Cheetah Mobile, Inc. Q3 FY2024 earnings call

Service Robotics - CEO Fu Sheng mentioned that industry demand for service robots continues to rise, especially in overseas markets and various sectors like restaurants, hotels, factories and offices. In September, a new robot for factory and fulfillment center use was launched. LMs are used to improve robot interaction through voice, enhance autonomous driving, and add robotic arms. Visits to customers and partners in Europe and Southeast Asia are ongoing to strengthen the local distribution network. ### Internet Business - The legacy Internet business remained resilient, achieving solid revenue growth and margin expansion. CFO Thomas Ren noted the Internet business grew 26% year - over - year and 18% quarter - over - quarter, and operating margin (excluding share - based compensation expense) improved to 10% from 6%. The company also focused on cost management, such as consolidating teams and reducing certain costs, and decisively invested in AI, with non - GAAP R&D expenses increasing 25% quarter - over - quarter in Q3 and about 60% of revenues invested in R&D.

In the third quarter of 2024, Cheetah Mobile's total revenues increased by 16.6% year - over - year. The real - based service robots were a key driver of growth. The legacy Internet business grew by 26% year - over - year and 18% quarter - over - quarter. Non - GAAP gross profit rose 14% year - over - year and 7% quarter - over - quarter to RMB131 million, with non - GAAP gross margin expanding to 68% in Q3 from 65% in Q2 and 63% in Q1. Non - GAAP operating loss was RMB61 million in Q3, reduced from RMB63 million in Q2 and RMB66 million in Q1.

Robot Business - For 2025, specific goals are still in the process of formulation. The long - term goal is to have commission income account for more than half of the entire company within three years and be among the top few globally, at least one of the top three service system providers. ### Internet Business - No specific forward - looking guidance on revenue and profit margin trends for 2025 was clearly stated, but long - term goals were emphasized.

Q: What specific goals have been set for the robot business in 2025 in terms of shipment volume, revenue growth and revenue proportion?

A: For 2025, some specific goals are still being formulated. The long - term general direction is to look at a long - term goal, for example, to make the commission income account for more than half of the entire company within three years and be among the top few globally, at least one of the top three such service system providers. Detailed goals still need discussions and careful deductions.

Q: How should we predict the revenue and profit margin trends of the Internet business in 2025? Will it show a steady growth trend or gradually slow down and decline?

A: The future trend of the Internet business in 2025 was not explicitly stated as a steady growth or gradual slowdown and decline; long - term goals were emphasized.

Q: What is the subsequent pace and specific plan for reducing losses? Is there a clear timetable for achieving profitability?

A: Reducing losses is the top priority. Due to participation in large language model R&D, the pace of loss reduction slowed down a bit. But after review, efforts will be put into agent development and robot intelligence implementation, with R&D costs related to this being significantly reduced compared to past large language models. There is an internal loss reduction plan and a timetable for profitability, but specific details were not clearly disclosed externally due to market and technological changes.

Q: How do you view Agentic AI or AI agent? How difficult is this technology? Can AI agent accelerate the application of large models in the C and D ends? How should we consider the value distribution in this?

A: AI agent is a new software paradigm. It can accelerate the application of large models in C and D ends, especially in many C and D apps. However, for an AI agent to truly land, a lot of effort is needed in ensuring stability. The value distribution will bring a wave of practice for application manufacturers, but for a startup like Cheetah in China, landing on the C end is not particularly optimistic due to strong competition from large domestic manufacturers.

Q: What is your view on the slowdown of the scaling law? What are the impacts of the slowdown of the scaling law on the development of the large model application industry?

A: The slowdown of the scaling law is beneficial for startups like Cheetah as it allows focusing on application development instead of participating in model selection competition. It means everyone can focus on better utilizing model capabilities with agents, which is beneficial for application - related work.

Q: How do you view the robot industry trend over the past three months? What promoting effects do large models have on the implementation of robots? How do you view the future competitive landscape of the robot industry?

A: In the robot industry, there is a gap between the concept and practical implementation. Large models are helpful for the general direction of robots but need step - by - step advancement in combination with scenarios. The future competitive landscape will likely see more pragmatic robot manufacturers emerging, and the final winner will be the one that polishes products within specific scenarios.

Q: Are there any plans for share buybacks or dividends in the future?

A: Cheetah has an open attitude towards shareholder returns. In the future, if the Board of Directors approves share buyback and dividend plans, an announcement will be made to the market as soon as possible. Currently, maintaining sufficient cash reserve is important for business development considering economic environment uncertainties.

Q: At present, the development of large models in China is rapid from the perspective of performance and efficiency. Have you felt any significant differences among various models in actual operations or has the competition of large models gone beyond the model capabilities themselves and extended more towards productization and ecological construction?

A: It's not appropriate to evaluate which model is better. The competition of large models has already gone beyond model capabilities themselves, moving towards productization and ecological construction. Whoever can truly improve user satisfaction and product experience will gain an advantage, as the gap in underlying model indicators is being eliminated.