China: an internet built on different rules
Facebook is blocked. Twitter is not permitted. Google is a memory for most of the country. We are talking about a market where the vast majority of internet users are active on social media every day, on a mobile network that reaches almost the entire population. If you still think of China as a smaller version of the western internet, you are planning your entry wrong before you even start.
China banned Google search and the Google Play store years ago. Chinese developers filled that gap themselves, and they did it well. Phone makers built their own app stores, internet giants built their own discovery tools, and an entire generation grew up never knowing what Google Play looks like. That gap is why China’s mobile app world runs on different rules than everywhere else you have launched.
Apple still sells well in China, but its position has moved around a lot over the past decade, squeezed by Huawei, OPPO, Vivo and Xiaomi in different years. What has not moved is how central the phone is to daily life here. You buy groceries with a QR code scan. You pay a street vendor with WeChat Pay. You order a taxi, book a doctor’s appointment and split a dinner bill, all without opening your wallet once.
I’m Olivier Verot, founder of GMA in Shanghai. I’ve watched Chinese consumers move from downloading apps one by one to living inside WeChat mini programs that need no download at all, and I’ve helped foreign brands make that same jump without losing the following they built along the way.
What makes China’s app market different
China has the largest smartphone user base on earth. Android leads by raw user count, though the gap with iOS has narrowed as domestic phone makers push premium models with better cameras and faster chips. Either way, betting on one platform and ignoring the other is not an option here.
How people find apps has changed a lot since this guide was first written. Tencent’s old MyApp store, once the default download hub for Android users, has faded as phone makers built their own stores directly into the operating system: Huawei AppGallery, OPPO’s app market, Vivo’s, Xiaomi’s GetApps. Users mostly download through whichever store came pre-installed on their phone. And a growing share of what they use every day, they never download at all. It lives as a WeChat mini program instead.
Local apps still dominate the charts
Foreign apps rarely crack the top of China’s charts, and that has not changed. Local developers understand local payment habits, local ID verification requirements and local content preferences better than anyone building from outside. If you are entering with a foreign-built app, plan for a local team to adapt it, not just translate it.
Baidu, Alibaba, Tencent: still the three companies that matter most
Baidu, Alibaba and Tencent, still referred to together as BAT, each run several social, search or e-commerce platforms with hundreds of millions of active users apiece. That has not changed since this guide first tracked them, only the specific apps they push have. What has stayed consistent is how Chinese developers design for these platforms: one app carrying many features, rather than the string of single-purpose apps western users are used to installing. A Chinese super app handles messaging, payment, ordering and customer service in one place. A western brand that ships five separate apps for five separate functions is, by local standards, making the user do unnecessary work.
How Chinese users actually find an app
It still helps to think about Chinese mobile users in three groups, a split this guide has used for years and one that still holds up.
The first group already knows exactly what they want. They see a product in a Douyin livestream or a Xiaohongshu post and buy it inside the same session, no browsing, no comparison, a “frictionless” path from seeing to buying is the only thing that matters to them.
The second group knows the outcome they want but not which app delivers it. They search app stores by keyword, read reviews, try something, and delete it fast if it disappoints. Mobile games are the clearest example: a player searches “MOBA” or a specific game title directly inside an app store rather than following a link from anywhere else.
The third group is open to suggestion, and this is the group today’s recommendation engines are built for. Douyin, Xiaohongshu and WeChat all now use behavior and interest data to surface an app, a mini program or a product the user never searched for. This is the mechanism behind interest e-commerce, covered in more detail below, and it is a far more sophisticated version of what this guide first described as store curated placement.
China’s mobile app market in 2026: what the numbers actually say
China’s 57th CNNIC report put the country’s internet population at 1.125 billion people as of the end of 2025, with penetration at 80.1 percent. Mobile accounts for 99.6 percent of that. There is no meaningful desktop audience left to plan for. Every asset you build, every product page, every video, gets judged on a six-inch screen first.
WeChat sits at the center of that mobile life. Tencent’s own first-quarter 2026 earnings put combined WeChat and Weixin monthly active accounts at 1.432 billion, up 2 percent year on year, according to Tencent’s earnings coverage. WeChat is not a messaging app anymore, if it ever really was one. It is where people pay, where they shop, where brands run customer service, and where a growing number of businesses run their entire storefront.
Inside WeChat, mini programs are the reason a native app is often unnecessary. A WeChat mini program loads instantly from a QR code or a chat link, needs no download, and disappears from the user’s phone the moment they stop using it, no storage taken up, no update prompts. WeChat’s own mini game category alone passed 500 million monthly active users this year, and mini shops built the same way now carry a meaningful share of social commerce for brands that never built a standalone app at all. If your China strategy still centers on getting users to download something, you are fighting a battle most competitors gave up years ago.
This guide talked about this shift before it had a name. An earlier version of it noted Tencent testing a new kind of “mini-app” inside WeChat that would need no download, along with an early chatbot platform for developers. That experiment is exactly what mini programs became. It is worth remembering when a platform tells you it is testing something small: in China, today’s experiment is often tomorrow’s default channel.
Douyin is the second pillar. Short-video platforms in China had never crossed the 1 billion monthly active user mark until this year: Douyin did it, reaching just over 1 billion monthly active users in early 2026, a 14.43 percent jump year on year, according to QuestMobile data reported by Sina. Average daily use runs past 1.5 hours per person. Douyin users do not scroll and leave. They watch a live host demonstrate a product, ask a question in the comment thread, and buy before the stream ends. That is live commerce, and it now works alongside a fast-growing local life services push, which puts Douyin in direct competition with local giants like Meituan, the food delivery and local services app Wang Xing founded back in 2010, for bookings, delivery and in-store deals, not just for attention.
Xiaohongshu (Little Red Book) plays a different role entirely. It is not where most people buy. It is where they decide to want something in the first place. Brands call this “grass planting” (种草): a product shows up in a real-looking post from someone who seems like a peer, not a spokesperson, and the desire it plants gets acted on later, on Douyin, on Tmall, or inside a WeChat mini shop. Recent platform reporting put Xiaohongshu’s e-commerce gross merchandise volume above 850 billion yuan, a scale that makes it far more than a discovery layer for brands that use it properly. Skip Xiaohongshu and you are skipping the exact moment where the purchase decision actually forms.
The KOC shift Western brands still underestimate
One change catches most new entrants off guard. Chinese consumers, especially in fashion and beauty, have grown skeptical of big-name KOLs with millions of followers and an obvious paid relationship with the brand. They trust KOCs instead: creators with roughly 10,000 to 100,000 followers, who read as real people sharing a real opinion rather than reciting a script. A single mega-influencer post used to be the goal. Now a batch of ten to twenty honest-sounding KOC reviews across Xiaohongshu and Douyin often outperforms it, at a fraction of the cost. This shift matters as much for a fashion label as it does for the wider China fashion market, where peer recommendation has quietly replaced celebrity endorsement as the trust signal that moves a sale.
Solutions that did not exist when this guide was first written
A few mechanisms are worth explaining properly, because most agency content mentions them by name and stops there.
GEO on DeepSeek and Doubao. Chinese consumers increasingly ask AI assistants for recommendations instead of typing a search query, the same shift happening globally but on domestic tools. Getting your brand mentioned inside a DeepSeek or Doubao answer works differently from ranking on Baidu: it depends on how clearly your brand is described across the web in structured, factual language these models can retrieve, not on backlinks or keyword density. WeChat has already integrated DeepSeek directly into search, which means a mini program with clean, well-structured content can now surface inside an AI answer without any paid placement at all.
Private domain traffic on WeChat. Instead of paying repeatedly to reach the same customer through ads, brands move buyers into an owned channel: a WeChat group, an Official Account, a mini shop membership. Once someone is inside your private domain, every future message is free to send and lands directly in their chat list. The mechanism only works if you give people a reason to stay, restock alerts, member pricing, direct access to customer service, not just more promotions.
Interest e-commerce on Douyin. Traditional e-commerce waits for someone to search for a product. Interest e-commerce shows the product to someone who never searched for anything, based on what they watch and how long they watch it, then lets them buy without leaving the video feed. The mechanism relies on Douyin’s recommendation engine being better at predicting desire than the shopper is at expressing it. That is why a demonstration video that holds attention for the first three seconds matters more here than a polished product shot.
AI-driven customer service. Chinese consumers expect an answer inside WeChat within minutes, day or night. AI-assisted first response, handling order status, sizing questions and return policy automatically, then routing anything complex to a human, has become close to a baseline expectation rather than a nice extra, and it directly affects conversion inside a mini shop, where hesitation before checkout is often just an unanswered question.
Torvald’s mini shop: a real pattern, not a real name
Torvald runs a small Norwegian skincare brand that had been selling in China through a local distributor for two years, with underwhelming results. His only direct presence was a WeChat Official Account with about 4,000 followers, mostly built from translated posts pulled straight from the European website. Over six months, that account produced eleven traceable sales.
His first fix was to spend more on WeChat Moments ads to grow the follower count, roughly 1,800 dollars a month for two months. Followers doubled to around 8,000. Sales stayed flat, about two orders a week, because there was still nowhere to buy without leaving the app and hunting for a separate store.
What actually worked was smaller and cheaper. He redirected that same budget toward building a WeChat mini shop directly inside the Official Account, and paired it with twelve Xiaohongshu posts from KOCs with 15,000 to 40,000 followers, reviewing the products honestly and linking back through a QR code straight into the mini shop. It worked for a simple reason: it removed every extra step between wanting the product and buying it, no separate app, no separate login, no separate site to trust. And the KOC content read as an opinion, not an advertisement, which matters enormously in skincare, where buyers are wary of brand copy but responsive to a stranger who looks like them.
Within four months, monthly mini shop orders went from around 15 to about 140. Repeat purchase rate crossed 20 percent, something the old setup had no mechanism to produce at all, since it never gave a past buyer an easy way to come back.
The apps still worth knowing
Xiaohongshu (Little Red Book)
Founded in 2013 for overseas shopping guides, Xiaohongshu grew into China’s closest equivalent to Pinterest, except with a far more direct line to purchase. It remains the platform most foreign brands underestimate on their first attempt, usually because they treat it like Instagram and post polished brand photography instead of the review-style content the platform actually rewards.
Read: Little Red Book Guide
Taobao
Taobao is still where a large share of Chinese shoppers go to browse, not just to buy. People scroll it the way others scroll social feeds, out of habit and curiosity as much as intent. Alibaba’s own reporting on user behavior, covered by Tech in Asia, has long noted this browsing pattern, and it still holds: a strong Taobao storefront gets discovered by people who were not looking for you.
Read: Taobao Guide
WeChat and its mini shops
QQ, WeChat’s older sibling from the desktop instant-messaging era, is still alive but firmly a legacy product now, kept going mostly by younger gamers and school-age chat groups. WeChat took over everything else: messaging, payments, and increasingly, direct retail through mini shops built inside the Official Account itself. For most foreign brands, WeChat is not one channel among several. It is the operating system their entire China presence runs on top of.
Douyin
Douyin did not exist in an earlier version of this guide, which tells you something about how fast this market moves. It is now the platform most likely to convert a viewer into a buyer without them ever leaving the app, through live streams where a host demonstrates the product and answers questions in real time. Brands watching Tmall traffic soften have been shifting budget here, a move covered in detail in our piece on why brands are leaving Tmall for Douyin.
Mobile Baidu
Baidu remains China’s dominant search engine, and the mobile version carries the vast majority of that traffic. What has changed is what happens after the search: Baidu’s own AI answer features now sit above the classic blue links for a growing share of queries, which means classic SEO and the newer discipline of getting cited inside an AI-generated answer now need to be planned together, not as separate projects.
Youku
Youku, Alibaba’s long-form video platform, is no longer the growth story it was when this guide first covered it. Short video on Douyin and Kuaishou pulled attention away from long-form viewing years ago, and Bilibili and iQIYI now compete harder for the audience that still wants longer content. Youku remains relevant for licensed film, TV and certain sponsorship placements, but it is not where a mobile-first brand launch should start in 2026.
See also: Top social media platforms in China and 8 social media apps to master in China, both useful next steps once you know which app world you are entering.
Frequently asked questions
Do I actually need a standalone app to sell in China?
No, and most foreign brands entering China skip building one. A WeChat Official Account with a mini shop, paired with a Xiaohongshu presence for discovery, covers what an app would do for a new entrant: content, payment and community, without download friction or an app store approval process. Build a dedicated app later, once your customer base is large enough to justify the ongoing maintenance.
Is Google Play really unavailable to everyone in China?
Yes, for practical purposes. Android users install apps through phone-maker stores such as Huawei AppGallery, OPPO’s store, Vivo’s store or Xiaomi GetApps, or they use a WeChat mini program that needs no download at all. If your go-to-market plan still assumes Google Play distribution, it needs to be rebuilt around these channels instead.
How long does it take to get a WeChat mini shop live?
Once your Official Account is verified, which itself needs a Chinese business entity or a licensed local partner, the mini shop build itself takes one to two weeks: product listings, WeChat Pay integration and basic storefront design. The slow part is almost always account verification, not the shop build.
If I only have budget for one platform, which one should I choose?
It depends on where your buyer already spends time. For considered purchases like skincare, fashion or home goods, start on Xiaohongshu to build trust, then move buyers into a WeChat mini shop to close the sale. For impulse or lower-cost items, Douyin’s live commerce format converts faster, since the purchase can happen inside the same viewing session.
Do I need a Chinese business license to open a mini program or mini shop?
For a full mini shop with local payment collection, yes, you need a registered Chinese entity or a licensed partner operating it on your behalf. Lighter, informational mini program formats exist without a license, but they cannot process payments directly. Most foreign brands start through a local partner rather than registering their own entity right away.
Where GMA fits in
We build and run WeChat mini shops, Xiaohongshu KOC campaigns and Douyin live commerce setups for foreign brands entering China, from the first Official Account verification through to a working private domain strategy. If your brand is still trying to figure out which of these apps deserves your first budget, that conversation is the one we have most often with new clients, and it usually saves them from repeating the mistake Torvald made the first time around. Reach out through our contact page to start that conversation.
