is far larger.
The obstacles are real and they are not just language. Sellers generally expect a Chinese-speaking counterpart, a domestic payment method, and delivery to a domestic address. Many have no export capability at all and no interest in acquiring it. There is no international dispute process standing behind your transaction, and product listings assume a buyer who already understands the category.
Used well, 1688 is the best available window onto what things actually cost before export packaging, agent margin, freight and duty. Used naively, it is a way to pay a domestic price and then discover that nobody in the chain is responsible for getting the goods to you.
Two of those obstacles are now partly addressed. 1688 operates a cross-border channel with an English interface and payment that works without a mainland Chinese bank account, covering a filtered subset of the catalogue rather than all of it. Which of the two routes you end up on changes a great deal — reading a domestic listing, how the money moves, and who makes the export declaration are different in each. That is worked through in how to buy from 1688.
The Canton Fair: the one that is about people
The Canton Fair is a large export trade fair held in Guangzhou, in phases covering different product categories. Its value is not the listings. It is that you meet people, handle products, and compare a dozen suppliers in the same category in an afternoon.
What you learn in person is difficult to get any other way. How a supplier reacts to a technical question. Whether they know their own product. Whether the sample on the table matches the photograph in the catalogue. Those signals are compressed into a few minutes of conversation and they are remarkably reliable.
Two practical points about how it runs. It is held twice a year, in spring and autumn, and it is split into phases covering different product categories — so the dates that matter are the ones for your category rather than for the fair as a whole. Attendance requires registering for a buyer badge in advance, and the official site carries the current phase breakdown.
It is also worth being clear about which products justify the trip. Anything whose specification is hard to write down benefits most: texture, finish, weight in the hand, how a fabric drapes, how a mechanism feels under a thumb. For a product that can be fully described in a drawing and a materials list, a fair adds less than a factory visit would and considerably less than a sample does.
The costs are equally real: travel, time, and the fact that exhibiting is itself a selection filter — a stand costs money, which again favours companies with an export sales function.
The channels this leaves out
Three channels is a simplification. Made-in-China.com is a substantial export directory in its own right, Global Sources has held a long-standing position in electronics and runs its own trade shows in Hong Kong, and regional and category-specific fairs run across the mainland, many of them far more focused than the Canton Fair.
They are left out above not because they are worse but because they do not change the shape of the decision. Every export-facing directory carries the same selection effect as Alibaba, for the same economic reason. Every trade fair carries the same one as the Canton Fair. Adding platforms widens the shortlist; it does not change what the shortlist still has to be checked for.
How they actually get used together
In practice these are stages rather than choices.
- Understanding the market. 1688 for a realistic sense of the domestic price floor and the range of what exists; Alibaba for how the category is presented to export buyers.
- Building a shortlist. Alibaba for suppliers you can transact with directly; 1688 if you have someone who can work in Chinese and handle export.
- Deciding. In person if the order justifies it, whether at a fair or at the factory. Nothing else compresses as much information into as little time.
| Channel | What it is good at | What it costs you |
|---|---|---|
| Alibaba.com | Transacting directly in English, with platform cover on the payment | You see only companies paying to be seen, weighted towards trading companies |
| 1688 cross-border | Domestic pricing, payable from outside China | A filtered subset of the catalogue, and still no exporter of record |
| 1688 domestic | The full catalogue and the real domestic price floor | Chinese throughout; you supply the language, the payment route and the exporter |
| Canton Fair | Judging people and handling product; a dozen suppliers in an afternoon | Travel and time, and exhibiting is itself a paid filter |
The channel that suits order one rarely suits order ten
A first order is mostly a trust problem. You want cover on the payment, a written record, and a counterparty with something to lose — which is a fair description of what an export platform provides, and roughly what the premium buys.
By the tenth order that same apparatus is largely cost. You know the company, you have a payment history with them, and the platform is taking a share of a relationship it is no longer underwriting. This is the point at which buyers move to contracting directly, or move to the domestic market with someone handling export, and the saving is real.
The mistake runs in both directions: paying platform economics for a supplier you have used for three years, or skipping platform cover on a first order with a company you found yesterday.
The price gap and what it is
The gap between a domestic 1688 price and an export quotation is not simply margin, and treating it that way leads buyers to conclusions that do not survive contact with reality. Between the two sit export packaging, domestic transport to the port, customs documentation, export declaration by an entity with the right to make it, quality control, and someone carrying the risk if the goods are wrong.
Some of that gap is margin. Much of it is work that has to be done by somebody. The useful question is not "why is the export price higher" but "who is doing each of these things, and am I paying a fair price for it". That is also why our cost pages quote landed cost rather than factory price — the factory price on its own answers almost nothing.
What none of them tell you
The thing every one of these channels is silent about is the only thing that decides whether an order works: whether this particular company can make your particular product, at your quality, at your volume, repeatedly. No directory answers that. No badge answers it. A good conversation at a fair only narrows it.
That answer comes from checks rather than from searching — reading the registration rather than a photograph of it, asking the operational questions a reseller cannot answer, and looking at goods while the line is still running rather than after it has stopped. Supplier verification covers the first two and when to inspect a China order covers the last. Picking the right channel gets you a better shortlist; it is not a substitute for checking the shortlist.
