
As Double 11 is coming to an end, Doubao has finally broken through the thin layer of window paper between Douyin and Douyin.
Recently, the alphabet list found thatDoubao began to actively recommend products based on user preferences.After entering the “Double 11 Dou Bao” entrance, the page is no longer dominated by preset questions, but has been upgraded to product recommendations based on user preferences. The products appear in a double column, the same as the Douyin mall model, and live broadcast room recommendations also appear.

To put it simply, Doubao began to actively recommend products and gave the mall a secondary entrance.It is equivalent to moving Douyin Mall into Doubao.
Previously, Alphabet had judged in “In order to bring goods, Doubao spoke first” that the e-commerce exploration of Doubao still needs to follow the business logic of Douyin, discover needs based on user interests and usage scenarios, and then complete transactions through product recommendations.
Today, this logic becomes clearer on bean bags. This means,If Doubao wants to make money, it ultimately cannot avoid the mature business logic of the mobile Internet.
A
Doubao gives the “mall” a secondary entrance.
After entering the “Double 11 Dou Bao” entrance, the page shape is close to a small Douyin mall.The top is for coupon collection, and the bottom is the product recommendation information flow. It is similar to the model of shelf e-commerce. Swipe up and down to instantly view more products. The bottom integrates e-commerce functions such as shopping cart and order checking.
Previously, when entering this entrance, there were coupon collection at the top, preset scenes such as “choose according to budget” and “product comparison” in the middle, and preset questions at the bottom.

The most noteworthy thing this time is that Doubao began to actively recommend products based on users’ preferences. Page tips:“I picked these based on your preferences.”
When refreshing the interface multiple times, some prompts will also appear in the upper right corner: “Shop around more and I will know you better” and “Looking for products you are interested in.”

Alphabet Ranking’s actual measurement found that part of Doubao’s so-called “according to your preferences” comes from user searches in Doubao. For example, Alphabet asked me a lot about laundry detergents in Doubao, and Doubao Mall recommended relevant laundry detergents to me.
Another part comes from Douyin, which is related to users’ browsing, searching and purchasing behaviors on Douyin.
Recently, Alphabet browsed some videos on Douyin of consumers buying wooden beds in the live broadcast room of Youhui. It also saw that the live broadcast room was selling wooden beds in the live broadcast of Youhui on November 8. Subsequently, a large number of bed products appeared on Doubao’s product recommendation page, some of which were brands that appeared in the live broadcast room with Hui.
This correlation in recommendations at least shows that,Doubao will combine the information users browse on Douyin to recommend brands or products in a personalized way.
Alphabet asked Byte for verification on this matter, but as of press time, it has not received a reply.
During Double 11, Doubao began to push e-commerce coupons to some users and actively popped up small cards to introduce its e-commerce capabilities. At that time, Alphabet predicted that it would not be impossible for Doubao to push multiple products to users. If Doubao is expected to use personalized recommendation algorithms to maximize the value of traffic like Douyin, it is just a matter of breaking the window paper. Now, this layer of window paper has finally been broken.
In addition to product recommendations, there are two other things worth noting about the product flow of bean bags.
首先,Doubao embeds product recommendations and live broadcast rooms in the secondary entrance mall.When users browse products, they will be directed to live broadcast rooms such as Walking with Hui and Lanzhi Chunxu. They can either purchase directly or learn more about the products through live broadcasts.
If the user wants to reduce the recommendation of a certain product, he can long press the product or the corresponding live broadcast room, click “dislike”, and select options such as “recommendation is not relevant”, “repeated recommendation”, “price is more expensive than other platforms”, etc. When the product card disappears, the system will pop up a prompt saying “Feedback is successful and the recommendation results will be optimized for you”.
Compared with Douyin Mall, Doubao Mall goes one step further in this regard. Among them, options such as “recommendation is not relevant” and “price is more expensive than other platforms” are not currently provided by Douyin Mall.
Note: Doubao is on the left, Douyin is on the right
其次,Doubao adds an AI question and answer entrance to the product.
Long press a product on the product recommendation page, and the “Question” option will appear.For example, if you press and hold a solid wood dining table and click to ask, some preset questions will appear in the search bar at the bottom, including the dining table’s material, load-bearing capacity, applicable scenarios, etc. Users can also directly enter content they care about without using preset questions.
After submission, a bean bag chat box will pop up on the page to provide further introduction to this product.
Overall, Doubao looks like it has been moved over from Douyin Mall, but it goes one step further than Douyin Mall.
B
Doubao has become more aggressive in brokering e-commerce transactions.
In the past period of time, Doubao’s actions in e-commerce have been relatively restrained.
Initially, Doubao was connected to Douyin Mall, and the purchase link of Douyin Mall was attached when recommending products, but the transaction still needed to be redirected to Douyin. Later, the payment function was improved, and purchases could be completed within Doubao without having to jump to Douyin. On the occasion of June 18th this year, the bean bag was launched on the online shopping portal “Ask Before Buying Bean Bao”.
Before the start of Double 11, Doubao embedded live broadcast room recommendations in answers, and users could jump from the AI conversation to the live broadcast scene.
Nowadays, Doubao has taken another step forward, aggregating a mall internally, actively recommending products and live broadcast rooms based on user preferences, and at the same time integrating e-commerce functions such as product flow, shopping cart, and order inquiry.
This is a rather radical move. This means that the commercialization of bean bags is undergoing a change:From waiting for users to put forward shopping needs and then recommending products, to actively discovering user needs and recommending products to users.
說激進,What’s more, this is the only large native AI model in China.
Although Qianwen has connected with Taobao to achieve jump-free shopping, it has also embedded “Double 11 discounts” in the navigation bar of the first-level entrance. However, this entrance still displays coupons, preset scenes, and preset questions, and does not actively recommend products. In addition, Qianwen has not connected with Taobao’s live broadcast room, and there is no live broadcast room recommendation when recommending products.
Not to mention Yuanbao, its product recommendations and transactions still rely on third-party e-commerce platforms, and currently transactions still need to jump to WeChat and be completed within the WeChat mini program.
Therefore, Doubao is one step ahead in actively recommending products and setting up a secondary entrance for the mall. On June 18 this year, Li Chengdong, the founder of Dolphin Club and an e-commerce expert, once analyzed to the alphabet list that for Doubao, user experience is the priority rather than commercialization. The rapid commercialization process may have to wait for Doubao to mature.
36Kr has also reported that Doubao may further integrate e-commerce function updates to improve the payment scenario in the third quarter of this year, and use subsidies to attract traffic to Douyin Mall, and enter the operational period in the fourth quarter.
Judging from the results, Doubao’s e-commerce actions are indeed more aggressive.
It’s not just the user side that has changed, the merchant side has also made efforts.According to DoNews, Doubao has recently added a “Doubao Operation” section for merchants, providing tools such as operating data and demand insights. The products on sale in Doudian are automatically synchronized with Doubao, enabling independent attribution of AI dialogue traffic and separate calculation of channel effects.
In other words, at the level of channel attribution and merchant operations within Byte,Doubao has been treated as an independent channel.
During this year’s Double 11, some media reported that the exclusive subsidy coupons of the Doubao platform cannot be used at the same time as the full discount coupons of the Douyin platform. There is a hard restriction behind this, that is, platform coupons are not interoperable.
This means that Doubao is becoming an independent “e-commerce portal.” But at the same time, Doubao is not yet an independent e-commerce “platform”. It relies on Douyin’s mature merchants and fulfillment network.
However, judging from various actions,Doubao is taking e-commerce as an important commercial direction.
It is not difficult to understand that large AI models have a large number of users. How to convert users’ searches and interests into actual consumption is a problem that this industry must face. Doubao chose to integrate Douyin’s mature e-commerce ecosystem and gradually increase its priority.
C
The method of making bean bags shows that at this stage,If Doubao wants to make money, it still needs to rely on the logic of mobile Internet.
目前,The business model of the AI native platform has not yet been fully developed.
Subscription membership, API calls, and enterprise services are all monetization methods that large AI model companies are exploring. But there is still a long way to go before large-scale profitability.
This can also be seen from the paying users of OpenAI. Three years after ChatGPT implemented the subscription system, the payment conversion rate was only 6.1%.
Even Anthropic has yet to achieve large-scale profitability. Previously, media reported that Anthropic had revenue of US$16.3 billion in the first half of 2026, but incurred a net loss of US$80.7 billion. Anthropic maintained operating profits in the second quarter of this year, but fell into losses again in the third quarter due to computing power purchases.
In contrast, mobile Internet already has a mature business model:First acquire users and traffic, and then monetize through advertising, e-commerce, etc.
OpenAI has already explored this traditional path to monetization, launched an advertising business, and ultimately achieved annualized revenue exceeding US$1 billion in 200 days. However, OpenAI claims that its advertising logic is different from traditional traffic advertising. ChatGPT’s advertisements are more context-dependent, such as what users are researching, comparing, and solving problems.
In other words, OpenAI still borrows the monetization framework of the mobile Internet, but the matching logic is replaced by “conversation context”.
For Doubao, it has even brought in Douyin’s e-commerce transactions. This choice is easy to understand.
Doubao is not a product that has completely grown out of the AI era. It is a product that has both mobile Internet traffic and AI capabilities.Byte does not need to wait for the AI industry to figure out a new business model before it starts thinking about how to make money. It can completely mobilize existing business resources first.
The e-commerce pie is much bigger. According to reports from brokerage industry experts quoted by Huxiu, Douyin’s GMV in the first half of 2026 will be approximately 2.4 trillion yuan.
Trading traffic for transactions is Byte’s core business.
Doubao has a huge AI user base, with monthly active users reaching 419 million. Douyin has mature product supply, merchant resources, live broadcast ecosystem and transaction infrastructure. Putting the two together, Doubao is responsible for understanding user needs and providing product recommendations, while Douyin is responsible for providing products and undertaking transactions.
This is much more realistic than trying to figure out a native AI business model from scratch.
There is also a user base between the two. Research by QuestMobile shows that as of May 2026, the overlapping users of Doubao and Douyin will reach 330 million. This means that there is already a considerable user overlap between the two products, and the ecological integration has a certain user base.
Of course, this does not mean that moving Douyin Mall into Doubao will definitely lead to success.
It still takes time to verify whether AI recommendations can truly understand consumer needs, whether the recommendation results can gain users’ trust, and how many actual transactions these recommendations can ultimately generate.
Related Reading
- Nvidia live updates: Jensen Huang ousted, CFO Colette Kress to serve as interim CEO2026-11-13
- Are mangoes not afraid of red fruits?2026-11-12
- Original: After Huang Renxun, who is most likely to take over Nvidia?2026-11-12
- Samsung’s roll-screen mobile phone is exposed, creating a new form ahead of Huawei and Apple, and folding screens are expected to be eliminated2026-11-12
- Musk’s favorite car is going to be discontinued.2026-11-12