Q2 2024 Earnings Summary
- Baidu's AI Cloud revenue grew by 14% year-over-year in Q2, driven by strong demand for generative AI and large language models (LLMs). Gen-AI-related revenue increased from about 5% of total AI Cloud revenue in Q4 last year to almost 9% this quarter. The company expects the AI Cloud revenue growth to maintain strong momentum, with the normalized margin for Gen-AI-related cloud business projected to exceed that of the traditional cloud business.
- ERNIE, Baidu's AI model, handles about 600 million API calls and generates over 1 trillion tokens daily, believed to be the highest in China. The developer community has grown to around 14.7 million, and hundreds of thousands of enterprises are engaged with Baidu. This application-driven approach and expanding ecosystem strengthen Baidu's leadership in the rapidly evolving AI landscape.
- Apollo Go, Baidu's autonomous ride-hailing service, is the world's largest in terms of the number of rides. With regional breakeven approaching, Apollo Go achieved 100% fully driverless operations in practically the entire city of Wuhan, operating a fleet of around 400 vehicles serving 9 million residents. Baidu's focus on cost reduction and innovation positions it to capitalize on the enormous potential of China's ride-hailing market, driving long-term sustainable growth.
- Baidu's advertising revenue is under pressure, facing challenges due to macroeconomic weakness, competition from leading short video and social media platforms, and the increasing amount of non-monetized Gen-AI content, which currently accounts for 18% of search results and is expected to grow. This shift may further impact ad revenue, as users spend more time on other platforms and advertisers reduce spending. ,
- Despite significant investments, Baidu's robotaxi service, Apollo Go, holds only about 1% market share in Wuhan, and scaling operations to achieve meaningful market share could take many years. The robotaxi business requires long-term investment and strong capital reserves, potentially impacting profitability due to high costs and slow adoption rates.
- The accelerated AI-driven renovation of Baidu's search may put pressure on profit margins in the short term, as new AI features require more computing power, increasing operating costs, while not yet being monetized. This could affect Baidu Core's profitability before monetization catches up. ,
-
Advertising Revenue Impact
Q: How is soft advertising revenue affecting Baidu?
A: Baidu's advertising business faces pressure from macro weakness and competition. Sluggish consumer spending has led advertisers, especially SMEs, to cut ad budgets, particularly in real estate, franchising, and automobile sectors. User behavior shifts towards leading short video and social media platforms have also impacted ad revenue. Additionally, 18% of search results now contain Gen-AI content, which is not monetized yet. While this may negatively impact revenue short-term, Baidu is focused on long-term growth by renovating its search with AI. -
AI Transformation of Search
Q: What progress has Baidu made in AI search?
A: Baidu is transforming its search capabilities using cutting-edge AI technologies and ERNIE. Since Q2, they've accelerated renovating search with Gen-AI to improve user experience and engagement. Currently, 18% of search results are provided by Gen-AI, offering more comprehensive and relevant information. New interactive functions allow users to refine questions through multi-round conversations. Baidu is also rolling out ERNIE agents within search results to provide intelligent assistance. -
AI Cloud Revenue Growth
Q: What's driving the growth in Baidu's AI Cloud revenue?
A: Baidu's AI Cloud revenue grew by 14% year-over-year, driven by strong demand for Gen-AI and LLMs across various industries. Gen-AI and foundation model-related revenue increased from 5% of total AI Cloud revenue in Q4 last year to almost 9% this quarter. The number of daily API calls surged from around 200 million in mid-May to over 600 million, generating over 1 trillion tokens daily. Baidu believes the normalized margin for Gen-AI-related cloud business should exceed that of traditional cloud business. -
Competition in AI
Q: How does Baidu view AI competition in the next 2–3 years?
A: Baidu expects fierce competition in AI over the next 2 to 3 years. They believe that companies who can monetize AI effectively will survive. Baidu maintains its leadership with a first-mover advantage and technological expertise, exemplified by their ERNIE models. ERNIE handles about 600 million API calls and generates about 1 trillion tokens daily, believed to be the highest in China. Baidu focuses on an application-driven approach and aims to make ERNIE increasingly accessible and affordable. -
AI Agents and Monetization
Q: How are ERNIE agents used and monetized?
A: Baidu has over 10,000 ERNIE agents on its platform, created by users, creators, service providers, advertisers, or companies. ERNIE agents improve user experience in Baidu's mobile ecosystem. Advertisers, especially in legal, education, and B2B sectors, are using ERNIE agents to enhance sales conversions. As ERNIE agents become smarter, they have the potential to drive direct sales, moving from a CPC model to a more effective CPS model. ERNIE agents could become the new website system in the AI era. -
Margin Outlook
Q: Is Baidu Core's margin expansion sustainable?
A: Baidu aims for high-quality growth balanced with investments. For AI Cloud, they continue to phase out low-margin businesses and expect normalized margins for Gen-AI-related cloud to be higher than legacy cloud business. The accelerated AI transformation of search may put pressure on margins short-term due to increased computing costs and unmonetized AI features. However, Baidu plans to reduce inefficient AI infrastructures and improve efficiency over time. -
Robotaxi Business Scaling
Q: How will Baidu scale its robotaxi business?
A: Baidu's Apollo Go is the world's largest autonomous ride-hailing service by number of rides. In Wuhan, with a fleet of around 400 vehicles serving 9 million residents, they achieved 100% driverless operation in practically the entire city. Their share of the ride-hailing market in Wuhan is around 1%, and scaling will be gradual, possibly taking many years. Baidu focuses on cost reduction through innovation and is open to various business models and partnerships for future development.