Zhike | Baidu Adheres to AI Strategy and Is Prepared for a "Protracted War"

On November 21, 2024, Baidu Group released its financial report for the third quarter of 2024 (2024Q3).

Data shows that in 2024Q3, the company recorded operating income of 33.6 billion yuan, of which the core income was 26.5 billion yuan, and the core net profit was 7.5 billion yuan, with a year-on-year growth of 17%, exceeding market expectations.

Overall, in the third quarter of 2024, Baidu's "AI content" has further increased. Currently, AI has become the engine driving Baidu's long-term growth, not only driving the steady growth of the cloud business, but also accelerating the high-quality transformation of traditional businesses.

 Overview of Third Quarter Results

This quarter, Baidu achieved operating income of 33.6 billion yuan. Among them, Baidu Core achieved operating income of 26.5 billion yuan, the same as the previous year. Further splitting Baidu Core, the company's online marketing income within the quarter was 18.8 billion yuan, a year-on-year decline of 4%; non-online marketing income was 7.7 billion yuan, a year-on-year increase of 12%, mainly due to the growth of the intelligent cloud business. In 2024Q3, the revenue of the intelligent cloud business reached 4.9 billion yuan, with a year-on-year growth of 11%.

In terms of profits, in 2024Q3, the company achieved an operating profit of 5.9 billion yuan, of which the core operating profit was 5.7 billion yuan, corresponding to a core operating profit margin of 21%; in the same period, the net profit attributable to the parent company was 7.6 billion yuan, with a year-on-year growth of 14%. Among them, the core net profit was 7.5 billion yuan, with a year-on-year growth of 17%, corresponding to a core net profit margin of 28%.

Overall, Baidu's third-quarter report, against the backdrop of the overall pressure on the industry, performed beyond market expectations. It is particularly worth noting that the intelligent cloud business continues to exert its strength and begins to act as a new performance engine. At the same time, the remodeling of traditional businesses by AI also lays the foundation for Baidu's long-term, high-quality and sustainable development.

In fact, since Baidu announced the AI reconstruction of its business in 2023, the focus of the capital market's attention on Baidu has shifted more to its progress in the field of AI, rather than simply the financial data performance.

In the conference call after the release of this financial report, Baidu mentioned that as of November, the call volume of the Wenxin Model reached 1.5 billion, and the daily processing of Tokens text exceeded 1.7 trillion. Compared with the daily call volume of 50 million disclosed for the first time a year ago, it has increased by 30 times; meanwhile, in September this year, the monthly active users (MAU) of the AI function of Baidu Wenku exceeded 50 million, with a year-on-year growth of more than 300%; the AI function of Baidu App has covered nearly 70% of monthly active users, and more than 20% of the search result pages contain AI-generated content. These data all indicate that the AI content of Baidu's business has further increased.

The Initial Effects of the Long-Term AI Strategy

As mentioned earlier, in this quarter's financial report and conference call, three sets of numbers disclosed by Baidu deserve special attention: the call volume of the Wenxin Model reaching 1.5 billion, the MAU growth of the AI function of Baidu Wenku by 300%, and the coverage of the AI function of Baidu Search reaching nearly 70% of monthly active users and more than 20% of the search result pages containing AI-generated content.

Why are these three sets of numbers so important? Because they are directly related to Baidu's long-term strategy.

So, what exactly is Baidu's long-term strategy?

After proposing a comprehensive shift to an "artificial intelligence-first" strategic direction half a year ago, In this financial report, Robin Li once again emphasized that Baidu's long-term strategic focus is the improvement of artificial intelligence capabilities, which is also the cornerstone for promoting its transformation to AI.

In this vigorous AI revolution, many large overseas companies have focused their strategic priorities on the disruption of new technologies, thus neglecting the implementation of technology in the application end. But in the long run, the key to whether AI can truly achieve commercial value lies in whether the technology can empower applications to produce different chemical reactions.

Different from the strategies of large overseas companies, Baidu is forming a core strategy with AI as the long-term strategy and application-driven as the main implementation path. The essence of this strategy is to maximize the demand-benefit, and to better utilize the capabilities of the large model by achieving the greatest fit between products and technologies. In simple terms, it is to take the user's usage needs as the foothold, and on this basis, to promote technological progress through continuous optimization of the model capabilities.

Robin Li also mentioned in a recent interview, "The first principle of Baidu's strategic choice is application-driven. We hope to come into contact with scenarios and applications as early as possible to see what problems we encounter in this process. The problems that most people encounter are the problems that we prioritize to solve."

Judging from the information disclosed in this financial report and conference call, the results of this core strategy have begun to emerge initially.

First, In the past year, the call volume of the large model as the base has shown an exponential growth trend. As of November, the daily call volume of Baidu's Wenxin Model has exceeded 1.5 billion, and the daily processing of Tokens text exceeds 1.7 trillion. Compared with the daily call volume of 50 million disclosed for the first time a year ago, the call volume has increased by 30 times, which represents the explosive growth of large model applications both inside and outside Baidu. In the conference call, Baidu also disclosed that the wide application of AI has driven the growth of API call volume, especially in products such as search and Wenku. In addition, external API calls are also growing rapidly, especially in fields such as online education, social media, and catering services;

Second, With the continuously improving AI capabilities, Baidu's "old" applications are rejuvenated. Taking Baidu Wenku as an example, it is achieving transformation through AI reconstruction. In September this year, The monthly active users (MAU) of the AI function of Baidu Wenku exceeded 50 million, with a year-on-year growth of more than 300%, and the users' willingness to pay is continuously increasing. The subscription income in the third quarter increased by 23% year-on-year. At Baidu World 2024 held recently, Baidu Wenku and Baidu Netdisk jointly created the tool-type intelligent agent "Free Canvas", which attracted more than 200,000 people to queue for reservations on the first day of its launch.

Finally, This quarter, Baidu Cloud's revenue increased by 11% to 4.9 billion yuan, and the proportion of AI revenue increased to more than 11%, building Baidu's second growth curve. According to the conference call, in the third quarter, the growth of Baidu Intelligent Cloud was mainly driven by the high demand for model training and reasoning in industries such as the Internet, education, and finance. Among them, the incremental income from mid-sized enterprise customers increased by 170% quarter-on-quarter. According to the IDC report, in the Chinese AI public cloud market, Baidu Intelligent Cloud has ranked first for five consecutive years.

 Intelligent Agent + Search, Further Opening Up the Imagination Space

In recent years, under the general backdrop of the overall weakness of the industry, the search business is facing certain growth pressure, but as Baidu's fundamental business, the importance of search has never decreased. According to the disclosure of the conference call, At present, the AI function of Baidu APP covers 70% of monthly active users, and more than 20% of the search result pages on Baidu Search contain AI-generated content. This is a more future-oriented data. In contrast, Google's current coverage is less than 50%.

Baidu's various actions in AI applications indicate that Search will likely become the first entry point of its new AI ecosystem in the future.

Baidu has repeatedly emphasized that the intelligent agent is the best form of AI application. Simply put, an intelligent agent is an application autonomous program that can realize multiple AI functions, and at the same time, various applications can work collaboratively to maximize the satisfaction of user needs. For example, the no-code tool "Miaoda" recently launched by Baidu is a combination of the large model and the intelligent agent. Based on the thinking, planning and other capabilities of the Wenxin Model, Miaoda can realize the scheduling and arrangement of different types of intelligent agents, thereby achieving intelligent agent collaboration.

It is not difficult to see that under the users' demand for multi-scenario intelligent agents, search naturally has the ability to connect multi-dimensional supply and demand. Therefore, Baidu Search has become the best entry point for intelligent agents. As Baidu said in the conference call, "No matter when users have needs, we can provide them with unrestricted, on-demand personalized content by combining multiple formats."

In the previous financial report disclosures, Baidu often only disclosed the proportion of AI-generated search results, but this financial report disclosed the coverage rate of AI applications. The change in the disclosure caliber also represents Baidu's attitude towards the future AI, that is, the AI function is not only AI generation, but a larger proportion is the "intelligent agent".

Therefore, with the continuous construction of Baidu's intelligent agent ecosystem in the future, search can also benefit from the explosion of AI technology and return to the fast growth track. In the new ecosystem of intelligent agent + search, the imagination space of Baidu's fundamental business is further opened up.

From the perspective of the secondary market valuation, the current PE-TTM of Baidu Hong Kong stocks is only slightly over 10 times, with the percentile at 3.43% in the past three years and 18.66% in the past five years, which is obviously at a low position. Standing at the critical time window of Baidu's comprehensive AI transformation, the market obviously has not fully priced in the AI capabilities and the remodeling effect of future AI on traditional businesses. Considering the above factors, the current Baidu may have reached the best time point for long-term investment.

Figure: Baidu's PE-TTM  Source: Wind, Compiled by 36Kr

*Disclaimer:

The content of this article only represents the author's views.

The market is risky, and investment should be cautious. Under no circumstances does the information in this article or the opinions expressed constitute investment advice to anyone. Before making an investment decision, if necessary, investors must consult a professional and make a cautious decision. We do not intend to provide underwriting services or any services that require a specific qualification or license to engage in for the trading parties.