AI competition never stops, landing masters focus on search and cloud (Google 1Q23 earnings call transcript)

Below is the summary of Google-A.US Q1 2023 conference call, and the financial report can be referred to "Google: Exceeding Expectations Against the Trend? There is Joy and Worry" (https://longportapp.com/en/topics/5476099?invite-code=)

I. Key Financial Data vs Market Expectations

1. Revenue: $69.8 billion, YoY+2.6%, higher than market expectations (1.2% beat).
2. Gross profit margin: 56.1%, YoY-0.4ppt, higher than market expectations (1.3ppt beat).
3. Key Business:

1. Advertising: revenue of $54.5 billion, YoY-0.2%, higher than market expectations (1.48% beat).
2. Cloud: revenue of $7.45 billion, YoY+28.1%, basically in line with market expectations (-0.14% inline).

II. Management's Remarks

(I) General Situation In Q1 2023, the search business performed well, and the cloud business is developing vigorously. A major product update based on deep computer science and artificial intelligence, North Star, is providing users with the most useful answers, which contains enormous opportunities and continues the company's long-term innovation record. In terms of cloud computing, the company will continue this long and exciting journey to build this business; cloud computing achieved profitability this quarter and still focuses on creating long-term value.

The following content is divided into 3 aspects: 1. Progress of artificial intelligence and how it drives the development of search and other fields; 2. The company's focus on efficiency improvement; 3. The development of cloud, and 4. The progress of YouTube.

1. Artificial Intelligence Artificial intelligence is a tremendous development opportunity for consumers, partners, and company businesses, similar to the successful transition from desktop computers to mobile computers more than a decade ago. Thanks to the investment and breakthroughs made in artificial intelligence over the past 10 years, the company is currently in a favorable position.

The opportunities in the field of artificial intelligence cover three aspects: 1. Continuously develop the most advanced large-scale language models and make significant improvements to products to better serve users; 2. Provide tools to support developers, creators, and partners' work; 3. Enable companies of different scales to benefit from artificial intelligence. Currently, the company has made good progress in the above three areas.

In March of this year, an experimental dialogue AI service called Bard was launched. Since then, the PaLM model has been added to make it even more powerful. Bard can now help people with programming and software development tasks, including code generation, and more functionalities will be added in the future.

For developers, PaLM and the new MakerSuite tool were released. It provides a simple way to access large language models and quickly start building new generative artificial intelligence applications. Many organizations use generative artificial intelligence large language models in Google Cloud platform, Google Workspace, and network security products. On the search front, for years the company has been focused on making search more helpful: from Google Lens to multi-faceted search, visual exploration in search, immersive views in Maps, Google Translate, and now with all language models that support search, the company has opened up powerful avenues for knowledge acquisition through AI. In the future, we will continue to combine advances in generative AI, pushing for better search results.

Because billions of people trust Google to provide accurate information, we will continue to test and iterate, guided by data, years of experience about people's needs, and a high standard of quality. We will unlock new experiences in search and elsewhere, just as camera, voice, and translation technology have opened up new categories for search and exploration.

For more than a decade, AI has been the foundation of our advertising business. For example, Performance Max leverages the power of AI to help advertisers find untapped, incremental conversion opportunities. Details to follow.

The highest standards of AI principles and information integrity remain core to all our work. For example, the perspective API aids in identifying and reducing toxic text used for training language models, significantly improving information quality, and aimed at ensuring the safety of generative AI applications before they are released to the public.

We are proud to have a world-class research team that has been driving breakthroughs and supporting this new era of AI. Last week, we announced that we are combining Google Research's Brain Team and DeepMind into one department, integrating all the talent into a key team supported by Google's computing resources, which will help accelerate progress and develop the most powerful AI systems safely and responsibly.

2. Focus on Efficiency and Discipline

Last quarter, we promised to invest responsibly and with discipline, finding areas where we can operate more efficiently at a faster pace. For years, the company has been working to save costs, such as increasing machine utilization, and finding more scalable and efficient ways to train and serve machine learning models.

We are currently improving data center efficiency, reallocating workloads and devices in areas where servers are not fully utilized. This is important work as the company continues to heavily invest in infrastructure to drive AI development.

Data shows that improving external procurement is another area where significant savings can be found. Specific actions are being taken to manage the real estate investment portfolio to ensure it meets current and future needs. The company will continue to use data to identify other areas where savings can be made.

3. Google Cloud and YouTube

1) Google Cloud

The company has systematically expanded its product roadmap and market organization, establishing itself as one of the world's largest enterprise software companies; revenue continues to grow, and annual operating profit margins continue to improve, a trend that will continue this quarter. The growth comes from strong relationships with large enterprises, a robust partner ecosystem, and leading product positioning.

(1) Strong relationships with large enterprises

In the past 3 years, GCP's annual transaction volume has grown by nearly 500%, with over $250 million in bulk transactions, growing by over 300%. Nearly 60% of the world's largest 1000 companies are Google Cloud customers, and many leading startups and millions of small and medium-sized enterprises are using Google Cloud.

(2) Robust partner ecosystem

Over the past 4 years, the number of certified Google Cloud partners has increased by more than 15 times globally. The world's largest system integrators have established 13 dedicated practices using Google Cloud (from zero at the beginning). Today, more than 100,000 companies have joined the Google Cloud Partner Advantage Program.

(3) Leading product positioning

Providing generative AI technology to cloud computing customers: PaLM generative AI model and Vertex AI platform are helping Behavox to identify internal threats, Oxbotica to test its autonomous vehicles, and Lightricks to develop text-to-image functionality quickly.

In Workspace, new generative AI capabilities make it easier for customers like Standard Industries and Lyft to create content and collaborate. This builds on the popular AI tools Smart Canvas and Translation Hub, used by over 9 million paying customers.

Leading product positioning also extends to the field of data analysis, enabling customers to use AI to integrate and better understand data. New advances in Data Cloud enable Ulta Beauty to expand new digital and omnichannel experiences while focusing on customer loyalty, Shopify uses AI to provide better search results and personalization, and Mercedes-Benz launches new products to the market faster.

Generative AI has been introduced to identify network threats and prioritize them, automate security workflows and responses, and help scale network security teams. Cloud network security products have helped protect more than 30,000 companies, including innovative brands such as Broadcom and Europe's Telepass. Mandiant has successfully integrated with products, including Mandiant Threat Intelligence and Breach Analytics.

The company's open approach to AI development, coupled with industry-leading TPUs and NVIDIA's first-class GPUs, enables innovative companies to quickly and flexibly handle any AI workload. Many companies, such as AI21 Labs, Replit, and MidJourney, are building and training foundational models and generative AI platforms. With the launch of the G2 virtual machine, the company is the only cloud service provider to announce the release of NVIDIA's new L4 Tensor Core GPU.

In regards to YouTube Shorts, we continue to see strong growth from creators. Last year, the number of channels that uploaded to Shorts daily increased by over 80%, and those who posted to Shorts weekly attracted a large number of new subscribers. The Living Room remains the fastest-growing program for screen time in 2022, with strong global momentum.

As for subscription services, several updates have been made to YouTube Premium: premium users can now seamlessly switch video playback between their mobile and tablet devices, and automatically download recommended videos for offline viewing. YouTube TV and YouTube Prime Time channels are showing strong development momentum: the pricing for the NFL Sunday Ticket product has been announced, which will help drive subscriptions and attract new audiences for YouTube's supported paid and ad experiences, and create new opportunities for creators.

Finally, we are pleased to be able to help people, businesses, and society fully realize their potential through artificial intelligence. We will share how we are using AI in our products, including Pixel, and share new developments in Android at Google I/O.

In terms of performance, the following three areas show significant long-term growth momentum and contain huge advertising opportunities: 1. Google AI; 2. Retail, which covers all advertising, products, and services; and 3. YouTube.

1. Google AI

As mentioned earlier, artificial intelligence has always been an important driving force for our business. Technological advances have strengthened our ability to help businesses of all sizes respond in real time to rapidly changing market and consumer changes, and provide measurable return on investment when it is needed most.

In 1Q23, we will continue to innovate in our products. Taking core search as an example:

(1) In terms of positioning, the latest natural language AI update from the MUM model has improved search keyword relevance, increasing the relevance and performance of displayed ads when there are multiple overlapping keyword bid conditions.

(2) In terms of bidding, we have improved the intelligent bidding model to more accurately bid based on the differences in search ad formats - in other words, more effective bidding depends on how users want to interact with ads.

(3) In terms of creativity, automatic asset creation has been opened up to all English ad customers. ACA will generate text materials alongside adaptive search ads, using artificial intelligence to help reduce manual workload and keep ad materials relevant to users’ queries, context, and ad client’s business.

(4) To further unlock core search and maximize conversion rates, we are actively helping more ad clients combine it with Performance Max. Using PMax, the average conversion rate for advertising clients can achieve more than 18% under similar CPA. Due to advancements in AI-based bidding, creative, search query matching, and new formats such as YouTube Shorts, this number has increased by 5 percentage points in just 14 months.

(5) Travel is also a contributing factor to growth. In March, PMax was launched specifically for travel goals. Now, even the smallest hotel operators can benefit from PMax's hotel ad coverage. For example, the family-run Corissia hotel group saw a 32% increase in revenue and a 26% increase in direct bookings in just one month using PMax. More AI-based features will be launched in the future, so stay tuned for Google Marketing Live this May.

1. Retail

The focus is on three pillars: 1) making Google the core of consumers' shopping journey and a place for businesses to connect with users; 2) enabling more businesses to participate in free product listings and advertising experiences; and 3) further driving retail performance through excellent ad products. In the current macro environment, the company's tools and solutions have proven that even in challenging times, Google can provide value for retailers, online and omnichannel, and attract high-value customers.

Caraway, a consumer-facing cookware manufacturer, used target ROAS to remove budget caps and PMax to optimize and deploy spending on Google's Black Friday ad campaign. PMax drove a 46% increase in revenue and a 31% increase in ROAS, resulting in Caraway's best performance in history, leading to reinvestment in its AI-first strategy in 1Q.

For omnichannel retailers, the company recently launched store sales reporting as well as bidding and performance maximization for store goals. This not only helps retailers optimize for online conversions, but also for in-store optimization, attracting high-value customers who are more likely to make purchases in stores. Danish department store Magasin recently increased its omnichannel ROAS by 128% using the company's physical store sales solution. Thanks to dynamic in-store value data from first-party sources, including high-value customers, Magasin can now measure the comprehensive impact of its online investments on e-commerce and physical store revenue.

1. YouTube

Creators are driving YouTube's success. Whether it's long-form or short-form video, music or podcasts, vertical or horizontal, YouTube is where creators are inspired to create their best work - meaning the best content, more audiences, and more opportunities for advertisers. As mentioned in last quarter's performance conference, the creator ecosystem and multi-format strategy will be key drivers of long-term growth for YouTube. To continue growth, the focus will be on 1) Shorts; 2) interactivity on CTV; 3) investing in subscription products. 4) Long-term efforts to make shopping on YouTube easier.

1) Shorts

Viewing time is growing strongly, and monetization is progressing smoothly. People are increasingly participating in Shorts advertising at higher rates and achieving conversions.

2) CTV

Significant global development momentum, audiences love to watch YouTube creators and their favorite content on the big screen, and advertisers are also leaning towards this.

- YouTube as a whole

YouTube as a whole is helping brands benefit from wide coverage and driving profitability improvements. In the largest marketing portfolio modeling study to date, according to custom analysis of Nielsen-Compass ROI benchmarks from January 2020 to March 2022 (measuring a total media spend of 19 billion dollars in 16 countries), YouTube's ROI is 40% higher than linear TV and 34% higher than all other online videos, proving that YouTube has the ability to greatly increase effectiveness on a large scale.

3) Subscription products

The goal is to become a one-stop shop for a variety of types of video content in ad-supported and paid services, such as launching Multiview and the groundbreaking NFL Sunday Ticket ordering service on YouTube TV, with more updates expected in the coming quarters.

4) YouTube shopping

It's still too early to shop on YouTube. One highlight to share is that last year, we partnered with commercial platforms such as Shopify to bring shopping experiences to more creators and brands. Now, more than 100,000 creators, artists, and brands connect their stores to their YouTube channels to sell products.

Another example of how the company brings the best to partners to accelerate innovation is the partnership with Mercedes-Benz. In February of this year, it announced a groundbreaking partnership with Mercedes-Benz, bringing the Google Maps platform and YouTube into Mercedes-Benz cars equipped with the next-generation MB.OS operating system. In addition to allowing the luxury automaker to design custom navigation interfaces, it will also provide AI and Data Cloud capabilities to advance its autonomous driving work and create better customer experiences.

(III) 2Q23 Outlook

1. Macro environment - Forex

1Q23 performance reflected the sustained unfavorable factors brought by the challenging economic environment, and the prospects are still unclear. The unfavorable factors in foreign exchange have eased, and it is expected that the unfavorable factors in foreign exchange will decrease in 2Q23 based on the current spot exchange rate.

2. Advertising

1Q23 performance reflected the elasticity of search, which has a unique ability to bring demand to the surface and provide measurable ROI. Excluding the exchange rate impact, search business revenue growth was similar to that of the previous quarter. YouTube ad spending showed signs of stability on a month-on-month basis, and will continue to focus on increasing Shorts engagement, feeling encouraged by its monetization progress. 3. Other Revenue

In terms of YouTube subscriptions, YouTube Music Premium and YouTube TV's subscription users continue to grow significantly. In terms of Play, revenue decreased year over year, mainly due to the continued impact of foreign exchange transactions in the Asia-Pacific region, although performance improved after lowering costs last year.

4. Google Cloud

As more and more customers in various industries and regions rely on Google Cloud to achieve business digital transformation, the company's investment in product innovation, market organization, and partner ecosystem has achieved strong results. However, in 1Q23, as customers optimized GCP costs, consumer growth slowed, reflecting macroeconomic uncertainty. In terms of operating performance, the company will still focus on driving long-term profitable growth of cloud computing while considering continuing to invest.

5. Other Bets

In 1Q23, we are committed to improving our strategy and prioritizing the work of the entire investment portfolio, including layoffs.

6. Progress in Key Areas

1. Reshaping cost structure

(1) Use AI and automation to improve Alphabet's operational efficiency and the efficiency of our technology infrastructure.

(2) More effectively manage related spending with suppliers and retailers.

(3) Continue to optimize work methods and locations.

This year, we will intensify efforts in these three areas and plan to continue them in 2024 and beyond.

1. Improving operating expenses-employee numbers

At the end of 1Q23, the number of employees included all employees affected by the announced layoffs in January; it is expected that most of the affected employees will be removed from the employee count by the end of 2Q. We are slowing down our hiring pace for 2023 while still investing in priority areas, especially top engineering and technical talent.

1. Key investment areas-AI

We have integrated Google Research's Brain Team and DeepMind to establish Google DeepMind to accelerate innovation. Starting in 2Q23, costs associated with the team and activities will be transitioned from Google Services to Google DeepMind and included in Alphabet's undistributed enterprise costs.

7. Capital Expenditures

Total capital expenditures in 2023 are expected to be slightly higher than in 2022, and investment in data center construction and servers will accelerate in 2Q23 with sustained growth throughout the year.

Q: Which artificial intelligence tools that the company will launch are most anticipated by users, developers, and advertisers?

A: The use of AI in the search field has been around for some time, and the use of LLM has been a key factor in promoting search and improving search quality in the past few years. Now there is an opportunity to use LLM more, and we will continue to iterate and innovate with the same caution as providing the right service for users like North Star. Based on experience, we are confident that users will return to search and engage in tasks they have done before, which is why the company can use LLM to better serve these cases. For example, in the context of YouTube, there is an opportunity to truly improve the experience of creators and consumers in watching videos, among other things. Workspace has changed and this will be an area with the greatest progress - because productivity is a powerful use case and generative AI can provide assistance.

In cloud computing, this is an important moment, as almost every organization is considering how to use AI to drive transformation; therefore, from startups to large companies, they are partnering with companies, this is a turning point.

Q: What AI tools are used internally to improve the productivity of engineers, sales teams, and G&A

A: We are advancing many efficiency initiatives, one of which is the use of AI and automation to further enhance productivity across Alphabet. For example, AI is helpful in many analyses of finance organizations' operations and management.

Excitingly, the company has the opportunity to share this with cloud customers. Google Workspace is another example of how AI benefits us internally, and is also something that users and broader enterprise customers can use. In addition, computing power and infrastructure innovation are also helpful for our work, while benefiting our customers.

Q: Bard will gradually integrate into search products, what percentage of search queries will use responses of large language model types and the relative cost of searching on these models compared to those currently running

A: (1) Brad & LLM

We have launched Bard as a complementary product to search, but it will also introduce the experience of large language models into search. It will be launched gradually for testing, iteration, and innovation.

Overall, it can be applied to a wide range of queries and better help users solve a class of problems (perhaps without a right answer, but with more creativity, etc.). Even in existing query categories, we have already shared some of the heavy lifting with users and used AI to better guide them, and we will continue to explore in these directions in the future. It may still be too early now, but there will be many innovations in the future.

(2) Cost

Continuously improving the efficiency of hardware, software, and models is the company's habit, and it is not new. The steeper the technology curve, the more excited we become, because the company has established world-class capabilities, then sequentially reduce costs and then deploy on a large scale globally. Overall, all of these factors will be considered in promoting innovation, and currently we are satisfied with how we are handling costs.

Q: The reason for the moderate increase in capital expenditure on a sequential basis

A: Total capital expenditures for 2023 are expected to slightly exceed 2022, with an increase in 2Q and continuous growth throughout the year.

AI is a key component and the foundation of everything we do, so we will continue to invest in supporting AI and supporting users, advertisers, and cloud customers. 2023 saw a significant increase in capital expenditure, reflecting a surge in investment in technology infrastructure, while office facilities saw a slight decline compared to last year.

Q: When it comes to optimizing cloud computing, how do you balance the headwinds of declining revenue or backlog growth against the potential contributions long-term consumption growth and AI can make to improving cloud computing?

A: Given the uncertain economic environment and the slowdown in consumption growth, there are unfavorable factors to consider. However, clients are indeed hoping to optimize costs due to the macro environment. We are currently leaning towards optimization - this is a critical moment to help clients and focus on long-term goals. Therefore, this is definitely an area the company is dedicated to, doing our best to help clients improve efficiency.

Q: Progress on Monetization for Shorts

A: There has been a rapid growth in the number of Shorts views. In 4Q22, daily views exceeded 50 billion, higher than the 30 billion in the spring of last year. People are increasingly participating in Shorts advertisements at a faster pace and achieving conversions. Our top priority is to narrow the gap between Shorts and long-form videos, and we will continue to create a better creator and user experience.

Shorts advertisers can now acquire video action, app discovery, and Performance Max events, and product feed. Shorts can also be used for shopping. The company is the only place where creators can make a living through various formats and screens, and last year more than 80% of the number of channels uploading Shorts daily increased. The YouTube Partner Program was launched in February for revenue sharing. The company's sustainable revenue sharing model is unique in the industry, and many creators use this model. The ultimate goal is to make YouTube the best destination for Shorts viewers and creators, which is currently the company's focus.

Q: Given Microsoft's ambition to increase its share of paid search and Alphabet's cooperation with Android and iOS, how does Alphabet maintain unit economics and its impact on long-term profitability?

A: (1) Competition

Being in a competitive environment, the most important thing is to create the best products that focus on providing value to users. When working with partners, we strive to create a win-win experience, and therefore, partners ultimately choose Alphabet. This is also the reason why search is distributed widely. It all starts with continuing to innovate and improve search and ensuring that Alphabet stays ahead in the industry, and we will continue to do so in the future.

** (2) Long-term profitability **

We will continue to invest in growth to ensure overall improvement. There are currently many ongoing projects, and the cost basis is constantly being updated. We are particularly excited about the development opportunities for AI and other areas of long-term growth, such as search, advertising, cloud, YouTube, and hardware, and hope to have the ability to continue investing in these areas. We hope to ensure that cost growth does not outpace revenue growth by prudently promoting revenue growth and improving costs. Q: Combining language models with search has what effect on revenue (does it disadvantage advertising placement by providing faster answers to users)?

A: Firstly, the company has experienced many changes in the field of search over the years and has enriched its experience in how to develop advertising. The company can provide value to users in a reasonable way.

The basic driving factor is that people are looking for relevant information, and in the commercial category, they find that advertising has a high degree of relevance and value - this is the factor that drives this virtuous cycle. Also, users don't really want relevant commercial information, they want choice. The company focuses on sending traffic, and advertisers want to engage users. Therefore, the dynamic links that the company has relied on for survival for a long time still exist. Along with development, continuous iteration and testing will be carried out to promote innovation as usual.

Q: How to optimize operating expenses?

A: $2.6 billion in severance and office space costs resulted in an increase in operating expenses. Due to the changes in useful life, depreciation decreased by US$988 million, but this is a one-time benefit. Another benefit is that stock-based compensation was deferred from 1Q to 2Q.

Currently, we are very concerned about reducing costs and increasing efficiency in various workflow processes. The first is the speed of recruiting, and then the use of artificial intelligence and automation to improve various workflow processes. Everything we do with suppliers is to maximize efficiency, optimizing the way we work and the places we work.

Currently, many projects are underway, and the results will be apparent in 2023, and the effects will be even more significant in 2024.

Q: What is the demand for services and products in search ads?

A: Search business revenue has grown slightly year-on-year, reflecting growth in vertical retail and tourism, partially offset by declines in financial and media entertainment businesses. In fact, excluding the impact of foreign exchange, it is similar to the previous quarter. Although there are headwinds, the continued performance of search does reflect its resilience - search can emerge demand in uncertain environments and provide measurable ROI.

Many companies are very concerned about short-term profitability in uncertain macro environments, so budgets have been reduced.

Q: How does the high capital cost affect the evaluation and management of other bets (other businesses)?

A: We are very concerned about investment pace and financial returns, and the key to producing attractive returns has always been the use of deep technology to drive innovation, and core operational and long-term operating models will be the most relevant. Capital costs have indeed risen, but the core is value creation and returns.

The bets are in the development stage, and similar layoffs have been taken, and currently, costs are being improved to ensure the realization of value.

Q: Why has network advertising revenue fallen compared to search and YouTube?

A: Last quarter, advertiser spending continued to decline. Compared to the previous quarter's decline in YouTube and the network, ad spending is currently stable. Since this is a mixed business, the network continues to decline on a quarter-on-quarter basis for YouTube. A: We are very concerned about the cost factors and attractive long-term profitability path of cloud computing. At the same time, we will continue to invest in supporting long-term growth, particularly in providing opportunities for customers to use artificial intelligence features. We are satisfied with our current level and will continue to focus on profitability and long-term value creation.

Q: Product planning after the integration of the artificial intelligence team

A: Brain and DeepMind both can gather talents and work together in a coordinated way, which is helpful for pooling computing resources.

The core product is to build stronger models securely and responsibly, fully considering customers' needs on the consumer and cloud level, and able to iterate and achieve virtuous cycles. PaLM API has been launched and is being integrated into the product.

Q: Retail planning for the platform in 2023

A: Retail is an important vertical area and driving force for the company. As mentioned before, the growth rate of retail and other businesses is comparable, and can create value for online and offline retailers even in challenging times. The company is helping them achieve business goals, and retailers are increasingly focused on maintaining profit margins and improving ROI.

PMax and broad match are key levers to provide more incremental conversions, and a deep understanding of bidding and budget can indeed help retailers discover growth opportunities and improve efficiency in the company's product suite. For example, the omnichannel solution reaches nearby shoppers through artificial intelligence, promotes local inventory, optimizes in-store foot traffic, and helps bridge the gap between online and offline sales; YouTube app deep linking, new customer acquisition goals, and Pmax help retain loyal customers and attract new ones; and tools such as virtual cards on Chrome make checkout easy.

Overall, the company is providing the best AI-driven tools and solutions for retailers to maximize coverage and ROI, and create a truly seamless experience, including possibly providing payment experiences for customers, which will continue to be our focus.

Risk Disclosure and Statement in this Article: Longbridge's Disclaimer and General Disclosures