Baidu Conference Call: 11% of Baidu search results are generated by AI, Luobo Kuaipao will achieve UE breakeven, advertising business growth may be weak

Before the U.S. stock market opens on May 16th, Baidu released its financial report for the first quarter of 2024. The report shows that Baidu's revenue in the first quarter was 31.5 billion RMB, a slight increase of 1% year-on-year, slightly exceeding the expected 31.43 billion RMB.

The net profit attributable to Baidu in the first quarter (Non-GAAP) was 7.011 billion RMB, a year-on-year increase of 22%. The net profit attributable to Baidu's core (Non-GAAP) was 6.628 billion RMB, a year-on-year increase of 26%, both exceeding market expectations. The core operating profit margin (Non-GAAP) was 23.5%.

During the subsequent conference call, Baidu's founder, chairman, and CEO, Robin Li, stated that with the expansion of autonomous driving operations and continuous improvement in cost structure, hardware depreciation costs will be significantly reduced this year. With the continuous improvement in operational efficiency and cost reduction, Luobo Kuai Pao will be the first to achieve UE breakeven in the Wuhan area in the near future.

Currently, 11% of search results on Baidu are generated by AI, making search more accurate, organized, and direct in answering user questions. The AI reconstruction work on Baidu Search is still in its early stages, and overall, search is most likely to become the Killer App of the AI era.

The following is a summary of the analyst Q&A session during the conference call:

Q1: Has AI technology helped Baidu improve its advertising monetization rate? In which areas do advertisers see improvements? Are there any areas that can be further strengthened?

Answer: Baidu's advertising monetization is the first business to benefit from AI, generating hundreds of millions of RMB in incremental revenue each quarter since the second half of last year. Baidu has been using the Wenxin large model upgrade monetization system to enhance various aspects of advertising technology, including improvements, feature additions, refining delivery systems, automated creative generation, and adding strategic information for advertisers. These improvements have led advertisers to increase their spending on Baidu, with incremental AI-related advertising revenue growing quarter-on-quarter in the first quarter, and this trend is expected to continue.

Incremental revenue has helped Baidu alleviate broader macroeconomic weakness, allowing advertisers to build their online image and interact with potential customers through multi-round conversations using natural language. With the introduction of Agent Builder, advertisers can easily create customized AI assistants. When advertisers express their intentions to AI assistants, they can more effectively achieve their goals. Agents also help enrich Baidu's content and improve the user experience on Baidu.

Although still in the early stages, some Agents have already helped some advertisers achieve better ROI. For example, an online education client used Agent Builder to create its AI agent, injecting key insights such as product introductions and subject matter expertise, and providing continuous feedback for Baidu's improvement. This Agent significantly enhanced the company's online customer service by providing round-the-clock, high-quality consultations, increasing ad conversion rates by 20%.

AI assistants will be the main form of content and services in the new AI era, and Baidu will continue to increase profitability, improve advertiser conversion rates and ROI, and promote an increase in transactions directly generated on the Baidu platform. This transformation should help Baidu shift from the traditional CPC model to a more effective CPS model

Q2: How do price reduction measures by some of Baidu's peer companies affect its AI cloud business? With intensifying competition, what is the profit potential of Baidu Cloud? What should be the level of sustainable growth for customers?

Answer: GAI and large models are transforming the cloud computing industry from general computing to AI computing. Baidu AI Cloud revenue went from a year-on-year decline in the third quarter of last year to an 11% growth in the fourth quarter of last year, accelerating to a 12% growth rate in the first quarter of this year. In the first quarter of this year, revenue from GAI and basic models accounted for 6.9% of AI Cloud revenue. Traditional cloud businesses are also leveraging GAI and basic models for support.

This quarter, Baidu continues to focus on achieving high-quality revenue growth by cutting low-profit businesses. The market is still in a very early stage regarding GAI and basic model businesses. Baidu's focus remains on the education market, expanding its influence in more enterprises. It is expected that AI-related profit margins will further increase and surpass traditional cloud businesses.

Changes in competitors' pricing policies and adjustments to the pricing of certain products by cloud providers are actually quite common. This is a trend we have observed many times in the past. Baidu's cloud products have expanded from traditional CPU clouds to high-value AI products and services. Therefore, industry changes have minimal impact on AI cloud business development compared to CPU cloud pricing.

Q3: Given the chip shortage, how is Baidu maintaining its delivery commitments, differentiation value, and enhancing its leading edge in large models?

Answer: Baidu is optimizing model costs and performance through a unique 4-layer AI architecture stack to ensure that customers and developers can easily build applications using tools like Agent Builder, App Builder, and Model Builder.

Baidu recently made a breakthrough by integrating GPUs from different suppliers into a large computing cluster, enabling efficient model training and inference using less advanced chips. Baidu's deep learning framework Pedal has helped reduce model training costs through continuous innovation and enhancement. Pedal is compatible with over 50 different chips, and its developer community has grown to 13 million.

Baidu makes AI more accessible and affordable through lightweight LLM and model development toolkits, and is confident in becoming a leader in the Chinese AI ecosystem.

Q4: What are the main drag factors for advertising growth compared to peers? What trends are seen in advertising budgets and advertiser sentiment, especially in the second quarter, April, and May, and how should normalized advertising growth for 2024 be considered?

Answer: Baidu's online marketing revenue grew by 3% year-on-year in the first quarter, with a weak macroeconomic environment leading to softness in the advertising business. Advertisers mainly come from small and medium-sized enterprises in various industries and are highly sensitive to the macro environment. In the first quarter, sentiment among advertisers in verticals such as real estate and franchising remained weak. As we enter the second quarter, no improvement in advertiser sentiment has been observed, and online marketing revenue growth is expected to be soft. Despite facing challenges, online marketing is expected to remain Baidu's main business

Q5: Can the management provide updates on the development, strategic initiatives, and geographical coverage of the autonomous driving taxi service for this year? Additionally, can you explain the strategy for continuously optimizing the operational unit economics (UE) to achieve breakeven in Wuhan as mentioned earlier? What is the expected fleet size for this year? How will these measures impact costs?

In 2023, Apollo has made significant progress in improving the operational unit economics in key cities. Let me illustrate how we have achieved this at Apologos, our largest operational base in Wuhan.

Since the commencement of commercial operations in 2022, our operational unit economics have continued to improve, primarily due to the expansion of the autonomous driving service scale and the reduction in per-vehicle costs.

Currently, the fully autonomous driving fleet in Wuhan has grown to 300 vehicles, triple the number from a year ago. The operational area and service hours have significantly expanded due to government recognition of our technology, covering a population of over 7 million. Our operating hours have been extended to provide 24-hour service. These expansion initiatives have increased the daily passenger trips and travel distances per vehicle, thereby boosting revenue.

In terms of cost control, by demonstrating a consistent safety record, we have increased the deployment of fully autonomous driving services. The proportion of fully unmanned driving orders has risen to 70% and is expected to reach 100% soon, which will help us reduce costs related to safety personnel.

Simultaneously, the large-scale production of our sixth-generation autonomous driving taxi, RT6, is underway, with an expected cost per vehicle (excluding the battery) of less than $30,000. RT6 will become the primary model for our future fleet expansion, helping to reduce per-vehicle depreciation costs, further enhance operational unit economics, and drive us towards profitability.

Looking ahead to this quarter, we plan to expand the fully autonomous driving fleet in operation in Wuhan to 1,000 vehicles by the end of the year, more than triple the number from the end of last year. Our focus remains on improving regional unit economics, reducing losses per Apologo vehicle, while continuously enhancing operational efficiency and lowering costs. We believe Apollo will first achieve breakeven in operational unit economics in Wuhan. Once achieved, we will be able to rapidly scale up operations