What impact will AI have on Microsoft's performance?

The following is a summary of the conference call. For financial analysis please see ""Microsoft " Can It Fly high Again with ChatGPT?".

1. Management Statement

The quarterly revenue of Microsoft Cloud exceeded USD 28 billion, with an increase of 22% and 25% respectively calculated at a fixed exchange rate, which indicates that we continue to hold a leading position in the entire technology field. We continue to focus on three priorities:

First, helping customers to leverage the breadth and depth of Microsoft Cloud to attain maximum value.

Second, investing in leading the new AI wave in our solution field and expanding our TAM.

Third, driving operational leverage to keep our cost structure consistent with our revenue growth.

Infrastructure: Azure continues to seize market share - from the cloud to the edge. When every application is driven by AI, we have the most powerful AI infrastructure. Many of our partners are using it to train large models.

Our Azure OpenAI service combines advanced models such as ChatGPT and GPT-4 with Azure's enterprise features. We now have more than 2,500 Azure OpenAI service customers, an increase of 10 times month on month.

Last week, Epic Systems shared that it is using Azure OpenAI services to integrate next-generation AI with its industry-leading EHR software. Azure also supports OpenAI APIs. We are pleased to see brands such as Shopify and Snap using APIs to integrate OpenAI's models.

More broadly, the world's leading enterprises continue to migrate critical workloads to our cloud. Unilever is fully committed to Azure this quarter, marking one of the largest cloud migrations in consumer goods history. We see many enterprises using Azure Arc to run Azure services in local, edge, and multi-cloud environments. We now have over 15,000 Azure Arc customers, with a year-on-year growth rate of over 150%.

We are using Azure for Operators to extend our infrastructure to the 5G network edge. We are the preferred cloud for communication companies. At last month's MWC, AT&T, Deutsche Telekom, Singtel, and Telefonica shared how they are using our infrastructure to modernize and monetize their networks.

Data: Our intelligent data platform combines databases, analytics, and governance, so organizations can spend more time creating value and less time integrating their data assets. Cosmos DB is the preferred database and can support the most demanding workloads in the world at any scale. OpenAI has relied on Cosmos DB to dynamically scale their ChatGPT service—one of the fastest-growing consumer applications ever, combining high reliability with low maintenance costs.

Developers: From Visual Studio to GitHub, we have the most popular tools to help every developer go from idea to code, and then from code to cloud—all while keeping them in their flow. Today, 76% of the Global 500 use GitHub to build, ship, and maintain software. With the first large-scale AI developer tool, GitHub Copilot, we're fundamentally changing the productivity of every developer, from novice to expert.

Copilot for Business has seen over 10,000 companies sign up in the last three months, including Coca-Cola and General Motors, as well as Duolingo and Mercado Libre, all thanks to Copilot speeding up developer velocity.

We're also bringing the next generation of AI to the Power Platform, so anyone can automate workflows, create apps or websites, build virtual agents, and analyze data using natural language, with nearly 33 million monthly active users on the Power Platform, up nearly 50% year-over-year.

Business Applications: From customer experience and service to finance and supply chain, we continue to share share across all categories we serve, as organizations like Asahi, CH Robinson,$Eni Spa-Sponsored(E.US)N, and Franklin Templeton choose our AI-driven business applications to automate, simulate, and predict every business process and function. We're taking it further with Dynamics 365 Copilot working across CRM and ERP systems, bringing next-generation AI to every job role, reducing manual data entry, content generation, and note-taking.

Future of Work: Microsoft 365 Copilot combines next-generation AI with business data in Microsoft Graph and Microsoft 365 applications, eliminating tedious work and unleashing creative potential. Copilot works alongside the user, embedded in Microsoft 365 applications used by millions every day, and also supports Business Chat, which uses natural language to surface information and insights based on business content and context. We're encouraged by early feedback and look forward to bringing these experiences to more users in the coming months.

Teams usage hits an all-time high, with over 300 million monthly active users this quarter. As we add value for existing customers and win new ones like ABN AMRO Bank, Jaguar Land Rover, and Vodafone, we're again gaining share across categories—from collaboration, chat, and meetings to calling. We have released a new version of Teams that can increase performance by up to two times and reduce memory usage by 50%, allowing customers to collaborate more efficiently and prepare for experiences like Copilot.

Teams is also expanding our TAM. Nearly 60% of enterprise Teams customers purchase Teams Phone, Rooms, or Premium. Teams Phone is the undisputed market leader in cloud calling, helping our customers reduce costs with a three-year ROI of over 140%. Teams Rooms revenue has grown by more than double year-over-year. Teams Premium meets enterprises' needs for AI-powered features like intelligent insights. Now generally available, it is one of our fastest-growing Modern Work products, with thousands of paying customers just two months after launch.

Through Microsoft Viva, we have created an entirely new suite of employee experiences. Viva brings together goals, communication, learning, workplace analytics, and employee feedback. Across industries, companies like Dell, Mastercard, and SES are using Viva to help their employees thrive. Just last week, we announced Copilot for Viva, a new way for leaders to build high-performing teams that prioritize productivity.

With Viva Sales, we are extending the platform to specific job functions, helping sellers apply large language models to their CRM and Microsoft 365 data so they can automatically generate content like customer emails. All of these innovations are driving growth across Microsoft 365. Ferrovial, Goldman Sachs, Novo Nordisk, and Rogers have all chosen E5 to provide their employees with our best-in-class productivity apps, as well as advanced security, compliance, voice, and analytics.

Windows: While the PC market continues to face headwinds, we are seeing record monthly active Windows devices and higher usage compared to pre-pandemic levels. With Windows 365 and Azure Virtual Desktop, we continue to transform how companies like Mazda and Nationwide access Windows. So far, more than a third of our enterprise customer base has purchased Windows delivered through the cloud. The new Windows 365 Frontline extends cloud PC capabilities and security to shift workers for the first time.

Security: Our comprehensive AI-driven solutions covering all clouds and all platforms provide defenders with an agile advantage. With Security Copilot, we combine large language models with domain-specific models informed by our threat intelligence and 65 trillion daily security signals to transform all aspects of SOC productivity. We have added new governance controls and policy protections to better protect identities and their access to resources.

Nearly 720,000 organizations are now using Azure Active Directory, a year-on-year increase of 33%. Overall, nearly 600,000 customers now have four or more secure workloads, up 35% year-on-year.

LinkedIn: We continue to see record engagement with over 930 million members turning to our professional social network for connections, learning, selling and hiring. As we expand to new audiences, member growth accelerated for the seventh consecutive quarter. We now have 100 million members in India, up 19%. As Generation Z enters the workforce, we're seeing a 73% year-on-year growth in student sign-ups.

In this continued tight labor market, LinkedIn Talent Solutions continues to help recruiters connect with job seekers and professionals. Our recruiting business saw market share grow for the third consecutive quarter.

LinkedIn is increasingly becoming a place for learning, conversation and skills development, with over 100 AI courses offered. We've also launched new AI features including writing suggestions for member profiles and job descriptions, as well as collaborative articles. LinkedIn Marketing Solutions continues to be a leader in the B2B digital advertising space, helping companies deliver the right message to the right audience on a safe, trusted platform.

More broadly, we are continuing to expand our opportunities in advertising. Our exclusive partnership with Netflix brings differentiated premium video content to our advertising network. Our new copilot is reshaping everyday search and browsing habits. We're encouraged by user feedback and usage patterns since the launch of new Bing and Edge two months ago, with over one billion active users on Bing daily. We're winning new customers on Windows and mobile devices. Daily installations of the Bing mobile app have quadrupled since launch. Edge has now held market share for eight consecutive quarters while Bing's market share grew again in the US.

We continue to innovate with groundbreaking AI features, including the ability to set chat tone and create images from text prompts, powered by Dall-E. Over 200 million images have been created to date. And we see that as people use these new AI features, their engagement with Bing and Edge increases. With chat entry points supported by Bing, Edge, Windows and OpenAI's Azure, businesses can truly choose the model and manner of GPT chat. We look forward to continuing this journey in the generational shift in the biggest software category - search. Gaming: We are quickly realizing our ambitions to become people's first choice to play exciting games anytime, anywhere.

We created records for monthly active users and monthly active devices in the third quarter. This quarter, our subscription revenue reached nearly $1 billion. We also brought PC Game Pass to 40 new countries/regions this quarter, nearly doubling the number of markets available to us. Over 500 million lifetime unique users for our first-party games have been achieved.

I've never been more excited about our gaming pipeline, including the releases of Minecraft Legends and Redfall.

In a few weeks, we'll be hosting the Build conference, where we'll share how we're building the most powerful AI platform for developers.

Guidance:

Based on current exchange rates, we anticipate that foreign exchange will reduce total revenue growth by about 2 percentage points, without affecting COGS or operating expense growth.

We anticipate negative foreign exchange impact of about 2% on productivity and business process as well as intelligent cloud revenue growth, and about 1% on individual calculations.

Microsoft Cloud gross margin is expected to grow by about 2% YoY. Excluding this impact, cloud computing gross margin next quarter will be relatively flat, with Office 365's improvements offsetting the lower Azure profit margin and the impact of expanding our AI infrastructure to meet growing demands.

We anticipate that capital expenditures will substantially increase driven by investment in Azure AI infrastructure. However, there may be normal quarterly spending fluctuations in cloud infrastructure construction time.

Productivity and Business Process:

Office Commercial: Revenue growth will once again be driven by Office 365, with seat growth in customer segment markets and ARPU increase through E5 growth. We anticipate Office 365 revenue growth to be around 16% (calculated by fixed exchange rate).

Office Consumer: Revenue is expected to achieve mid-single-digit growth driven by Microsoft 365 subscriptions.

LinkedIn: Talent Solutions is expected to drive mid-single-digit revenue growth.

Dynamics: Revenue is expected to grow at the mid-to-high level, driven by the continued growth of Dynamics 365.

Intelligent Cloud:

We expect revenue to grow at a fixed exchange rate of 15% to 16%, or $23.6 to $23.9 billion.

Revenue will continue to be driven by Azure, mainly from per-user business and period revenue recognized based on contract combinations.

Azure: Fixed-exchange rate revenue growth is expected to reach 26% to 27%, including about 1% from AI services. On-premises server: We expect revenue to decline to low single digits as demand for our hybrid solutions (including Windows Server and SQL Server running in multi-cloud environments) is offset by adverse forex impact.

Enterprise services: Revenue should be relatively stable YoY, as growth in enterprise support services is offset by decline in Microsoft Consulting Services.

Personal computing:

PC demand should be similar to Q3 and, despite normalization in the market, our revenue will lag behind overall market growth given high channel inventory. Therefore, Windows OEM and devices revenue is expected to decline around mid-single digits by 2025.

Windows Commercial and Cloud services: Revenue is expected to decline to mid-single digits.

Search and news advertising: Ex-TAC revenue growth should be around 10%, about 5 points higher than overall search and news advertising.

Gaming: We expect revenue to achieve mid to high single-digit growth.

COGS:

Fixed-rate forex translation should increase by 3-4%, i.e. $16.8 to $17 billion, while operating expenses should increase by about 2%, i.e. $15.1 to $15.2 billion.

Other income and expenses should be around $300 million due to expected interest income exceeding interest expense, adding quarter-to-quarter volatility.

We expect our Q4 effective tax rate to be consistent with our full-year rate of approximately 19%.

Finally, as a reminder for Q4 cash flow, we expect to pay $1.3 billion in cash taxes related to capitalized R&D.

Q&A

Q1: How is revenue recognized in partnerships with OpenAI and any other potential relationships?

A1: We have talked about the foundation of our partnership activities, which is a win-win. Sometimes we view a partner as our customer, just like any other customer that will be using Azure infrastructure services to support their end customers. When they do so, we recognize revenue on behalf of that customer.

Q2: Is the decline in revenue for Azure's IaaS/PaaS due to macro factors or more so due to fundamental factors?

A2: I think it's important to differentiate between macro performance and relative performance, first of all, optimization is still ongoing. In fact, we encourage our employees to help customers optimize because we believe it's the best way to ensure loyalty and long-term contracts with customers—they can rely on companies like us to help constantly optimize workloads.

The second thing is we've seen new workloads move to the cloud, both new workloads and extending existing workloads, during the pandemic. But prior to the pandemic, there was a balance between optimization and new workloads. We're seeing both optimization and new workloads again.

We talked about transitioning away from starting a lot of new workloads and that we've been in a transition period for almost a year now. We continue to optimize, but in some cases workloads can't be further optimized. ——So you can see some year-over-year impact in our guidance.

Q3: Regarding the profitability of Copilot, how do you consider the magnitude of price increases?

A3: Overall, we do plan to monetize a set of separate monetization measures in all technology stacks, whether it is consumer metering or subscriptions per user. The pricing of GitHub's Copilot is a good example of how we monetize. Expect us to do things like what we did on GitHub Copilot.

Q4: How do you view the calculation cost of Artificial Intelligence workload compared to classical workloads? How do you think it will develop over time?

A4: It is obvious that accelerated computing is used to drive AI. We are very concerned about ensuring that computing resources are used effectively. The function of supercomputers is not only the hardware of the stack. They use software to optimize the performance of a given workload so that we can continue to push for computing resource optimization. For a given generation of large models, one can also see changes in costs within a quarter. We will bring the continuous optimization of COGS to different workloads.

In addition to AI, we have many other workloads. For example, one of the main reasons we collaborate with OpenAI is that training or actuarial insurance will be more relevant to us, not just in the context of AI.

Q5: How do you view the evolution of gross margin, especially in support of Generative AI?

A5: Our current investment and focus are relative performance and share. Currently, we have the largest commercial cloud, with more and more customer commitments to workloads and new TAM. Our focus will be on continuing to occupy more and more share, whether it is E5 products, Microsoft 365, or the help of Windows 365 and perhaps the optimization of computing and PC costs.

With these opportunities, plus our consumer business, where the most devices are in the Windows ecosystem, this allows us to focus on gaming share, bringing services to PCs and mobile devices, which are the opportunities we will focus on next year.

We hope to focus on investing in AI, which will increase COGS growth, but we are committed to ensuring our healthy profitability by keeping operating costs low.

In fact, this is our way of looking at long-term success, being in a favorable position in the market, capturing market share, making sure we lead this trend, and then continuing to focus on improving gross margins where possible. Given our commitment to building, over time, we will charge for these AI capabilities, which will ultimately bring operating profit.

Q6: Regarding the number of employees in the Office365 business, can you evaluate the sensitivity of current slow recruitment and layoffs?

A6:

We still have the opportunity to continue to grow in the frontline scenarios of small and medium-sized enterprises. In enterprises, Microsoft is a basic productivity tool, and most labor markets in most areas are still tight, and we continue to see customers gaining value. Our focus is actually on continuing to get healthy renewals, adding new products when renewing to save customers money and increase value. We did have a good E5 quarter, which helped with ARPU.

Q7: How do you see the potential for regulation of artificial intelligence, and concerns about data, customer privacy, and governance? What do you think you can do to alleviate government, customer, and organizational fears?

A7: Overall, we take an approach of not waiting for regulation to appear. The approach we take is that any unintended consequences of new technology, we consider from day one to be first class and build into our engineering processes all of the safeguards. So, for example, when we introduced our AI principles back in 2016, we turned those AI principles into a set of internal standards, which were then further turned into implementation, and we audited ourselves against. That's the framework that we have.

We have a chief AI officer who is responsible for thinking about what those standards are, but also responsible for the people who help us audit ourselves through that traceability process. But, obviously, we'll participate in any regulatory action that shows up in any jurisdiction around AI. But we think that the more trust you have in artificial intelligence, the more we’re going to benefit.

Q8: What is your view on Microsoft integrating its business in the current environment? If there are certain areas you are particularly interested in or see more attractiveness.

A8: Whether it's Microsoft 365 or Office 365, we're focused on customers being able to get value from our products. We want our products to be deployed, used, and obviously that leads to share growth.

It's the same on top of the Azure stack, right? So when you think about artificial intelligence, it’s not just an AI model that’s needed. For example, ChatGPT uses Cosmos DB as its core database. So we want to make sure our services, that they are competitive, used together, whether it’s at the IaaS layer, the PaaS layer or at the SaaS layer.

The other aspect of that is some new business processes that are being automated, like Dynamics workloads, can also benefit from artificial intelligence. From core Azure functionality, and some front-end products, to working loops being realized in a new way in Microsoft 365 products. This may not be the traditional definition of integration.

But when people think about creating value and the new tools that are required, I think we have excellent value, and quite frankly, in every vertical tool where we're competing it’s a leading tool.

Risk Disclosure and Statement: Dolphin Disclaimer and General Disclosure