Ultimate Guide To eCommerce In China

China is the largest e-commerce market on earth. More goods are bought online there than in the United States and Europe combined, and a very large share of that shopping happens on a phone, inside an app, often while watching a video. If your brand sells anything a Chinese consumer might want, the question is not whether the market matters. The question is whether you can afford the way it works.

Last updated: 27 August 2026.

In this guide:

• Why the market size is not the argument you think it is
• The platforms, and what each one is actually for
• Cross-border or domestic: the first real decision
• Discovery and transaction are two different jobs
• Logistics and payment, briefly
• What a realistic first year looks like
• When not to do this
• AI shopping is no longer a feature, it is a new front door
• Instant retail is pulling share away from next-day delivery

This guide maps the whole picture for a brand starting from zero: the platforms, the two ways in, the split between discovery and transaction, the plumbing of logistics and payment, and what a first year actually looks like when nobody is selling you a dream.

Why the market size is not the argument you think it is

Brands hear “hundreds of millions of online shoppers” and assume volume solves everything. It does the opposite. A big market with cheap traffic rewards patience. A big market with expensive, contested traffic rewards preparation. China is the second kind. Customer acquisition on the major platforms is bought, not earned, and the auction is crowded with domestic brands who understand the mechanics better than you do.

So treat the size as a reason to be serious, not a reason to be optimistic.

The platforms, and what each one is actually for

There is no single “Chinese Amazon”. There are several platforms with different jobs.

• Tmall (Alibaba): the flagship store venue. Brand-controlled, brand-facing, and where most foreign consumer brands build their official presence.
• Taobao: the open marketplace beneath Tmall. Huge, messy, dominated by small sellers. Rarely the right first move for a foreign brand.
• JD.com: strong in electronics, appliances, and anything where consumers care about authenticity and fast delivery. JD runs much of its own logistics.
• Pinduoduo: price-led, group-buying roots, enormous in lower-tier cities. A poor fit for premium positioning.
• Douyin (Chinese TikTok): short video and livestream selling. Both a media channel and a shop.
• Xiaohongshu (RED): review and recommendation platform, especially for beauty, fashion, wellness and travel. Increasingly sells directly too.
• WeChat: not a marketplace, but the layer that holds CRM, mini-programs, payment and private traffic together.

Most brands should not open on more than two. Spreading across five stores with one team produces five neglected stores. If you want the detailed side-by-side on the two biggest, our comparison of Tmall and JD goes deeper than we can here.

Cross-border or domestic: the first real decision

Everything downstream depends on this choice, so make it deliberately.

Cross-border means selling from outside China into China, typically via Tmall Global or JD Worldwide. You keep stock in Hong Kong, a bonded warehouse in a free trade zone, or your home country. You do not need a Chinese company. Product registration requirements are lighter, which matters enormously for cosmetics, supplements and food. The trade-off: slower delivery, per-order tax treatment, personal-import annual limits per consumer, and a harder time joining certain platform promotions.

Domestic means selling from inside China, on Tmall or JD proper. You need a Chinese legal entity or a partner acting as one, local registration and certification for your goods, local warehousing and an RMB bank account. Delivery is next-day, margins on logistics are better, and you get full access to promotions and traffic tools.

Our position, unchanged for years: start cross-border unless you already have a Chinese entity or your product needs no registration. Cross-border is a test that costs tens of thousands of RMB. Domestic is a commitment that costs a great deal more, and it is far easier to move from cross-border to domestic once you have sales data than to unwind a domestic set-up that never found demand. The practical steps for opening a Tmall Global store are worth reading before you budget anything.

Discovery and transaction are two different jobs

This is the concept most Western brands miss, and it explains most failed launches.

Chinese consumers rarely discover a foreign brand on Tmall. They discover it on Xiaohongshu, in a review from someone they follow, or on Douyin, in a video that ends with a demonstration. Then they search the brand name on Tmall or JD to buy. Search demand on the transaction platforms is largely created elsewhere.

The practical consequence: a Tmall store with no content behind it is a shop on an empty street. You will pay for every visitor through paid search, your conversion will be poor because nobody has heard of you, and you will conclude that China does not work. What did not work was the sequence.

Build a modest volume of credible content first. Fifty honest posts from small creators who genuinely use the product beat one campaign with a famous name. Livestreaming with top hosts is largely not worth it for a new foreign brand: the fees are high, the discount demanded is deep, and the sales stop the moment the stream ends.

Logistics and payment, briefly

Payment is simple. Alipay and WeChat Pay together hold the large majority of the third-party payment market, and both are built into the platforms. You do not need to solve payment. You do need a route for money to leave China legally, which is where your entity structure or your partner matters.

Logistics is where cross-border sellers lose customers. Chinese consumers are used to next-day delivery. A fourteen-day wait from Europe reads as a broken promise, whatever your terms page says. Bonded warehouses in zones like Ningbo, Zhengzhou or Hangzhou let you pre-position stock inside China while keeping cross-border tax treatment, cutting delivery to a few days. If you are serious about cross-border, budget for bonded stock in year one rather than shipping parcel by parcel.

Returns and after-sales are handled in Chinese, fast, on the platform’s own chat tool. Slow replies damage your store rating, and store rating affects the traffic you receive. This is not a task to leave to head office in another time zone.

What a realistic first year looks like

1. Months 1 to 3. Register trademarks in China, in Chinese characters as well as Latin script. Pick a Chinese brand name properly. Choose your entry route. Prepare documents and open the store. Nothing sells yet.
2. Months 3 to 6. Store goes live. Content begins on Xiaohongshu or Douyin. Seed products with small creators. Sales are low and that is normal. You are gathering data on which product, which price, which message.
3. Months 6 to 9. You know your best-selling SKU. Spend shifts towards it. Paid search on the platform becomes worth running because there is now branded demand to capture. First real promotion participation.
4. Months 9 to 12. Singles’ Day in November is the first honest test of whether the machine works. Review the numbers. Decide whether to scale, change platform, or stop.

Breaking even inside twelve months is uncommon. Two years is a fairer expectation for a brand starting with no awareness. Any agency promising profit in six months is selling you the pitch, not the market. Running a store day to day is a full-time operational job, which is why most brands hand it to a team that manages Tmall accounts professionally rather than adding it to someone’s existing role in Europe.

When not to do this

Do not enter China if your annual marketing budget for the market is under roughly 300,000 RMB. Do not enter if nobody on your team can work in Chinese, and you are unwilling to hire or outsource that. Do not enter if your product has a strong domestic equivalent at half the price and no story that justifies the gap. And do not enter simply because a distributor emailed you asking for exclusivity: that is their strategy, not yours.

The brands that succeed here are rarely the biggest. They are the ones that picked one platform, one product, one audience, and stayed long enough to learn.

Sources: Alibaba and JD annual reports 2025-2026; QuestMobile and CNNIC China internet data; EAC client account experience.

AI shopping is no longer a feature, it is a new front door

2026 is being called the first year of mass AI shopping adoption in China, and the shift is not cosmetic. Consumers are starting a purchase inside an AI assistant rather than inside a store search bar. By the end of February 2026, close to 140 million users had tried AI-driven shopping for the first time through the agent features inside Alibaba’s Qwen app, and on Taobao’s own merchant side, more than 95% of sellers surveyed by Alibaba say they already use AI in some part of their operation, with 88% planning to keep or increase that spending. Amazon’s China-facing sellers report similar numbers: over 98% of surveyed cross-border sellers now use AI tools somewhere in running their store, and 16% have moved past single-purpose tools into running actual AI workflows or agents, according to TMO Group’s 2026 review of agentic commerce in China.

What changes for a foreign brand is the layer where the buying decision actually gets made. Survey data circulated in China’s 2026 AI shopping consumer report finds that 53% of shoppers say they would try a new brand because an AI recommended it, and 39% say AI has directly changed what ends up in their cart, per National Business Daily’s 2026 AI shopping insight report. In practice this means the competition is shifting from being remembered by a human scrolling a feed to being recommended by a model answering a question. A product page written only for a human reader, with vague copy and no structured specification data, is increasingly invisible to the assistant doing the shortlisting on the buyer’s behalf. Brands serious about the next two years should treat clean, structured product data (materials, certifications, use cases, comparison points) as a ranking input, not a nice-to-have, the same way search engine optimization once was.

The platforms are also simplifying in response. The June 2026 mid-year shopping festival was widely reported as the first “AI-native” major promotion, with platforms cutting the cross-store discount stacking, team-building and point-collecting mechanics that used to define Chinese sales events, in favor of flatter, more transparent pricing, according to Sina Finance’s coverage of the 618 shopping festival. That simplification is partly a direct response to AI shopping agents, which cannot reliably navigate the old maze of coupons and rules on a shopper’s behalf. A brand still pricing for the old game, deep nominal discounts offset by complicated conditions, is optimizing for a shopper who increasingly is not the one making the final click.

Instant retail is pulling share away from next-day delivery

The other shift worth planning around is speed. Instant retail, ordering through an app and receiving goods within one to two hours from a nearby warehouse or store rather than the next day, grew far faster than traditional e-commerce during the 2025 Double Eleven shopping period: roughly 28% growth for instant retail against about 12% for traditional online retail over the same window, according to industry data cited in a 2026 e-commerce outlook report. Chinese consumers increasingly hold two separate expectations at once: patient, planned buying for considered purchases on Tmall or JD, and near-instant fulfillment for anything they decide they want right now, from skincare to snacks to small electronics. The market is starting to describe this as a “far-field planned, near-field instant” hybrid, rather than one delivery standard replacing the other.

For a cross-border brand this trend mostly plays out indirectly for now, since instant retail depends on physical stock already inside China, which favors domestic-entity brands and larger platforms’ own dark-store networks. But it changes the delivery bar a foreign brand is judged against. A Chinese shopper who is used to receiving a coffee order in twenty minutes does not become more patient about a cross-border parcel simply because the product is foreign. If anything, the gap reads worse. This is one more argument, alongside the tax and personal-import limits already covered above, for treating bonded warehouse stock as a year-one investment rather than a year-three upgrade, since it is the only lever a cross-border brand has to close part of that speed gap.

Explore each platform in depth

This overview is the starting point. Each platform has its own dedicated guide, kept current with 2026 rules and data.

• Tmall (domestic store): the step-by-step process for opening a domestic Tmall store, from legal entity to product registration.
• Tmall Global (cross-border): how CBEC works on Tmall, entry models, deposits, and the 2026 category rules.
• Tmall vs JD: a category-by-category comparison to decide which platform to launch on first.
• Douyin: content strategy, budget, and entry requirements for foreign brands.
• Xiaohongshu (RedNote): KOC seeding, search visibility, and the 2026 closed-loop commerce shift.
• WeChat: account types, mini-programs, and why WeChat is a retention channel, not an acquisition one.
• Baidu SEO and SEM: how AI search is changing content strategy in 2026.
• CBEC strategy: the full compliance and entry sequencing for a brand with zero China presence.
• Shopping festivals: the full festival calendar and how to plan a promotion budget around it.

FAQ: Entering the China eCommerce Market

How long until we see real traction in China, realistically?

Six to nine months to reach meaningful, repeatable sales, not the first Tmall or Douyin transaction. That first sale can happen in week two if the store is set up correctly. Traction is different: it means organic search visibility, a KOL or livestream channel that converts without heavy subsidy, and a repeat purchase rate you can plan around. Brands that budget for three months usually pull out right before the curve turns, because months one to three are almost entirely store setup, platform approval, initial content production and the first rounds of paid traffic testing. The brands that hit traction fastest share three traits: they already have product-market fit somewhere in Asia, they commit a real content and livestream budget instead of treating China as a side channel, and they pick one platform to dominate before spreading thin across five. If a consultant promises profitability in sixty days, ask what they are counting as profit, because it is rarely the same number your finance team would sign off on.

Do we need a Chinese legal entity before we can start selling?

No, not to start. Cross-border eCommerce (CBEC) exists specifically so foreign brands can sell into China through bonded warehouses on Tmall Global, JD Worldwide or Douyin, using their existing overseas entity and a local operating partner or TP (Tmall Partner) to hold the store license. This is how most brands should enter: no WFOE, no local bank account, no China-based staff, and products can stay under simplified CBEC tax treatment instead of full import duties. A local entity becomes necessary only once you want a domestic Tmall or JD store instead of the cross-border version, need to hold your own ICP license for a China-hosted site, or want to run a Chinese payroll and hire local sales staff directly. Setting up a WFOE before you have any sales data is the single most common expensive mistake we see: it commits capital, requires ongoing accounting and tax filings, and buys nothing that CBEC does not already give you for the pilot phase.

Is it true the rules for cross-border eCommerce have changed in 2026, and does that affect us?

Yes, and it matters most for brands scaling past a specific revenue point. Regulators have tightened enforcement around a threshold: once a brand’s annual CBEC sales through one platform or channel cross roughly RMB 1 million, authorities increasingly expect that brand to meet general trade compliance standards for product registration and qualification, not just the lighter CBEC filing rules that cover smaller sellers. Mid-size brands doing the equivalent of RMB 500,000 to 2 million a year are the most exposed, because they are large enough to draw attention but often still running on documentation built for a much smaller operation. This sits inside a wider revision: China’s updated Foreign Trade Law takes effect in March 2026 and gives regulators broader tools to manage compliance and platform responsibility. Details are covered by China Policy Guide’s regulatory breakdown. If your CBEC revenue is approaching that threshold, get your product registration file audited before the platform flags it, not after.

Harry Huang is the founder of EAC. He spends most of his week inside store dashboards rather than in strategy decks, which shapes how he advises brands.

Thinking about this market? Send us your questions.