The Zhitong Finance App learned that Ignite, the Microsoft (MSFT.US) annual technology conference, will be held from November 17 to 20 in San Francisco, USA. Unlike the Build Conference for developers, which focuses on showcasing software development tools and new technologies, Ignite focuses more on enterprise customers, and Microsoft will use this to showcase its latest developments in the fields of cloud computing, artificial intelligence, data, security, and intelligent devices.
Morgan Stanley sees this year's Ignite as an important point to test the progress of the implementation of Microsoft's corporate AI strategy, not just a new product launch. The bank maintains Microsoft's “Overweight” rating, and the target price is $650. Damo's core judgment is that Microsoft is building an “enterprise AI flywheel” composed of Azure, Fabric, Foundry, and Microsoft 365 Copilot: Azure provides the underlying computing power, Fabric integrates enterprise data and business context, Foundry helps customers select and orchestrate models and build intelligent bodies, and Copilot embeds AI capabilities into employees' daily work processes.
According to Damo, what is really worth paying attention to at this conference is not how many new features Microsoft has launched, but rather whether this technical system can be transformed into more enterprise production deployments, higher product consumption, and more sustainable revenue growth.
Important conference schedule: Nadella's keynote speech will be held in the early morning of the 18th Beijing time
According to Microsoft's official Ignite schedule, the conference will be held at the San Francisco Moscone Center from November 17 to 20, and online events will also be offered. The opening keynote will be held at the San Francisco Chase Center, co-hosted by Microsoft CEO Satya Nadella (Satya Nadella) and Judson Althoff.
The schedule of important public events is as follows:
| activity | US Pacific Time | Beijing time |
|---|---|---|
| Nadella's opening keynote | November 17, 10:00 — 12:00 | November 18 02:00 — 04:00 |
| Live at Microsoft Ignite: Cloud & AI Innovation | November 17 12:00 — 13:00 | November 18th 04:00 — 05:00 |
| Azure and Nvidia Accelerated Computing Topics | 11/17 15:15 | 11/18 07:15 |
| Foundry and Nvidia Nemotron Smart Device Topics | 11/18 11:30 | 11/19 03:30 |
| Cloud to edge intelligent AI factory topic | November 19, 14:45 | 11/20 06:45 |
In addition, the Damo report also listed investor communication arrangements with Microsoft management: an exchange with Nadella will be held at 16:00 EST on November 17; a meeting with Chief Financial Officer Amy Hood (Amy Hood) and Vice President of Investor Relations Jonathan Neilson (Jonathan Neilson) will be held on November 18 at 15:00 EST. The above are investor events listed in the report, not conference keynote speeches open to the public.
The company's AI flywheel is already supported by performance, but the key is whether it can continue to deliver
Damo's optimistic judgment on Microsoft's AI strategy comes first from the latest quarterly results.
Microsoft announced the results for the first fiscal quarter of the 2027 fiscal year on October 28. For the quarter ending September 30, the company's revenue reached US$92.3 billion, up 19% year on year; operating profit increased 21% year over year, and the operating profit margin was about 50%. Azure revenue increased 48% year over year, and Microsoft 365 commercial cloud revenue increased 18% at a fixed exchange rate.
Commercial remaining fulfillment orders (RPO) reached $677 billion, an increase of 73% year over year. Of this, approximately 30% are expected to be recognized as revenue within the next 12 months. This huge reserve of orders provides high visibility for Microsoft's future cloud business growth.
Notably, Azure's growth is not solely dependent on AI infrastructure requirements. Microsoft management said that Foundry's revenue increased more than threefold year-on-year, and the additional computing power brought about by the increased efficiency of GPU clusters, as well as improvements in CPU resource utilization, all contributed to Azure's growth. Meanwhile, Microsoft Cloud revenue reached $63.9 billion, up 30% year over year and 31% at a fixed exchange rate.
Damo believes that Microsoft's advantage is that it can connect cloud infrastructure, data platforms, model services, and office applications, so that enterprises can complete multiple aspects from data governance to intelligent deployment within the same ecosystem. However, whether this model can actually form a closed commercial loop, we still need to find more evidence in this edition of Ignite.
Five highlights: Copilot monetization, Foundry orchestration, and computing power efficiency
I. Commercialization of Copilot: Shifting from seat growth to ARPU increase
Whether Copilot can further increase the average revenue per customer (ARPU) is one of the most important observations of this conference.
Damo believes that Microsoft's commercialization path is expanding from simply selling Copilot seats to three directions: increasing the penetration rate of direct-paid seats, promoting customers to upgrade to higher-level Microsoft 365 E7 packages, and obtaining additional revenue from smart devices to perform tasks through a pay-per-use billing model.
Microsoft revealed during the latest fiscal quarter conference call that Copilot daily active users increased nearly 50% month-on-month, with a year-on-year increase of more than fivefold; the number of customers actively using pay-as-you-go Copilot Credits increased by more than 75% month-on-month. Additionally, the number of customers with more than 50,000 Copilot seats increased more than eight times over the same period last year.
This means that Microsoft is trying to expand the AI business model from traditional software subscriptions to charging for actual usage. As agents take on more workflows, it is expected that enterprise spending will no longer be limited to the IT department's software procurement budget, but will gradually extend to the business unit's operating budget.
However, Microsoft also changed the way Copilot business data is disclosed in the latest fiscal quarter. The company previously announced the number of paid seats continuously, and disclosed that the Microsoft 365 Copilot paid seats exceeded 30 million in July 2026; however, in the latest fiscal quarter, Microsoft did not update this figure and instead emphasized daily user and customer usage and revenue momentum.
Chief Financial Officer Hood made it clear during the conference call that the company did not disclose an update on Copilot seat numbers this quarter. Although usage metrics have increased, the latest base of paid seats is lacking, and it is still difficult for outsiders to accurately determine the extent to which the increase in user activity translates into additional paid revenue.
Changes in pricing models are also worth watching. The Microsoft 365 E7 package launched by Microsoft is priced at $99 per user per month and integrates products such as Microsoft 365 E5, Copilot, Agent 365, and Entra. At the same time, Microsoft plans to enable pay-per-use billing by default in new Microsoft 365 Copilot Business licenses sold through cloud solution provider channels, and set a default quota of 4,000 Copilot Credits per user per month. The arrangement has been adjusted to the third quarter of fiscal year 2027, which is to be rolled out gradually from January to March 2027.
Therefore, the key question that Ignite needs to answer this year is: can Microsoft prove that Copilot not only has more active users, but can also continue to increase revenue through seat upgrades and pay-as-you-go billing.
2. Foundry: In the multi-model era, can Microsoft master intelligent system orchestration entrances
If Copilot is responsible for delivering AI to end users, then Microsoft Foundry acts as a platform for enterprise AI application development and deployment.
Microsoft said in the latest fiscal quarter that the number of Foundry customers has exceeded 120,000, and related revenue has more than tripled year over year; the number of customers using multiple models to build intelligent solutions has increased fivefold year over year, while revenue from the open source model on Foundry has increased tenfold year over year.
Damo believes that the increasingly rich selection of models will not necessarily reduce the value of Microsoft's platform. Conversely, a unified model management and orchestration platform may become more important as businesses need to choose between different models and adjust task allocation based on quality, cost, and speed of response.
Microsoft also stated during the earnings call that 90% of Copilot queries use the Auto mode, where the system dynamically routes between different models to balance quality, speed, and cost. This also reflects that Microsoft is incorporating multi-model scheduling capabilities into its product system.
What is worth watching for this edition of Ignite is whether Microsoft can further showcase Foundry's actual progress in model selection, agent development, deployment, and operation management, and whether these capabilities can lead to more enterprise workloads and cloud service consumption.
3. Fabric: Can enterprise data become a differentiating advantage in the AI era
As basic model capabilities gradually become more popular, an enterprise's true competitive advantage may increasingly depend on being able to make AI understand its own business data and work processes.
Microsoft Fabric is an important part of this strategy. The platform is designed to integrate enterprise data and provide actionable business information and context for agents.
According to Microsoft's latest fiscal quarter, the number of paid Fabric customers exceeded 44,000, an increase of more than 50% over the previous year; there were more than 19,000 customers using Fabric and Foundry at the same time. This means that more and more companies are beginning to combine data platforms with AI application development platforms.
Damo believes that the governed enterprise data and business context is an important advantage that distinguishes Microsoft from simple model suppliers. If enterprises are willing to migrate more data and workflows to the Microsoft platform, the collaboration between Fabric and Foundry is expected to further drive the growth of Azure consumption.
4. The popularity of smart devices drives security requirements, and Agent 365 and corporate governance are worth paying attention to
As AI agents move from answering questions to performing tasks across applications, managing their rights to access enterprise data, call tools, and perform operations will become more important.
Microsoft is integrating authentication, permission control, behavior monitoring, and data protection capabilities into the agent management system. Products such as Entra, Defender, and Purview are expected to gain more demand from enterprises as a result, while Agent 365 takes on the role of managing, monitoring, and governing AI agents.
Microsoft previously included Agent 365 in the Microsoft 365 E7 plan. During the latest earnings call, Nadella said that the number of agents registered with Agent 365 has exceeded 50 million, an increase of more than 30% over the previous month.
For Microsoft, the increase in the number of smart devices means not only potential cloud computing needs, but also the need for new security and governance tools for enterprises. If Microsoft can combine intelligent management with existing enterprise security products, it may further push customers to upgrade to higher-priced enterprise software packages.
5. Maia 300: Can self-developed chips improve AI infrastructure costs
Demand for AI computing power continues to grow, and data center construction and chip procurement are also putting huge capital expenditure pressure on Microsoft. Therefore, whether self-developed chips can reduce the cost per unit of computing power is an important variable affecting Microsoft's long-term profitability.
Microsoft announced the next generation of self-developed AI inference chip Maia 300 in October, and plans to expand the deployment scale of Azure Maia series chips starting in 2027. The company expects the number of tokens the Maia 300 can provide per megawatt to be about 40% higher than the previous generation, and that the rack-scale system's token throughput can reach up to seven times that of the Maia 200.
However, an increase in chip performance does not necessarily mean an immediate reduction in overall costs. Investors still need to focus on its actual deployment pace, supported workloads, and cost and efficiency advantages over third-party GPU solutions.
Damo believes that as Microsoft expands the deployment of self-developed chips, improvements in AI infrastructure efficiency are expected to ease pressure on gross margins and enhance the long-term returns of the Azure business.
Gross margin is under pressure, but can operating leverage offset cost pressure?
Another aspect of the rapid growth of Microsoft's AI business is that capital expenses and infrastructure costs continue to rise.
Microsoft's capital expenditure in the first fiscal quarter reached 45.2 billion US dollars, of which about three-quarters was spent on short-term assets such as CPU, GPU, memory, and storage. The company expects capital expenditure for the second fiscal quarter to exceed $50 billion, and estimates capital expenditure for the 2026 calendar year to be approximately US$170 billion.
The company's free cash flow for the same period was $18.6 billion. Microsoft said it expects free cash flow to remain positive in the 2027 fiscal year, but it also indicated that it will continue to use financing channels flexibly in the future according to capital costs and market conditions.
According to Damo's model, Microsoft's gross margin may gradually decline from 67.9% in FY2026 to 65.9% in FY2027, 64.5% in FY2028, and 63.5% in FY2029. However, the bank expects operating profit margins of 46.8%, 46.6%, 46.7%, and 47.2%, respectively, during the same period, to remain stable overall.
This means that Damo expects Microsoft to offset the gross margin pressure brought about by some AI infrastructure investment through revenue growth, platform efficiency improvements, and cost control. For investors, what really matters is whether gross margin declines in the short term, but whether AI investment can bring sufficient revenue and operating leverage to keep operating margins resilient.
Behind the target price of $650: Valuation depends on AI commercialization
In terms of valuation, Damo maintains Microsoft's “Overweight” rating and a target price of $650. The bank's benchmark scenario is based on earnings of $24.41 per share for the fiscal year 2028, which corresponds to a price-earnings ratio of about 26 times.
Damo's target price for the bull market scenario is 800 US dollars, corresponding to earnings of 27.41 US dollars per share for fiscal year 2028 and a price-earnings ratio of about 29 times; the target price for the bear market scenario is 300 US dollars, which corresponds to earnings per share of 22.25 US dollars.
The bank believes that Microsoft is expected to continue to expand its AI market share through collaboration between Azure, Foundry, Fabric, and Copilot. However, this judgment also means that Microsoft needs to continue to prove that the commercialization progress of AI products can match the market's expectations for growth and profit.
The biggest test: Can revenue growth keep up with capital expenditure?
Behind this year's Ignite, there are also two conditions that need to be established at the same time: on the one hand, after Copilot expands from seat subscriptions to pay-as-you-go billing, it must actually bring revenue growth; on the other hand, Microsoft must increase the utilization rate of AI infrastructure to avoid continuing to rise in capital expenditure without being able to simultaneously improve return on investment.
Microsoft's capital expenditure reached 45.2 billion US dollars in the first fiscal quarter, and is expected to exceed 50 billion US dollars in the second fiscal quarter. Despite strong growth in Azure and Foundry, investment in AI infrastructure is still huge. Microsoft needs to prove to the market that every dollar of additional capital expenditure can bring more realizable computing power and higher return on revenue.
Therefore, what is most noteworthy on Ignite is not a new model or agent function, but three more commercially meaningful developments: whether CoPilot can improve ARPU through seat upgrades and pay-as-you-go billing; whether Foundry can become a core platform for enterprise intelligence deployment in a multi-model environment; and whether self-developed chips and data center efficiency improvements can reduce the cost per unit of computing power.
For Microsoft, competition for enterprise AI is no longer just about model capabilities, but also about full-stack commercialization capabilities from infrastructure to applications, from data to security. Whether Ignite can provide a more clear implementation roadmap will be an important window for the market to evaluate the return on its AI investment.