According to Zhitong Finance APP, Microsoft’s (MSFT.US) annual technology conference, Ignite, will be held in San Francisco, USA, from November 17 to 20. Unlike the Build conference, which targets developers and focuses on software development tools and new technologies, Ignite is more focused on enterprise customers, and Microsoft will use this event to showcase its latest advancements in cloud computing, artificial intelligence, data, security, and agents.
Morgan Stanley views this Ignite event as a critical milestone in assessing the progress of Microsoft’s enterprise AI strategy, rather than merely a product launch. The firm maintains an “Overweight” rating on Microsoft with a price target of $650. Morgan Stanley’s core thesis 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 agents, and Copilot embeds AI capabilities into employees’ daily workflows.
In Morgan Stanley’s view, what truly matters at this event is not how many new features Microsoft launches, but whether this technology ecosystem can be translated into increased enterprise deployment, higher product consumption, and more sustainable revenue growth.
Key event schedule: Nadella's keynote speech will be held at 18:00 Beijing time on the 18th.
Microsoft’s official Ignite schedule shows that the conference will take place from November 17 to 20 at the Moscone Center in San Francisco, with virtual events also available. The opening keynote will be held at the Chase Center in San Francisco and co-hosted by Microsoft CEO Satya Nadella and Judson Althoff.
The schedule for important public events is as follows:
| Event | Pacific Time (PT) | 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 18, 04:00–05:00 |
| Azure and NVIDIA Accelerated Computing Special Topic | November 17, 15:15 | November 18, 07:15 |
| Foundry and NVIDIA Nemotron Agent Special | November 18, 11:30 | November 19, 03:30 |
| AI Factory: From Cloud to Edge Intelligence | November 19, 14:45 | November 20, 06:45 |
Additionally, the Morgan Stanley report outlines investor engagement schedules with Microsoft’s management: a meeting with Nadella on November 17 at 4:00 PM Eastern Time, and a meeting with Chief Financial Officer Amy Hood and Vice President of Investor Relations Jonathan Neilson on November 18 at 3:00 PM Eastern Time. These events, as listed in the report, are investor engagements and not public keynote presentations.
The enterprise AI flywheel already has performance backing, but the key lies in whether it can continue to deliver.
Morgan Stanley's optimistic assessment of Microsoft's enterprise AI strategy stems first and foremost from its latest quarterly earnings.
Microsoft announced its first-quarter fiscal year 2027 results on October 28. For the quarter ended September 30, the company reported revenue of $92.3 billion, a 19% year-over-year increase; operating profit rose 21% year-over-year, with an operating margin of approximately 50%. Azure revenue increased 48% year-over-year, and Microsoft 365 commercial cloud revenue grew 18% at constant exchange rates.
The commercial remaining performance obligations (RPO) reached $677 billion, a 73% year-over-year increase. Approximately 30% of this is expected to be recognized as revenue over the next 12 months. This substantial order backlog provides high visibility into Microsoft’s future cloud business growth.
It is worth noting that Azure's growth is not solely driven by demand for AI infrastructure. Microsoft management stated that Foundry revenue more than tripled year-over-year, and increased computing power from improved GPU cluster efficiency, along with better CPU resource utilization, collectively fueled Azure's growth. Meanwhile, Microsoft Cloud revenue reached $63.9 billion, up 30% year-over-year, or 31% at constant exchange rates.
Morgan Stanley believes Microsoft’s advantage lies in its ability to connect cloud infrastructure, data platforms, model services, and productivity applications, enabling enterprises to complete multiple stages—from data governance to agent deployment—within a single ecosystem. However, whether this model can truly form a commercial闭环 still requires further evidence from this year’s Ignite.
Five Key Highlights: Copilot Monetization, Foundry Orchestration, and Computing Efficiency
I. Copilot Monetization: Shifting from Seat Growth to ARPU Improvement
Whether Copilot can further increase the average revenue per user (ARPU) is one of the most important observations at this conference.
Morgan Stanley believes Microsoft’s monetization strategy is expanding beyond simply selling Copilot seats to three directions: increasing penetration of directly paid seats, driving customers to upgrade to the higher-tier Microsoft 365 E7 plan, and generating additional revenue through usage-based pricing for agent-executed tasks.
Microsoft disclosed on its latest earnings call that daily active users of Copilot increased nearly 50% quarter-over-quarter and more than fivefold year-over-year; the number of customers actively using pay-as-you-go Copilot Credits increased by over 75% quarter-over-quarter. Additionally, the number of customers with more than 50,000 Copilot seats increased by more than eightfold year-over-year.
This means Microsoft is attempting to expand its AI business model from traditional software subscriptions to usage-based pricing. As agents take on more workflows, enterprise spending is expected to extend beyond IT departments' software procurement budgets and gradually move into business units' operational budgets.
However, Microsoft also changed its disclosure approach for Copilot business data in the latest fiscal quarter. Previously, the company had consistently reported the number of paid seats, disclosing in July 2026 that Microsoft 365 Copilot had exceeded 30 million paid seats; however, in the latest fiscal quarter, Microsoft did not update this figure, instead emphasizing daily active users, customer usage, and revenue momentum.
Chief Financial Officer Hu De clearly stated on the earnings call that the company did not provide an update on the number of Copilot seats this quarter. Although usage metrics have increased, the lack of updated figures for paid seats makes it difficult for outsiders to accurately determine how much of the rise in user activity has translated into new paid revenue.
The change in pricing models is also noteworthy. Microsoft’s Microsoft 365 E7 plan is priced at $99 per user per month, bundling Microsoft 365 E5, Copilot, Agent 365, and Entra. Meanwhile, Microsoft plans to default to usage-based billing for new Microsoft 365 Copilot Business licenses sold through the Cloud Solution Provider channel, with a default allocation of 4,000 Copilot credits per user per month. This arrangement is scheduled to roll out gradually during the third quarter of fiscal year 2027, from January to March 2027.
Therefore, the key question Ignite this year must answer is: Can Microsoft prove that Copilot not only has more active users but also continues to increase revenue through seat upgrades and pay-as-you-go pricing?
II. Foundry: Can Microsoft Control the Agent Orchestration Entry in the Multi-Model Era?
If Copilot is responsible for delivering AI to end users, then Microsoft Foundry serves as the platform for enterprise AI application development and deployment.
In its latest fiscal quarter, Microsoft reported that the number of Foundry customers has exceeded 120,000, with related revenue growing more than threefold year-over-year; the number of customers building agent solutions using multiple models has increased fivefold year-over-year, and revenue from open-source models on Foundry has grown tenfold year-over-year.
Morgan Stanley believes that the growing variety of models does not necessarily diminish Microsoft’s platform value. On the contrary, as enterprises need to choose between different models and allocate tasks based on quality, cost, and response speed, a unified model management and orchestration platform may become even more important.
Microsoft also stated on its earnings call that 90% of Copilot queries use Auto mode, with the system dynamically routing between different models to balance quality, speed, and cost—reflecting Microsoft’s integration of multi-model scheduling capabilities into its product ecosystem.
This year’s Ignite is noteworthy for whether Microsoft can further demonstrate tangible progress in Foundry regarding model selection, agent development, deployment, and operational management, and whether these capabilities can drive increased enterprise workloads and cloud service consumption.
III. Fabric: Can Enterprise Data Become a Differentiating Advantage in the AI Era?
As foundational model capabilities become increasingly widespread, a company’s true competitive advantage may increasingly depend on its ability to enable AI to understand its own business data and workflows.
Microsoft Fabric is a key component of this strategy. The platform is designed to unify enterprise data, providing agents with accessible business information and context.
Microsoft's latest fiscal quarter revealed over 44,000 paid Fabric customers, a year-over-year increase of more than 50%, and over 19,000 customers using both Fabric and Foundry. This indicates that an increasing number of enterprises are integrating their data platforms with AI application development platforms.
Morgan Stanley believes that governed enterprise data and business context are key advantages that distinguish Microsoft from mere model providers. If enterprises are willing to migrate more data and workflows to the Microsoft platform, the synergy between Fabric and Foundry is expected to further drive growth in Azure consumption.
IV. The widespread adoption of agents drives demand for security; Agent 365 and enterprise governance warrant attention.
As AI agents evolve from answering questions to executing tasks across applications, managing their permissions to access enterprise data, invoke tools, and perform actions will become increasingly important.
Microsoft is integrating identity authentication, access control, behavioral monitoring, and data protection capabilities into its agent management system. Products such as Entra, Defender, and Purview are expected to see increased enterprise demand, while Agent 365 will assume the role of managing, monitoring, and governing AI agents.
Microsoft previously integrated Agent 365 into the Microsoft 365 E7 plan. During the latest earnings call, Nadella stated that the number of registered agents for Agent 365 has exceeded 50 million, representing over 30% growth quarter-over-quarter.
For Microsoft, the growth in the number of agents not only signifies increased potential demand for cloud computing services but also highlights the need for new security and governance tools. If Microsoft can integrate agent management with its existing enterprise security products, it could further drive customers to upgrade to higher-tier enterprise software packages.
Five: Maia 300: Can Proprietary Chips Reduce AI Infrastructure Costs?
Demand for AI computing power continues to grow, placing significant capital expenditure pressure on Microsoft due to data center construction and chip procurement. Therefore, whether developing in-house chips can reduce per-unit computing costs is a critical factor influencing Microsoft’s long-term profitability.
Microsoft announced in October its next-generation custom AI inference chip, Maia 300, and plans to scale up deployment of the Azure Maia series starting in 2027. The company expects Maia 300 to deliver approximately 40% more tokens per megawatt than its predecessor, with rack-level system token throughput reaching up to seven times that of Maia 200.
However, improved chip performance does not necessarily mean an immediate reduction in overall costs. Investors should still monitor its actual deployment pace, supported workloads, and cost and efficiency advantages compared to third-party GPU solutions.
Morgan Stanley believes that as Microsoft expands its deployment of in-house chips, improved efficiency in AI infrastructure could alleviate gross margin pressure and enhance the long-term returns of the Azure business.
Gross profit margin is under pressure, but can operating leverage offset the cost pressures?
On the other side of Microsoft's AI business experiencing rapid growth is a continuous rise in capital expenditures and infrastructure costs.
Microsoft's capital expenditures for its first fiscal quarter reached $45.2 billion, with approximately three-quarters allocated to short-term assets such as CPUs, GPUs, memory, and storage. The company expects capital expenditures for the second fiscal quarter to exceed $50 billion and anticipates total capital expenditures for the calendar year 2026 to be around $170 billion.
During the same period, the company's free cash flow was $18.6 billion. Microsoft stated that it expects to maintain positive free cash flow in fiscal year 2027, but also noted that it will continue to flexibly utilize financing channels based on funding costs and market conditions.
Morgan Stanley’s model shows that Microsoft’s gross margin may gradually decline from 67.9% in fiscal year 2026 to 65.9% in fiscal year 2027, 64.5% in fiscal year 2028, and 63.5% in fiscal year 2029. However, the firm expects operating margins during the same period to remain relatively stable at 46.8%, 46.6%, 46.7%, and 47.2%, respectively.
This means that Morgan Stanley expects Microsoft to offset some of the gross margin pressure from AI infrastructure investments through revenue growth, improved platform efficiency, and cost control. For investors, what truly matters is not whether gross margin declines in the short term, but whether AI investments generate sufficient revenue and operating leverage to maintain resilient operating margins.
Behind the $650 price target: Valuation depends on the realization of AI commercialization
In terms of valuation, Morgan Stanley maintains an "Overweight" rating and a $650 price target for Microsoft. The bank's base case assumes earnings per share of $24.41 in fiscal year 2028, implying a P/E ratio of approximately 26 times.
Morgan Stanley's bull case target price is $800, based on 2028 fiscal year earnings per share of $27.41 and a P/E ratio of approximately 29x; the bear case target price is $300, based on earnings per share of $22.25.
The row believes that Microsoft is poised to further expand its enterprise AI market share through synergies between Azure, Foundry, Fabric, and Copilot. However, this assessment also implies that Microsoft must consistently demonstrate that the commercialization progress of its AI products aligns with market expectations for growth and profitability.
The ultimate test: Can revenue growth keep pace with capital expenditures?
Behind this year’s Ignite are two simultaneous conditions: first, Copilot must deliver actual revenue growth after transitioning from seat-based subscriptions to pay-as-you-go pricing; second, Microsoft must increase utilization of its AI infrastructure to prevent capital expenditures from continuing to rise without a corresponding improvement in return on investment.
In the first fiscal quarter, Microsoft's capital expenditures reached $45.2 billion, and are expected to exceed $50 billion in the second fiscal quarter. Despite strong growth in Azure and Foundry, investment in AI infrastructure remains substantial. Microsoft needs to demonstrate to the market that each additional dollar of capital spending generates more monetizable computing power and higher revenue returns.
Therefore, the most significant developments on Ignite are not a new model or an agent feature, but three more commercially impactful advancements: whether Copilot can increase ARPU through seat upgrades and pay-as-you-go pricing; whether Foundry can become the core platform for enterprise agent deployment in a multi-model environment; and whether in-house chips and improved data center efficiency can reduce the cost per unit of computing power.
For Microsoft, competition in enterprise AI is no longer just about model capabilities—it’s a full-stack commercial battle spanning infrastructure to applications, data to security. Whether Ignite provides a clearer roadmap for implementation will be a key window for the market to assess the return on its AI investments.




