ALIBABA OPENS ANOTHER ATTACK ON MEITUAN WITH AMAP AND AI

Amap (a.k.a. AutoNavi and Gaode) has been my Chinese navigation app of choice for a while now. This Alibaba app is now opening a full attack on Meituan’s review app Dianping, at a time when Meituan is already been pestered by Alibaba’s Taobao Instashopping.

Alibaba once invested in the local services platform Koubei. It was a competitor of Dianping (China’s ‘Yelp’), which Meituan eventually acquired. But Alibaba failed to capitalise on the market opportunity and eventually discontinued Koubei. Now, Alibaba is integrating the idea behind Koubei’s reference functionality into Amap, acquired by Alibaba in 2014, calling it ‘Amap Street Stars’. It offers rankings, ratings and reviews of businesses.

In recent years, Chinese consumers have become increasingly sceptical of online reviews, which are often found to be fake. Amap Street Stars therefore, takes a different approach and its rankings will be primarily based on ‘hard data’ of actual consumer behaviour:

▶️ the number of people visiting a store

▶️ the repurchase rate (‘Returning Customer Ranking’)

▶️ the number of people travelling to a store from 5-10 km (‘Tire Wear Ranking’)

▶️ the regional background of visitors; whether customers are locals or from all over the country (‘Local’s Favourites Ranking’).

This type of behavioural data is difficult to manipulate.

Amap assembled a team of scientists, primarily graduates from Peking University and Tsinghua University, in early June. They used AI products based on Alibaba’s Tongyi Qianwen model to gain insights into user behaviour data.

Amap used 1.2 billion navigation trips and 22.8 billion kilometres travelled by 51.43 million people in the past year, and screened from 1.18 million stores that users have repeatedly visited.

User-generated content, including images, text, and videos, is used as a secondary data source. Amap incorporated Alibaba’s Qianwen's multimodal capabilities to identify and select authentic, high-quality content. It also uses Alipay’s Sesame Credit to weigh reviews based on the scores of users (no, this is NOT the social credit system!) and AI to filter out fake reviews. Both the behavioural and UCG data ultimately flow into the team's developed AI algorithm system, generating dynamic ratings and rankings. Amap also uses AI programming and AIGC tools.

Amap Product Manager Li Gang proposed the project in May. In June, Alibaba separated 20 Amap staff members to a secluded building on the Xixi campus. By September, the team had grown to over 100 people, and the product was launched mid-September, in time for China’s National Day Golden Week public holiday.

AMap covers over 7 million restaurant locations nationwide, with 120 million daily searches related to lifestyle services and navigation to 13 million lifestyle service destinations. Amap Street Star already covers over 1.6 million offline service providers in more than 300 Chinese cities, including over 870,000 restaurants, 230,000 hotels, and nearly 50,000 scenic spots.

Alibaba claims that the new functionality won’t be commercialised. This is remarkable since Amap has lost 1 to 2 billion yuan each year and has only recently become profitable. With 890 million MAU (and 170 million DAU), Amap is the 4th most used app in China after WeChat, Taobao and Alipay and surpassing Douyin (the Chinese version of TikTok). Amap’s income primarily comes from advertising (60-70%); the rest comes from aggregation of ride-hailing and other local services.

Meanwhile, Dianping’s income comes from advertising by merchants (paid rankings, brand zones), transaction commissions for group-buying and take-out (3-10%), paid value-added services memberships, and enterprise data services.

Amap already integrated travel services (taxis, shared bikes, flights, etc) and now incorporates services like group buying, hotels & travel, leisure & entertainment, and errands. Alibaba has previously conducted group-buying on Taobao, Alipay, and Eleme. There are, however, no current plans to merge Amap and Taobao, and Eleme is also testing in-store business in some cities separately.

With Alibaba’s strong moves into instant retail and meal delivery, Amap could become an important entry point for those services. Amap can stimulate in-store business for restaurants that do meal delivery on Eleme, but also those that don’t offer such services.

As we reported at Tech Buzz China, Meituan has previously been under attack of Douyin in the in-store business. With its 800 million DAU, Meituan captured RMB 500 million of this market in 4 years and has a target of RMB 900 million this year. However, Douyin’s impulse-triggered sales of in-store group-buying vouchers see much lower redemption rates than Meituan’s search-based sales. With Alibaba’s Amap approach, it might be more successful than Douyin.

Alibaba will offer over CNY1 billion (USD140.4 million) of incentives to support the spending of 10 million customers on car rides, dining, and other services. Meanwhile, lifestyle app Xiaohongshu has become the default app for Chinese to search for user-generated information. It is now also trying to commercialise its users’ search behaviour for in-store and travel content (420 million searches per month).

With attacks on both flanks, meal delivery/instant retail and in-store (Meituan’s cash cow), it looks like Alibaba is out for Meituan’s blood. And considering Alibaba’s cash reserves, Amap usage numbers and potential for further integration of Amap with the Alibaba ecosystem, Meituan might need to be more worried than it was when Douyin entered the sector.

-Ed

Sources: LatePost,  Kazuji New Retail, Yicai, Tech Planet, Huxiu.