In Shenzhen, drones zip across the skyline delivering coffee and lunch orders, while at Chongqing’s Jiangbei International Airport, robots glide through terminals serving takeout. At the same time, millions of people across mainland China shop, scan products, plan travels, and book doctors’ appointments—all within minutes on a single app. These are everyday scenes of how AI-powered super apps are reshaping daily life in China.
China’s digital infrastructure evolved differently from the rest of the world. Unlike markets reliant on standalone websites and apps, China’s mobile internet coalesced around super apps (WeChat, Alipay, Douyin, Meituan) as the hook for traffic acquisition, engagement, and commerce. Lightweight Mini Programs (小程序) allow brands to operate as “tenants” inside these ecosystems, trading some infrastructure control for frictionless scale.
Chris Kwok, data marketing director at Omnicom Media Hong Kong, has been advising brands on bridging these worlds.
Closed-loop vs fragmented journeys
Outside mainland China, in Hong Kong and the rest of the world, brands build and maintain independent corporate websites and native iOS/Android apps. They own the full infrastructure stack and control the entire user journey from start to finish. But control comes at a cost. Users are forced to jump between social ads on Meta, Google, or TikTok, brand landing pages, e-commerce platforms, and separate checkout flows. The friction points accumulate fast: slow page loads, guest checkouts, manual sign-ups, app download barriers, and data trapped in silos across disconnected tools.
In mainland China, brands have been consolidating with Mini Programs within the super apps since 2016. This Mini Program architecture, or what Omnicom Media Hong Kong’s data marketing director Chris Kwok calls “tenant model”, reshapes the traditional consumer journey into a closed-loop system, collapsing intent, engagement, and transaction into minutes without leaving the app. “The strategy moves from driving users to [the company’s] websites or apps to converting users where they already are,” says Kwok, “this affects [user experience] design, data collection, measurement, CRM, and retargeting.”
The tenant model is only half the story. The real competitive edge for Chinese brands is a robust three-stage private domain strategy on top of it.
Stage one involves acquiring users through paid and organic activity on Douyin, Weibo, Kuaishou, and Baidu. Stage two will be pulling them into owned channels with hard incentives like discounts, free samples, VIP perks. Retain and monetise through WeCom (企业微信), WeChat's enterprise CRM tool, using one-on-one conversations, community groups, campaign invitations, and personalised promotions, is the last stage of this strategy.
Every interaction feeds into a CDP or CRM that builds a single customer profile with behaviour patterns, content preferences, personal data, and audience tags. Brands then run automated sequences such as onboarding journeys, content nudges, behaviour-triggered messages, and win-back campaigns for lapsed users. Those who still haven't converted get retargeted via lookalike audiences on Meta, Google, WeChat, and XHS, matched using phone numbers or WeChat IDs.
The measurement gap
When a Douyin ad converts inside the Douyin store, there is zero data loss as the entire journey lives on one platform. In other Asian markets, the handoff between, let's say, a Meta ad and a brand website is where iOS App Tracking Transparency and ad blockers do their damage and attribution falls apart. Third-party tools like Campaign Manager and AppsFlyer exist to patch this gap.
China's cross-platform picture has its own limitations. Tools like Miaozhen (秒针系统) and TalkingData offer multi-touch attribution across walled gardens but with restricted capability compared to global equivalents.
Inside Mini Programs, brands need an entirely different analytics infrastructure. Sensors Data (神策数据) and GrowIO provide native SDKs that integrate directly into WeChat and Alipay Mini Programs, track unique users by member ID or phone number, and unify behaviour across WeChat, Taobao, Alipay, and brand-owned channels into a single view. Google Analytics 4 does not work here.
The divide in advertising platforms is equally pronounced. Outside Mainland, Asian brands rely on Meta Ads Manager, Google Ads, and Display & Video 360. Within the Chinese ecosystem, ByteDance, Tencent Ads, Alibaba, and Xiaohongshu are the dominant players.
Each of these ecosystems operates with distinct targeting capabilities, creative formats, and measurement systems. As a result, they demand fundamentally separate strategies, localised creative approaches, and independent performance frameworks.
The agentic era
While much of the world has treated AI as an experiment until recently, China has taken a more decisive path with its next leap forward into embedding agentic AI directly into the shopping experience, supporting the full journey of discovery, recommendations, service, payment, and after-sales.
“The trust in platforms like WeChat was established long before embedded AI became a feature. Super apps were already offering a seamless ecosystem for functions ranging from payment and messaging to shopping and booking services. In short, AI adoption does not need to build trust from zero – it amplifies the trust these platforms have already earned,” says Kwok.
On Taobao, the AI assistant Qianwen (千问) lets users search by natural language, virtually try on clothing using a photo, optimise across budget constraints, manage orders, and check out—all without leaving the app. Meanwhile, Meituan's Xiaomei goes further and lets users delegate tasks to it, such as ordering a meal for a specific delivery time, rather than searching manually.
Another example is Douyin, whose e‑commerce gross merchandise value (GMV) reached $490 billion in 2024, up 30% year‑on‑year and projected to hit $536 billion in 2025. Within this surge, the number of brands running their own livestream sessions grew 113% year‑on‑year, and nearly 70% of livestream‑driven GMV now comes from store‑operated streams. At the same time, WeChat surpassed six million active Mini Programs in 2025. China’s e-commerce market is projected to grow to $6.2 trillion by 2035, while agentic commerce is expected to expand to $5 trillion by 2030. With trillions in potential value at stake, China offers a vivid example of how agentic AI may reshape consumerism globally.
Compare that to anywhere else in Asia: AI is still an isolated tool. Consumers use ChatGPT or Gemini to research products and brands deploy chatbots on their own websites for support queries.
Playing catch-up with China
The global market's answer to the super app ecosystem is the open protocol.
“Outside China, AI may appear less mature, but it is more likely to serve as the connective tissue across the open web – linking search, e-commerce sites, marketplaces, and brand‑owned platforms," says Kwok, "Instead of one super app owning the journey, AI could become the layer that unites many separate destinations. Google’s recent announcement of its open‑source Universal Commerce Protocol and Universal Cart are clear examples of where the rest of the world is heading.
The commercial logic is the same everywhere: remove friction, meet people where they are, and let AI do the joining. The race to build that experience across the open web is already underway.
