Google: Overinvesting is preferable to underinvesting (2Q24 conference call)

The following is the summary of the second quarter financial report conference call for $Alphabet(GOOGL.US) in 2024. For financial report analysis, please refer to 360-degree in-depth look, is Google really that stable? 》

I. Review of Core Financial Information:

II. Detailed Content of the Financial Report Conference Call

2.1. Key Points from Executive Statements:

1) Operational Highlights:

① Search Business:

a. Performance: The search business continues to perform strongly, showing outstanding results this quarter.

b. Progress in AI Overview: Significant progress has been made in product innovation, increasing search usage and user satisfaction.

c. Business Innovation:

• Providing better responses through AI, expanding types of search queries, introducing new search methods, with higher engagement of 18 to 24-year-old users when using AI overview.
• Prioritizing methods that can send traffic to various websites across the web, with ads appearing above or below the AI overview, providing valuable choices.

② Cloud Business:

a. Performance: The cloud business reached several milestones in the second quarter, with quarterly revenue exceeding $10 billion for the first time, and quarterly operating profit exceeding $1 billion for the first time. AI infrastructure and generative AI solutions have generated billions of dollars in revenue for cloud customers, with over 2 million developers using these solutions.

b. Infrastructure Development:

• Announced the establishment of the first data center and cloud region in Malaysia, and expansion projects in Iowa, Virginia, and Ohio.
• TPU Trillium is the sixth-generation custom AI accelerator, significantly improving performance and energy efficiency. The latest NVIDIA Blackwell platform will join Google Cloud in early 2025.

③ AI Business:

a. Gemini Model: Over 1.5 million developers use Gemini, launching more powerful and efficient new models. Gemini provides highly useful features in products such as search, Workspace, Google Messages, used by over 2 billion monthly active users across 6 products. New features showcased at I/O will soon be rolled out to Gmail and Google Photos.

b. Project Astra: Demonstrated multimodal understanding and natural conversation capabilities, showcasing early development of a universal agent. Future advancements in AI products will be based on long-standing research leadership and global network infrastructure c. Future Outlook: Continue to invest in AI technology and infrastructure, leveraging research and infrastructure leadership to drive technological development.

④ Youtube:

a. In the United States, YouTube ranks first in streaming watch time, with significant increases in YouTube Shorts and TV-connected viewing. By connecting creators and audiences through advertising and subscriptions, it helps advertisers reach their desired audience.

b. Viewing of YouTube Shorts and TV-connected content doubled last year, making it easier for creators to add subtitles and convert regular videos into Shorts.

⑤ Waymo:

a. Performance: Waymo has provided over 2 million rides and driven over 20 million miles on public roads. It offers over 50,000 paid public ride services per week, mainly in San Francisco and Phoenix.

b. Innovation: The waitlist in San Francisco has been eliminated, allowing anyone to ride, and fully autonomous driving tests are being conducted at other locations in the Bay Area.

⑥ Android & Pixel:

a. Business Progress: Collaborating with Samsung, Google has shared AI-driven updates on new devices. The new Pixel 8a, powered by the latest Google Tensor G3 chip, offers an outstanding AI experience.

b. Recent Activities: Looking forward to the Made by Google event in August, which will share content about the Android and Pixel device series.

2) Strong Performance of Google Services: The application of AI in advertising business and YouTube's position as a multi-format platform

① Application of AI in Advertising Business: Over 30 new ad features and products were announced this quarter, helping advertisers leverage AI to meet customer and user expectations. In the search, PMax, Demand Gen, and retail sectors, AI is used to streamline workflows, enhance creative asset production, and provide more engaging consumer experiences.

a. Feature Products: The new Merchant Center expands the Product Studio, allowing business owners to generate ad materials using AI. Virtual try-on and shopping ad features in testing are expected to be widely promoted later this year, with feedback showing a 60% increase in high-quality views and higher click-through rates.

c. Profit Optimization: AI-driven profit optimization tools have been expanded to Performance Max and standard shopping ad campaigns, resulting in an average profit increase of 15%.

d. Demand Generation: Soon to be expanded to Display & Video 360 and Search Ads 360, providing new image generation tools for social media marketers to create high-quality image assets. When used in conjunction with search or PMax, Demand Gen delivers an average 14% increase in conversion rates.

② YouTube Platform Dynamics: YouTube differentiates itself from other platforms through creators and their connection with fans, with the MrBeast channel surpassing 300 million subscribers

a. Viewing: Long-term investment in Connected TV (CTV) continues to pay off, with viewing times increasing by over 130% in the past three years. YouTube has been the most-watched streaming platform on American TV screens for the 17th consecutive month. Sports content has seen significant growth, with CTV viewing time on YouTube increasing by 30% year-on-year.

b. Monetization: CTV on YouTube continues to benefit from increased viewing time, audience and advertiser innovation, as well as the shift of brand advertising budgets from linear TV to YouTube. Verizon's advertising campaign with YouTube creators shows that AI-driven ad formats have lower costs and higher conversion rates. The monetization rate of Shorts continues to improve, with a significant increase in brand advertising contributions.

c. Shopping: Last year, 300 billion hours of shopping-related videos were watched, with viewing time increasing by 25%. Several updates have been launched for YouTube Shopping, including product tags, product collections, and a new Affiliate Center, to help creators sell products to viewers.

3) Financial Highlights:

① Business Outlook:

a. Google Services:

• Advertising: Strong performance in search across various verticals, satisfactory growth on YouTube this quarter, healthy growth in viewing time, narrowing monetization gap for Shorts, and sustained brand momentum for Connected TV. Looking ahead to the third quarter, we will face challenges in advertising revenue growth in the second half of 2023, partly due to retailers in the Asia-Pacific region.
• Subscriptions: Significant growth in subscription business, mainly driven by the price increase of YouTube TV, with this impact expected to continue for the remainder of the year.
• Platforms and Devices: Play performed well with an increase in buyer numbers. On the devices front, Made by Google releases are scheduled for the early third quarter, expected to contribute to revenue for the quarter.

b. Google Cloud: Strong growth in cloud business is expected to continue, benefiting from significant growth in GCP and Google Workspace, with AI infrastructure and generative AI solutions expected to continue generating revenue for cloud customers and being adopted by more developers.

② Investments and Capital Expenditures:

CapEx: $13 billion in the second quarter, mainly for technology infrastructure investments. Quarterly CapEx for the full year is maintained at or above $12 billion from the first quarter. Waymo investments have received a new multi-year $5 billion investment commitment.

Employee Count: Employee count is expected to slightly increase in the third quarter, while continuing to invest in top engineering and technical talent, especially in cloud services and technology infrastructure.

③ Profit Margin Guidance: Looking ahead, it is expected that Alphabet's operating profit margin for the full year 2024 will expand relative to 2023. However, due to increased levels of technology infrastructure investment and the early release of hardware in the third quarter, the profit margin for the third quarter may be impacted by increased depreciation and expenses.

2.2. Analyst Q&A

Q: In the past 18 months, which AI applications have exceeded expectations, and which ones have fallen behind?

A: The translation of technology into practical solutions, whether for consumers or enterprises, takes time.

On the consumer side, I am pleased to see products like search being introduced in a way that enhances the experience and positively impacts the overall user experience. We are making progress in organic growth for our consumer products, and monetization is the next step we need to focus on.

On the enterprise side, we currently have many models gradually converging into a set of foundational capabilities. I believe the next wave is to build solutions on top of these, with some applications in coding and customer service gaining attention, but more effort is needed to truly unleash the potential of these technologies.

Q: Regarding the structural redesign of operating expense foundations. Can you provide more specific examples of how to further improve company efficiency?

A: We emphasize "persistently redesigning cost foundations" because we are engaged in deep workflow changes, not temporary patches. Our focus is on the priority of products and processes, organizational efficiency, and structure, all of which are reflected in the reduction of workforce.

Recently, we merged the hardware and services product area with the platform and ecosystem product area, which helps to enhance the speed and efficiency of product execution. We are also continuing to advance efficiency improvements in technical infrastructure, AI applications across all of Alphabet, the establishment of centralized procurement organization, and optimization of real estate portfolio. These are areas that our leadership team continues to focus on and execute in order to achieve persistent optimization of cost foundations.

Q: Google Services' operating profit margin has remained around 40% for two consecutive quarters. Considering future investments, can this level be sustained?

A: The operating profit margin of Google Services does reflect our efforts in persistently optimizing cost foundations and the strong revenue growth of the search business. Looking ahead to the third quarter, the profit margin will be affected by increased depreciation and expenses due to investments in technical infrastructure, as well as additional costs from the early release of Pixel products. Nevertheless, we expect Alphabet's full-year operating profit margin in 2024 to increase compared to 2023.

Q: Regarding AI Overviews, can you share the progress of promoting AI Overviews, as well as the click-through rate and monetization level compared to traditional search?

A: AI Overviews have been launched in the United States, and we plan to expand to more countries within this year. Our intention is to ensure quality and maintain healthy metrics. We will gradually expand its application scope and reach more queries. The feedback we have received so far has been positive.

In terms of monetization, innovation and improving the search user experience have always brought new opportunities for advertisers. The transition from desktop to mobile devices is one example. Generative AI has expanded the types of questions we can help people answer, and people have found the ads appearing in AI Overviews helpful As announced at the GML conference, we will soon begin testing search and shopping ads in AI Overview in the United States. These ads will be clearly marked as sponsored content and will appear in positions related to queries and AI Overview information, providing further innovative opportunities.

Q: Regarding Chrome's decision not to phase out Cookies, can you share our experience and the reasons behind this decision?

A: We place great importance on enhancing user privacy in Chrome. The Privacy Sandbox project is a focus of our ongoing attention. Regarding third-party cookies, considering their impact on the ecosystem and feedback from all parties, we believe that allowing users to choose for themselves is the best solution. We will continue to protect privacy by providing choices and investing in privacy-enhancing technologies. We will also listen to feedback from all parties in the ecosystem, always prioritizing privacy.

Q: The growth rate of the search business exceeds that of YouTube, can you analyze the factors leading to this phenomenon? Additionally, does AI technology contribute to YouTube's growth, as it does in the search business?

A: The growth in search revenue does cover various verticals, with retail and financial services being the main growth drivers.

Regarding YouTube's year-over-year growth, we are satisfied with the team's performance, with growth mainly driven by brand advertising and direct response advertising. The slowdown in revenue growth in the second quarter is mainly due to the fact that compared to the first quarter of last year, YouTube's year-over-year growth was negative, resulting in a lower base for comparison. In addition, this year's first quarter benefited from an extra day due to leap year. Currently, YouTube is experiencing the anniversary effect of growth in the Asia-Pacific region, as well as facing adverse effects from foreign exchange. Despite some timing issues, we want to emphasize that YouTube's operational strength remains solid.

One more point to add, we do need more time to build multi-modal models, such as Gemini which inherently has multi-modal features from the start. While there has been progress in technical use cases, achieving high-quality generative audio and video experiences will take more time.

Q: Can you elaborate on the adoption of AI in enterprises and its potential impact on cloud business strategy and revenue growth?

A: The combination of cloud services and AI is undoubtedly a long-term significant driver. Our AI infrastructure and generative AI solutions on the cloud, including Vertex AI, Gemini in Workspace, and Gemini in Google Cloud, have attracted deep engagement, with over 2 million developers currently using these tools and early use cases starting to emerge.

We are at a stage where deep work is needed to ensure that these use cases and workflows can unlock value more deeply, and I am very optimistic about this. While it will take time, looking at the long term, especially considering our investments in AI and research infrastructure, as well as our leadership position, all of these will directly translate into our strengths

Q: What are the new discoveries in the subscription model for YouTube? How interested are consumers in subscribing to more media products?

A: The overall subscription platforms and devices have achieved healthy growth, mainly driven by subscription services. The growth is primarily fueled by the increase in subscribers for YouTube TV and YouTube Music Premium, as well as the strong growth in users and revenue for Google One.

The core issue lies with YouTube, which is the main source of revenue for this business line. We have seen a significant acceptance of subscription services by users, with strong growth in subscribers, which we are very pleased with. Despite a slight slowdown in growth for this business line due to the anniversary effect of the YouTube TV price increase, there is still a strong interest in subscription products.

Q: Observing the super-large data center industry transitioning from insufficient construction last year to now increasing capacity to meet demand, there is a concern that if capital expenditures continue to grow, there may be overbuilding next year. Is this view correct? How do we evaluate the return on investment for this round of AI capital expenditures?

A: We are currently in the early stages of a technological transformation. In the face of such a transition, we believe that actively investing in defining technological areas is reasonable, especially in areas where these technologies can have cross-domain impacts on our core products and services, including search, YouTube, and other services, while also driving cloud service growth and supporting our long-term innovation investments.

I believe that in this transformation, the risk of underinvestment far outweighs the risk of overinvestment. Even in the case of overinvestment, these infrastructures are beneficial for us in the long term, and we have enough time and space to leverage them. Our team is very cautious about every penny invested, whether it's in hardware, software, or the efficiency of model deployment, and we have put a lot of effort into optimization.

Q: Is the AI industry approaching a bottleneck in improving basic models in AI training, such as due to a lack of new data for training? What are your thoughts on this?

A: Regarding the scalability of AI models and whether they are reaching a bottleneck, I believe we are pushing forward with full force, continuously making breakthroughs in computing power to expand the boundaries of these models. Regardless of the outcome, we are continuously optimizing and driving the continuous improvement of model capabilities. More importantly, applying these technologies to actual consumer and business use cases, in this area, we believe there is still a lot of room for progress, which is also our focus.

Q: Currently, we mainly see use cases focused on cost savings and efficiency improvement. When do you think we will see AI products helping Fortune 500 and Fortune 1000 companies increase revenue, thereby creating greater long-term value, rather than just reducing costs?

A: Technology has broad applicability, capable of both improving efficiency and increasing revenue. For example, by enhancing customer service experience, we can not only reduce costs but also enhance the overall experience, increase conversion rates, optimize sales processes, and even increase the basket size for retail e-commerce. Currently, extensive experiments have been conducted in both these areas, and AI technology will play a crucial role in improving efficiency and increasing revenue.

Q: I noticed that brand advertising is the main factor, followed by direct response advertising. Considering the importance of shopping you mentioned earlier, when do you think direct response advertising will become the main driver of growth for YouTube, rather than just brand advertising?

A: In terms of direct response, we have found that AI-driven advertising campaigns integrated with videos can significantly improve conversion rates. Whether it's PMax or demand generation, advertisers using automated tools to create video content see a noticeable performance boost. In fact, advertisers running image and video ads can achieve an additional 6% conversion for every dollar spent compared to those using images alone.

On the brand advertising front, Google AI is making it easier for brands to be showcased alongside highly engaging content for users. Data shows that this is a very effective way to increase brand awareness and consideration. At the same time, we are very excited about some of the latest developments in YouTube shopping, which can be seen as part of direct response.

Q: Regarding the growth of cloud services, is it mainly being driven by AI demand, a more general resurgence in computing demand, or is AI driving this growth?

A: The cloud team is actively promoting AI solutions globally, especially generative AI, which strengthens our business. It is worth mentioning that most of our top 100 customers are already leveraging our generative AI solutions.

The growth rate of GCP exceeds the overall growth rate of cloud services, which is encouraging for us. We are also pleased to see an improvement in the profit margin of cloud services, reflecting strong revenue growth and efficiency improvement measures we mentioned earlier. Looking ahead to the third quarter, we expect the profit margin to follow the same seasonal pattern as last year and continue to invest in the business.

Q: How do you use AI to improve efficiency and reduce costs for internal cost savings? Has the efficiency of engineers been enhanced as a result?

A: We are committed to staying at the forefront in the field of engineering and coding. These tools are being used by some of the most efficient and rigorous engineers, and they are conducting in-depth testing on them. While we are still in the early stages of writing high-quality, secure code, the experience gained will help improve our models and products, forming a virtuous development cycle with a promising future.

Risk Disclosure and Statement of this Article: Dolphin Research Disclaimer and General Disclosure