36Kr jointly releases the second edition of "2026 Consumer Brand AI Recommendation Power Roster" with PureblueAI Qinglan
The entry point for consumer decision-making is changing.
In the past, when people purchased a product, they often relied on brand awareness, recommendations from friends, content reviews, or actively searched for answers in the search box.
But today, this process is being rewritten by AI.
"How to choose a new energy sedan priced between 200,000 and 300,000 yuan?" "What are the anti-aging serums suitable for oily skin?" "What mobile phone to buy with a budget of 5,000 yuan?" More and more specific consumer questions are being directly handed over to AI for answers.
According to QuestMobile data, as of June 2026, the monthly active users of Doubao, Tongyi Qianwen, and DeepSeek have reached 382 million, 167 million, and 130 million respectively. In addition, the "2026 Q2 China Consumer Willingness Survey Report" shows that in a survey covering 5,000 consumers in urban and rural areas, nearly 80% of respondents believe that AI has affected their consumption decisions.
The entry points are shifting, and the rules are also being redefined.
In May this year, 36Kr jointly launched the first "2026 Consumer Brand AI Recommendation Power List" with PureblueAI, making the first attempt to observe the performance of consumer brands in AI decision-making and recommendation scenarios.
The first phase of the study found that whether a brand can be included in AI recommendation results depends not only on market awareness, but also on whether its product value can be accurately understood by AI and matched with the specific needs of consumers.
In other words, brand competition is shifting from competing for consumer awareness in the past to competing for priority selection in specific consumption scenarios. What questions consumers raise, whether the product value of the brand is clearly expressed, and whether user reviews are effectively accumulated are becoming important factors affecting AI recommendation results.
Following the release of the first phase, 36Kr once again cooperated with PureblueAI to launch the second phase of the "2026 Consumer Brand AI Recommendation Power List", covering mainstream AI platforms such as DeepSeek, Doubao, Tongyi Qianwen, and Tencent Yuanbao. It is obtained through multiple rounds of collection and evaluation around 19 subdivided consumption intentions in five major consumer categories: mobile phones, home appliances, automobiles, skin care, and beauty.
On the basis of continuing the evaluation system of the first phase, this issue has made two major core upgrades: first, the observation scope has been further expanded, more subdivided consumption intention scenarios have been added, and brand performance has been observed from more specific questions; second, comparison of ranking changes with the shortlisted brands in the first phase has been added to further track the dynamic changes of brands in the AI recommendation environment.
Core Insights
After observing the two consecutive lists, we found that the change brought by AI recommendation is not simply adding a new traffic entry, but changing the evaluation system of brand competition.
In the traditional business logic, brand competition is more focused on cognitive scale, channel coverage and marketing volume. But after entering the AI decision-making scenario, consumers' questions become more specific, and brand competition has also shifted from "big brand competition" to "competition under specific needs".
Judging from the current list and related data, three changes are worthy of attention.
First, brand competition is moving from "overall cognition" to "scene matching".
What consumers ask AI is usually not "which brand is the best", but "what to buy under a certain budget" and "what to choose under a certain demand". This means that in AI answers, the degree of matching between products and specific needs is presented more directly.
Looking at the shortlisted brands in the list, the same brand may have advantages in one consumption scenario and perform ordinarily in another. Therefore, the key to future brand competition is not only to establish overall awareness, but also to form clear labels in specific consumption intentions one by one.
Second, AI recommendation is amplifying the importance of product expression capabilities.
In the traditional consumption environment, brands can rely on long-term accumulation to form cognitive advantages; but in the AI environment, whether product selling points are clear, parameter information is complete, and user feedback is fully accumulated may affect the presentation method and recommendation performance of brands in AI answers.
This also explains to a certain extent that brands need to re-examine the construction of their own information assets, not only to produce content, but also to enable product value to be understood, extracted and disseminated by machines.
Third, AI recommendation is forming a dynamic competition system.
Judging from the changes in the two lists, some brands have entered the forefront under specific consumption intentions, and some previously leading brands have experienced ranking changes. This shows that AI recommendation has not yet formed a solidified pattern, which means both challenges and opportunities for re-competition for brands.
In our view, when AI becomes a new decision-making entry for consumers, the capability that brands really need to build is not just "being remembered by consumers", but "being accurately recommended by AI when consumers need it".
The following is the complete list and interpretation content of this issue.
The second phase of "2026 Consumer Brand AI Recommendation Power List" and Interpretation Analysis
Automobile | 200,000-300,000 yuan new energy sedan recommendation
In the intention of "recommendation for new energy sedans priced at 200,000-300,000 yuan", Xiaomi won the first place with 63.77 points and continued to maintain its lead, but the point difference with the second place was significantly narrower than that in the first phase. Zeekr and BYD rose to the second and third places respectively, with recommendation power scores of 52.09 points and 51.64 points, the difference between the two is only 0.45 points; Tesla dropped from the second place to the fourth place, and the competition among leading brands in the AI recommendation scenario has further intensified.
Compared with the first phase, the rankings of mid-tier brands have been greatly adjusted. XPeng and AVATR remain at the 5th and 6th positions respectively, IM rose from 10th to 7th; Seres, BMW i and Hyundai entered the top ten for the first time in this issue, while Luxgen, Voyah and NIO dropped out of the top ten, reflecting that the AI recommendation pattern in this price range is still dynamically changing.
On the whole, the current list is shifting from the "one brand leading obviously" in the first phase to "multi-brand joint competition". With the improvement of the recommendation power of brands such as Zeekr and BYD, the gap between leading brands is narrowing. Whether a brand can continue to appear in relevant consumer decision-making questions, maintain a leading recommendation position and stable product expression is becoming an important factor affecting its AI recommendation performance.
Automobile | 7-seat new energy MPV model recommendation
In the intention of "7-seat new energy MPV model recommendation", Wey Gaoshan ranks first with 38.60 points, BYD Xia ranks second with 29.48 points, and Leapmotor, XPeng X9 and Buick Electra series rank third to fifth respectively. However, the recommendation power scores of the top ten brands did not exceed 40 points, and the overall performance was relatively low; except for the first place, the scores of other brands were relatively concentrated, and no prominent AI recommendation advantage has been formed yet.
"7-seat new energy MPV model recommendation" is a high-traffic automobile consumption question in the 30 days of June, and its attention may be related to the rising demand for summer family travel, multi-person riding and long-distance self-driving. As more users filter models through AI, space, battery life, riding comfort and intelligent experience have also become key comparison factors in relevant decisions.
On the whole, although 7-seat new energy MPVs have attracted the attention of users, the recommendation pattern of each brand in AI answers has not been fully established. Whoever can present the family travel value and differentiated advantages more clearly and stably will have a better chance to gain an advantage in subsequent AI recommendations.
Automobile | 200,000-300,000 yuan SUV recommendation
In the intention of "200,000-300,000 yuan SUV recommendation", Li Auto ranks first with 73.84 points, leading BYD, the second place, by 23.25 points, forming a relatively obvious leading advantage. BYD ranks second with 50.59 points, and IM and XPeng rank third and fourth with 32.20 points and 30.62 points respectively, forming the second echelon.
It is worth noting that this intention does not limit the energy type, but all the top ten brands are brands that mainly focus on new energy models, reflecting that AI shows a clear new energy tendency in the recommendation of 200,000-300,000 yuan SUVs.
From the perspective of score distribution, Li Auto and BYD have a large gap with other brands, and the scores of IM and XPeng are relatively close; starting from the fifth place, the scores drop further, and the mid-tier brands have not yet formed obvious advantages. Except for leading brands, most brands have not formed prominent recommendation advantages, and there is still much room for subsequent ranking changes.
Mobile phone | 2000-4999 yuan high-performance mobile phone recommendation
In the recommendation of "2000-4999 yuan high-performance mobile phones", iQOO ranks first with 38.75 points, followed by Redmi with 37.99 points, the difference between the two is only 0.76 points, and the head competition is relatively close. Xiaomi ranks third with 29.76 points, 8.23 points behind the second place, and the list shows a clear stratification starting from the third place.
From the overall distribution, the scores after the fourth place gradually narrow, among which the 5th to 8th places are concentrated between 17.48 and 20.40 points, and the competition of mid-tier products is relatively close. The top three in the list appear in brand names, and the 4th to 10th are mainly specific models; Honor has 3 products in the top ten, and OnePlus has 2 products on the list. The overall list presents a pattern of "two strong heads close, mid-tier competition concentrated".
Mobile phone | 2000-4999 yuan gaming phone recommendation
In the recommendation of "2000-4999 yuan gaming phones", iQOO Neo11 ranks first with 33.05 points, leading the second place iQOO 15 by 4.33 points, forming a clear single product advantage. The competition for the 2nd to 4th places is very tight. The scores of iQOO 15, Redmi K90 Ultra and RedMagic 10S Pro are 28.72 points, 28.42 points and 28.25 points respectively, with a maximum gap of only 0.47 points, indicating that AI platforms have relatively similar cognition of the gaming performance of these products.
From the perspective of recommendation power scores, iQOO, OnePlus and Redmi related products occupy 8 of the top ten seats, of which 3 iQOO products are on the list and 3 OnePlus products are also in the top ten, showing strong product matrix recognition. However, OnePlus products are mainly distributed in the 6th to 10th places, and there is still a certain gap with the flagship models. On the whole, AI platforms focus more on specific models with distinct performance labels in gaming phone recommendations, and the head competition mainly revolves around flagship chips, game frame stabilization, heat dissipation capacity and battery life performance.
Mobile phone | 500-1999 yuan cost-effective mobile phone recommendation
