Meta (META.US) 2025Q3 performance meeting: It is expected that advertising performance will continue to show strong momentum in the next two fiscal years.

Recently, Meta (META.US) held a performance meeting for the third quarter of 2025. Meta stated that it is taking a "preemptive" strategy for Compute investment to ensure sufficient resources for AI model expansion. With the continuous iteration of the AI advertising system, the company expects strong performance in advertising for the next two fiscal years. Looking ahead to 2026, the efficiency improvement of the advertising recommendation system will continue to be the main driver of growth. Regarding Meta's AI models, the company plans to continue iterating the Meta AI models and will regularly release enhanced versions of the Frontier series in the future. Each improvement in model performance brings higher user engagement and interaction time. Currently, Meta AI has over 1 billion monthly users, and improvements in reasoning, creation, and multimodal interaction of new models will further enhance user stickiness and the accuracy of the advertising system. The upgrade expected in 2026 will be a major leap forward, with significant performance improvements for AI assistants, advertising recommendations, and creative generation tools. Q&A Q: Regarding the 2026 AI advertising and computing investment, can you provide specific early results of internal A/B testing? How do these developments support your investment return expectations under high CapEx investment? A: We have made significant progress in AI advertising models, with ad conversion rates significantly increasing year-over-year when weighted by value. Since the introduction of the unified architecture Lattice in 2023, it has helped integrate around 100 small models into a more efficient general model. Expanding Lattice to apply to advertising in Q3 resulted in a 3% increase in conversion rates. These improvements demonstrate that even in the context of rapid CapEx growth, AI infrastructure investments can yield clear ROI. Looking ahead to 2026, the efficiency improvement of the advertising recommendation system will continue to be the main driver of growth. Additionally, we take a "preemptive" approach to Compute investment to ensure sufficient resources for AI model expansion. With the continuous iteration of the AI advertising system, we expect strong performance in advertising for the next two fiscal years. Q: The company mentioned "pre-building" computing power. How do you balance the rapid growth of CapEx and expenses, profitability, and free cash flow? A: We are currently in the budgeting phase for 2026. The company does want to pre-align computing power to meet potential demands for AI models. While CapEx growth is significant, we believe it will be an important foundation for driving long-term revenue growth. We expect that even with higher computing power investments, the payback period will remain manageable. When we add computational resources, they are almost always effectively utilized for ad optimization and model training in the short term. From past experience, building computing power in advance is more cost-effective than passively waiting. Even if there is temporary excess capacity, we can apply the computing power to external API services or new product incubation, thus keeping overall risk under control. Q: On the user level, how is the feedback on Meta AI usage? How will more powerful models change user behavior and the AI ecosystem? A: Currently, Meta AI has over 1 billion monthly active users, and we have observed a direct positive correlation between model performance improvements and user engagement. With the introduction of higher-tier Frontier models, we expect Meta AI to bring deeper changes in content creation, recommendations, and commercial interactions. In addition to assistant functions, AI will also drive the emergence of new product forms such as videos, images, and creative content. With Meta's unique scale advantage, we can quickly push AI products to billions of users and embed AI capabilities into core applications. In the future, AI will improve both user experience and advertising performance. Q: How sustainable are the future improvements in advertising? Will Super Intelligence Labs' new model be released in 2026? A: We employ a "knowledge transfer" method in our advertising system where large training models guide lightweight inference models, resulting in performance improvements while maintaining cost control. Even small percentage points of performance improvement, when applied to Meta's advertising scale, can bring substantial revenue. We will continue to advance new models and new product development, but we have not yet disclosed a specific timeline. The overall direction is to continue to enhance the intelligence level of AI systems in recommendation, creative generation, and ad matching. Q: How will AI products affect profit margins? Is it possible to compress the overall profit space of the company? A: It is currently too early to make a judgment, as the profitability characteristics of different AI products vary greatly. Our goal is to maximize user value and overall profitability, rather than focusing on short-term profit margins. With the commercialization of AI advertising, creative tools, and business services advancing, we expect the profit structure to gradually stabilize. Q: Does Meta Super Intelligence Labs pursue goals similar to AGI (General Artificial Intelligence)? How do you balance long-term scientific research with short-term commercial value? A: Our focus is on developing broadly applicable intelligent capabilities, rather than solely pursuing AGI. AI reasoning, content generation, ad optimization, business assistants, etc., are all core practical scenarios. These capabilities will be embedded in advertising systems, business AI, and content recommendations, bringing direct commercial value to the company. Meta aims to maintain leadership in key technological directions and scale the results for billions of users. Q: What is the commercialization path for Meta AI? Will you consider direct monetization in the future? A: We currently focus on improving model quality and user scale for Meta AI. The usage and model improvements of Meta AI show a positive feedback loop. In the future, AI capabilities will span advertising creation, material generation, customer interactions, and placement optimization. The ultimate goal is to allow advertisers to input their business goals and budget, and the system will automatically generate ad materials and placement plans, achieving end-to-end automation. Q: How is the adoption of the Advantage Plus automated advertising solution? Can you share feedback from advertisers? A: Advantage Plus has been fully rolled out to sales and prospective client advertising series, achieving end-to-end automation. Advertisers can use the system to automatically complete audience selection, budget allocation, and placement optimization. The number of advertisers using this solution has grown by 20%, and the system has an annual revenue run rate exceeding $60 billion. Additionally, we have introduced AI-generated music and video extension features, increasing the match between ad content and brand tone, and continuing to improve overall conversion performance. Q: Can you explain the expenditure plan for the future two years regarding the pace of AI infrastructure investment and CapEx management? A: We expect capital expenditures in 2026 to significantly exceed those in 2025, primarily for expanding GPU clusters and building new-generation data centers. Multiple projects are currently in various stages of construction, with some projects expected to come online in the second half of 2026. While the investment scale is expanding, we still expect to maintain a healthy free cash flow, benefiting from strong growth in the advertising business and increased operational efficiency. The expansion of AI infrastructure is part of a long-term strategic plan aimed at ensuring Meta has leading computational reserves to support future Frontier models and business-level AI products. Q: What are the strategic focuses for Reality Labs in the future? How do you balance long-term investment with commercialization in the product roadmap? A: We believe that mixed reality (MR) and wearable devices will become the next generation computing platforms. The focus of Reality Labs is on building smart glasses and spatial computing experiences with AI capabilities. Current main products include the Quest series headsets and Ray-Ban Meta smart glasses, the latter of which has shown excellent performance in visual, voice, and generative interactions. In the short term, this sector is still in the investment stage, but in the long run, it will form a strong moat supporting the integration of social, creative, and work scenarios. Q: How is the update frequency of Meta AI models and the increase in user engagement? A: We plan to continuously iterate Meta AI models and will regularly release enhanced versions of the Frontier series in the future. Each performance improvement in the model can bring higher user engagement and interaction time. Currently, Meta AI has over 1 billion monthly active users, and improvements in reasoning, creation, and multimodal interaction of new models will further enhance user stickiness and the accuracy of the advertising system. We expect the upgrade in 2026 to be a major leap forward, with significant performance improvements for AI assistants, advertising recommendations, and creative generation tools.