Focus: U.S. Stocks Q4 2025 Earnings Reports Alphabet announced its financial results for the fourth quarter and full year ended December 31, 2025. The report revealed that the company's annual revenue surpassed $400 billion for the first time, with Q4 revenue increasing 18% year-over-year. AI technology drove comprehensive growth across core businesses, Google Cloud maintained a high growth rate, overall profitability improved steadily, and all key metrics showed positive momentum. Details: Alphabet's 2025 Revenue Exceeds $400 Billion, Cloud Business Emerges as Core Growth Engine Following the earnings release, Alphabet CEO Sundar Pichai, Chief Business Officer Philipp Schindler, and CFO Anat Ashkenazi hosted an analyst conference call to discuss the results and answer questions. The following is a transcript of the call:
Morgan Stanley analyst Brian Nowak: I have two questions, one on Agentic Commerce and one on YouTube. First: Looking back at 2025, what do you consider Alphabet's most significant achievement regarding new Agentic Commerce products? Looking ahead to 2026, in which areas are you most confident of making greater progress to deliver more value for users and advertisers? Second, on YouTube: We see new content creation models like Genie emerging. Can you outline Alphabet's long-term vision for integrating Genie and other content creation tools into the YouTube platform? Sundar Pichai: Thank you for the questions. First, regarding Agentic Commerce. I believe 2025 was primarily a year of laying the foundation: we made our models more robust for agentic application scenarios, with particularly notable progress in the coding domain. In commerce, we collaborated with the ecosystem throughout the year to develop the underlying protocols needed for an agentic commerce world. This January, we launched the Universal Commerce Protocol with several founding partners at the NRF show, receiving excellent feedback. We have now built the interoperability foundation required for agentic commerce. This year, consumers will be able to practically use these functionalities, and I am excited about the future opportunities. Now, on the YouTube question: I am very enthusiastic about Genie and impressed by the incredible virtual worlds the team is creating. Genie has broad application scenarios, and our advantages in multimodality and real-world representation will be further highlighted; Genie represents a significant step towards building world models. Whether it's Lyria (for music) or Genie, all our innovations are integrated into products and offered to cloud customers, and YouTube will naturally become a key platform for creators to utilize these tools. Creators are already actively adopting these tools, and we always place creators at the heart of the experience. Ensuring YouTube remains a platform for creator expression is the foundation for all our efforts.
Goldman Sachs analyst Eric Sherman: During past earnings calls, we've frequently discussed the imbalance between AI demand and capacity, both internally and externally. Given your expectation of a significant capital expenditure increase in 2026, could you discuss the path to addressing the internal and external compute demand gap? Also, how will you continue to enhance operational efficiency within the business to fund these growth investments? Sundar Pichai: You are correct; despite our continuous efforts to expand capacity, we still face supply shortages. Clearly, this year's CapEx represents an investment for the future, and we must also consider factors like extended supply chain lead times, so we are always engaged in long-term planning. The scale of 2026 CapEx is based on the outcomes of work from previous years, noting there's a time lag effect. Currently, demand across our business segments—including Services, future projects at Google DeepMind, and Cloud—is exceptionally strong. Therefore, we anticipate remaining supply-constrained throughout 2026. Anat Ashkenazi: Thank you, Eric. As I've mentioned in previous calls, our approach to efficiency and productivity is this: it's not a one-time project but a core principle of our daily business operations. We constantly seek opportunities to enhance efficiency across the entire business. The more capital we can free up, whether from external customer demand or internal organizational needs, the better we can fuel the 'investment-growth' flywheel effect. Specifically, we are taking multi-faceted actions: First, ensuring every dollar invested in technical infrastructure achieves maximum efficiency, a process involving both scientific and technological innovation. As mentioned, we primarily build our own data centers, occasionally collaborating externally, to ensure construction efficiency aligns with our workload requirements. Second, we leverage AI to boost internal operational efficiency; for instance, portions of our code are now written by coding agents and reviewed by engineers, helping engineers increase productivity and accelerate progress. Furthermore, we optimize operations across the organization, from engineering teams to small back-office groups, even within my finance team. We've deployed agents in finance for accounting tasks and introduced them into payment processes. These methods help free up additional capacity to support future investments.
JP Morgan analyst Doug Anmuth: I have two questions. Competition in the Large Language Model (LLM) space has intensified over recent years, a trend expected to continue. What strategies does Alphabet employ regarding data distribution and product integration to consolidate Gemini's leading position? Additionally, what is the possibility of TPUs expanding beyond Google Cloud into external data centers, becoming a new revenue stream? Sundar Pichai: Thanks, Doug. First, the momentum in frontier LLM development is exciting, and we expect continued progress in this area during 2026. We are optimizing models across multiple dimensions, including pre-training, fine-tuning, and real-time inference. Concurrently, we are integrating multimodal capabilities, agentic functionalities, and have made significant strides in coding. We are bringing all these elements together to deliver superior experiences for customers through our products and APIs. There remains substantial room for development in this field; judging from the progress over the last two years, our pace of innovation is rapid, and we are confident in maintaining this momentum through 2026. Regarding TPUs: This should be viewed within the context of Google Cloud's overall advantages. A key attraction of Google Cloud is our offering of a wide choice of accelerators, meeting diverse customer compute needs. Combined with our strengths in data center interconnect efficiency, this has driven Google Cloud's robust growth. Given our overall investment scale, we anticipate maintaining this growth trajectory. That summarizes my view on TPUs.
Evercore analyst Mark Mahaney: I have two questions. First, YouTube advertising revenue grew 9% year-over-year. You mentioned strong direct response performance and relative strength in retail, yet YouTube's ad revenue growth doesn't fully reflect these positives. Could you explain why? Second, Sundar, a current market debate suggests we might be in a 'deep sea moment' phase. You previously stated Gemini is the AI engine for the world's most successful SaaS companies, yet the prevailing market view is that software companies are losing pricing power and their customer bases may be deteriorating. I find this hard to believe, but as a leader in AI, what is your perspective on AI's impact on software companies? Why would it not (or why would it) harm the unit economics of large software/SaaS company customer bases? Philipp Schindler: First, on YouTube: For the full year 2025, total YouTube revenue from ads and subscriptions surpassed $60 billion. The Q4 YouTube ad revenue growth was indeed driven by strong direct response performance. However, in brand advertising, the year-over-year growth rate was affected by a high comparison base from significant political ad spending during last year's U.S. election cycle, alongside some minor negative impacts in other brand-related verticals. It's important, however, to view YouTube's ad and subscription business holistically: When users shift from the ad-supported model to YouTube Music and Premium subscriptions, it creates a slight negative impact on YouTube ad revenue, but is positive for the company's overall business. This quarter, YouTube subscription revenue, particularly from YouTube Music and Premium, showed strong growth. More importantly, consider our future plans for brand advertising: the growth opportunity in Connected TV, innovative ad formats like the 'shoppable masthead' piloted during Cyber Monday, helping brands secure sponsorships with creators, and assisting advertisers in promoting services during high-attention moments. Additionally, we are expanding the functionality of the Creator Collaborations Hub, making it easier for brands to find creators and run campaigns. We are also investing heavily in brand collaboration and measurement, with more exciting developments to come. Beyond brand ads, there are growth opportunities in performance advertising: SME advertiser adoption of Demand Gen continues to increase, and ad innovations in areas like Shorts are promising. Overall, we are confident in YouTube's business outlook. Sundar Pichai: Regarding Gemini's application and its implications for the SaaS industry at this stage, from my perspective: Our SaaS customers are leaders in their respective fields. These successful companies are deeply integrating Gemini into their critical business processes, whether for optimizing product experiences, driving growth, or enhancing internal operational efficiency. AI is an enabling tool, much like those we use in Search, YouTube, and other products and services. I believe companies that seize the current opportunities will reap corresponding growth benefits in the future. Based on data like customer token usage, Q4 growth was very strong, and we are confident in these partnerships and the business momentum.
Bernstein Research analyst Mark Shmulik: I have two questions. The first is for Anat: Could you elaborate on the relationship between investment scale and core performance trends? Does Alphabet have specific operating profit or free cash flow targets guiding decisions? What are the primary factors considered when approving resources and projects? Second, a year ago we might have guessed the answer, but given the current environment, what is the single biggest concern for each of you when thinking about Alphabet's future development? Anat Ashkenazi: Thank you, Mark. First, on the investment framework, which is indeed an important topic for us. Internally, we employ a highly rigorous framework to evaluate all investment demands—whether from internal organizational needs or external customer requirements—and estimate the potential returns, including both short-term and long-term benefits. Based on this assessment, we make two types of decisions: First, determining the company's overall investment scale. For instance, our CapEx exceeded $90 billion in 2025, and the 2026 CapEx estimate follows the same logic, aiming to ensure the company's near-term and long-term growth through investment. Second, allocating funds across the organization. Throughout the year, we continuously monitor internal and external dynamics to ensure the rationality of our investment decisions. The encouraging aspect is that our investments in AI are already yielding returns. As seen in this quarter's results, AI is driving success across multiple segments like Cloud and Search, while frontier model development lays a solid foundation for future growth. We also consider cash flow generation, financial health, and the balance sheet to ensure investments are both financially prudent and maintain the company's strong financial position. Sundar Pichai: I'll address the 'biggest concern.' Overall, we have been pursuing an 'AI-first' strategy for over a decade, systematically advancing related work. For example, our development of TPUs for over ten years exemplifies this strategy. At this stage, my primary focus is undoubtedly compute capacity supply: encompassing various constraints and how to meet surging demand while ensuring efficiency and achieving world-class standards. We are actively addressing this challenge, dedicating significant time and resources to this area.
MoffettNathanson analyst Michael Nathanson: A question for Sundar (or others): You've frequently mentioned the Universal Commerce space. Could you discuss the rationale behind developing this area, the opportunities you see on platforms like Salesforce, and its significance for consumer product discovery paths? Also, Anat, could you comment on the proportion of CapEx allocated to long-cycle assets versus short-cycle assets? This would be very helpful. Sundar Pichai: Thank you, Michael. Clearly, users complete commercial journeys across our multiple platforms, and we also support retail partners through our Cloud and advertising businesses. Therefore, optimizing these commercial experiences holds potential for significant foundational improvement. The key is that advancing this area requires considering both user and merchant needs, defining the value proposition clearly. The design strength of the Universal Commerce Protocol lies in its ability to simplify shopping for users while allowing merchants flexibility in product presentation. These functionalities are built into the protocol design. I believe that only by delivering value to all ecosystem participants can we genuinely enhance the experience. We are currently implementing the protocol, and the agentic capabilities of the Gemini model are continuously improving. In the future, when users engage in product discovery and search, the entire commercial journey will be smoother, presenting expansion opportunities for the business. Anat Ashkenazi: Regarding CapEx composition: In 2025, approximately 60% of our CapEx was allocated to equipment, with the remaining 40% to the long-cycle assets you mentioned. We expect this ratio to remain largely consistent in 2026. You might also be interested in the depreciation differences: long-cycle assets typically have longer useful lives, while the depreciation period for servers is considerably shorter. Another important aspect is the allocation of CapEx: As mentioned previously, in 2026, over half of our ML compute resources will be allocated to the Cloud business. Thank you.
Barclays analyst Ross Sandler: A question on Gemini adoption: It reached 750 million monthly active users, adding 100 million in Q4. Could you discuss user engagement and retention for the Gemini native app overall? Is the 750 million MAU metric appropriate for gauging the company's progress relative to competitors like ChatGPT? Are there other user segments not captured by this statistic that we should consider? Sundar Pichai: Thanks, Ross. The growth momentum for the Gemini app in Q4 was indeed very strong. It's not just about the increase in MAU; user engagement intensity on the app has also improved significantly. All metrics—active usage rates, engagement intensity, retention, and performance across platforms and regions—are excellent. Optimizations to the product experience and continuous improvements to the Gemini model have driven this strong momentum, which continues, exciting us and prompting further investment. Additionally, it's important to consider the native experience of AI Mode within Search. This area has also seen robust growth, with the integration of Gemini 3 into AI Mode positively contributing to business growth. We will continue optimizing these experiences and anticipate further opportunities.
Wells Fargo analyst Ken Gawrelski: Thank you. Two questions on Search. First, how do you assess the extension of user session duration? As user value increasingly derives from on-platform results rather than outbound clicks, how is the company enhancing revenue opportunities to match this user value? Will premium subscription services play a larger role? Second, relatedly: In partnerships with companies like Apple, how do you ensure mutual success through appropriate models? Past collaborations involved revenue sharing, but the value created by AI search and Gemini on these platforms might be less tied to traditional 'search revenue.' How do you achieve value alignment with partners? Philipp Schindler: Firstly, it's important to note that the acceleration in Search growth isn't driven by a single factor but by the synergistic growth of multiple components. Vertically, retail, finance, and healthcare were primary contributors to Search revenue, and almost all major verticals saw improved growth rates in Q4. Specifically addressing your question: Continuous innovation is our core strategy. Optimizing both user and advertiser experiences is key to driving performance growth. We achieve growth through multiple initiatives: increasing overall query volume; enhancing search ads by using Gemini to improve query matching and ad creative generation; and deepening our understanding of user intent, enabling us to monetize long-tail, complex queries previously difficult to address. For example, AI Max is used by hundreds of thousands of advertisers, continuously unlocking billions in new query monetization opportunities. In ad creative, we've used Gemini to generate millions of creatives. These efforts are performing very well.
BofA Securities analyst Justin Post: A follow-up on two Gemini app questions. First, as users engage more with the Gemini app, have you observed any signs of cannibalization of Search traffic? Second, what is the progress on monetizing the Gemini app? With new formats like agentic ads emerging, could this business provide incremental growth contribution in the coming years? Sundar Pichai: Currently, we see users embracing a diversity of choice. They use both Search and the Gemini app. This diversification is actually driving an expansion phase for our business. Overall, it broadens the range of query types users can perform through Alphabet, and we haven't observed signs of cannibalization thus far. Philipp can add on monetization. Philipp Schindler: On monetization, as discussed earlier regarding Agentic Commerce, our priority remains creating excellent user experiences. Currently, we have incorporated monetization elements into early experiments within AI Overviews and AI Mode. We have more plans for the future. Regarding the Gemini app itself, our current focus is on the free tier and subscription services, with the subscription business showing good growth momentum. Advertising has always been key to scaling our products to billions of users. When done right, ads can provide valuable commercial information at the right moment. We will share specific plans in due course, but we are not rushing monetization at this stage. (End)
Comments