Tmall, JD and Douyin: How the Channels Differ

Tmall, JD and Douyin are three of the channels through which consumer goods are sold online in China, and they are often discussed as though they were interchangeable places to list the same products, when in fact they differ in why the shopper is there, and most of the commercial differences between them follow from that.

The core difference is consumer intent#

These are tendencies rather than laws, and all three now sell across almost every consumer category, but the reason a shopper opens each app has proved durable, and it is what makes generalising from one channel to another unreliable: a brand's position on one channel predicts surprisingly little about another.

Why figures are not directly comparable#

Each platform makes its own defensible choices about what a sale is:

• whether cancelled orders are removed, and after how long
• when a pre-sale transaction counts — at deposit or at balance payment
• whether platform-funded subsidies are included in transaction value
• whether shipping is included
• how variants group into listings, which changes any per-product figure

Every one of these moves the number. A cross-platform comparison assembled from each platform's own reporting therefore compares definitions at least as much as it compares markets. See GMV for what each of these adjustments does.

There are two sound approaches. Use a single independent source applying one consistent definition across all three — which makes the channels comparable to each other, though not necessarily equal to any platform's own announcements. Or establish each platform's conventions explicitly and adjust. What does not work is taking three published numbers at face value.

Data behaviour differs too#

History depth is not uniform. The established marketplaces have been observable for far longer than Douyin, which entered most datasets in 2022. Any multi-year category series spanning all three contains a point where a channel appeared. Growth computed across that point partly measures the dataset expanding rather than the market growing — a specific instance of the restatement problem in Data Coverage and Completeness.

Volume stability differs. Marketplace weekly figures are a reasonable sample of an underlying rate. Content-channel weekly figures are largely a record of whether an event fell inside the week. Applying one analytical window to all three produces a false impression of volatility in one and false stability in another.

Price distributions differ for the same goods. The same product frequently transacts at different prices across channels. A single blended price distribution can therefore show a cluster that exists on no individual platform — which is why Price Band Analysis should be run per channel where the mix differs materially.

Does a brand need all three?#

Presence on all three is often treated as the default, and it is how budgets get spread too thin to work anywhere. Each channel requires a genuinely distinct operating model — different content, different pricing cadence, different service expectations — not a copy of the first one with the logo changed.

The more useful question is: in which channel is my specific advantage legible? A brand whose advantage is best explained through comparison, specification and an established brand story is legible where shoppers come to compare, and a brand whose advantage shows best on camera, in a demonstration or a first reaction, is legible in content. A brand whose advantage is price and availability is legible in any channel where it can hold that price and keep the product in stock.

Then, separately: does a second channel justify the operating model it demands?

Reading them together#

The channels influence each other, and no platform's reporting can see it. Content exposure on one routinely produces search and purchase on another, because consumers often discover a product in one place and buy it in another.

Reading each channel in isolation therefore undercounts the discovery channels and overcredits the conversion ones — which leads directly to cutting the activity that was creating the demand. The practical corrective is to watch whether marketplace search and sales move following content activity, and to treat each platform's own attribution as a partial view. See Douyin E-Commerce for the mechanism, Xiaohongshu Marketing for a discovery platform that often sits upstream of both, and Share of Search for the signal that usually carries the effect.

Where to look next#

For the content channel in depth, see Douyin E-Commerce. For discovery and consideration, see Xiaohongshu Marketing. For comparing sales figures safely, see GMV. For per-channel price work, see Price Band Analysis. For history and coverage differences between platforms, see Data Coverage and Completeness.

Common questions#

How do the three channels actually differ?#

Principally in why the shopper is there. On Tmall the shopper is often browsing and comparing brands, on JD the shopper is often buying something specific they already had in mind, and on Douyin the shopper has come to be entertained and meets the product inside a video or a livestream, so demand there is created rather than matched. Those differences in intent shape how products are found and how sales build up over time on each channel, which is why a brand's position on one channel predicts surprisingly little about another.

Can sales figures be compared directly across the three?#

Not without normalisation. The platforms differ in how they treat cancelled orders, when pre-sale transactions count, whether platform-funded subsidies are included in transaction value, and how product variants are grouped into listings. Each is a defensible choice and each moves the number. A cross-platform comparison built on each platform's own reported figures therefore compares definitions as much as markets. Either use one independent source applying a single consistent definition across all three, or establish each platform's conventions explicitly before comparing anything.

Does a brand need to be on all three?#

Not necessarily, and treating presence as automatic is how budgets get spread too thin to work anywhere. The channels reward different things: a brand whose advantage is best explained through comparison and an established brand story fits the marketplace environments, where shoppers come to compare, while one whose advantage shows best on camera fits content. The more useful question is which channel your specific product advantage is legible in, then whether the others justify a distinct operating model — because each genuinely requires one, not a copy of the first.

How should a brand read performance across channels together?#

By expecting the channels to influence each other rather than to operate independently. Content-led exposure on one platform routinely produces search and purchase on another, and no platform's own reporting can see that. Reading each channel's numbers in isolation therefore undercounts the discovery channels and overcredits the conversion ones. The practical approach is to look for movement in marketplace search and sales following content activity, and to treat each platform's attribution as a partial view rather than as the answer.