Microsoft Corp (MSFT) (Q4 2026) Earnings Call Highlights: Record Revenue Surpasses $331 …

This article first appeared on GuruFocus.

Revenue: $90 billion for the quarter, up 18% year-over-year; full fiscal year revenue surpassed $331 billion, up 18%.

Microsoft Cloud Revenue: $59.3 billion for the quarter, up 27%; full fiscal year surpassed $214 billion, up 27%.

Azure Revenue: Grew 43% for the quarter; full fiscal year surpassed $100 billion, up 41%.

Operating Income: Increased 18% for the quarter; full fiscal year increased 21% to more than $155 billion.

Earnings Per Share (EPS): $4.74, an increase of 23% when adjusted for the impact from the investment in OpenAI.

Gross Margin Percentage: 67% for the quarter, down year-over-year.

Operating Expenses: Increased 10% for the quarter.

Capital Expenditures: $41 million for the quarter.

Cash Flow from Operations: $55.4 billion, up 30%.

Free Cash Flow: $19.6 billion.

Commercial Bookings: Grew 18% when excluding the impact from OpenAI.

Commercial Remaining Performance Obligation (RPO): Grew 84% to $678 billion.

Productivity and Business Processes Revenue: $37.8 billion, up 14%.

Intelligent Cloud Revenue: $39.3 billion, up 32%.

More Personal Computing Revenue: $12.9 billion.

M365 Commercial Cloud Revenue: Increased 16% on an adjusted basis.

M365 Consumer Cloud Revenue: Increased 24%.

LinkedIn Revenue: Increased 12%.

Dynamics 365 Revenue: Increased 13%.

Search Advertising Revenue (ex-TAC): Increased 10%.

Xbox Content and Services Revenue: Decreased 10%.

Windows OEM Revenue: Decreased 5%.

Release Date: July 29, 2026

For the complete transcript of the earnings call, please refer to the full earnings call transcript.

Positive Points

Microsoft Corp (NASDAQ:MSFT) reported a record fiscal year with annual revenue surpassing $331 billion, up 18%, and Microsoft Cloud revenue exceeding $214 billion, up 27%.

Azure revenue grew 43% in Q4, driven by strong demand and efficiency gains, with expectations for continued acceleration in H1 FY27.

Microsoft 365 Copilot paid seats surpassed 30 million, with net seat adds more than doubling quarter over quarter, and user satisfaction scores doubling over the last three quarters.

The company is expanding its total addressable market by introducing usage-based billing models for Copilot, Cowork, and GitHub Copilot, alongside per-seat licensing.

Capital expenditures are expected to grow in FY27, reflecting strong demand signals, with a focus on efficiency gains and monetization of new capacity.

Customer demand for Azure continues to exceed available capacity, leading to supply constraints that could limit near-term growth.

Xbox content and services revenue decreased 10% year over year, impacted by a strong prior-year comparable and impairment charges.

Windows OEM and Devices revenue is expected to decline in the high-teens for FY27 due to lower PC market demand and higher component costs.

Operating margins are expected to decline slightly in FY27 due to continued investments in AI infrastructure and R&D.

Foreign exchange is expected to decrease total revenue growth by less than 1 point in Q1 FY27, adding headwinds to financial performance.

Q & A Highlights

Here are the key highlights from the Microsoft Corp (NASDAQ:MSFT) fiscal year 2026 fourth-quarter earnings call, focusing on the most significant Q&A exchanges.

Q: Satya, you spoke about model choice and the protection of corporate IP. How material could traction be for open and custom models, and how does Microsoft benefit given your Frontier Lab exposure? A: (Satya Nadella, Chairman and CEO) The goal is for every firm to control its own destiny by building its own “learning machine.” Models are an input, not an extraction of enterprise knowledge. Our platform architecture is designed with a separate “harness” (memory, context) from the model, making any model swappable. This allows enterprises to use Frontier models for their strengths, low-cost models for efficiency, and even train their own models. This design pattern, which we use in Copilot and GitHub Copilot, is what we are democratizing. It ensures resilience and cost control, as no single model can hold a company hostage.

Q: Amy, Azure accelerated to 43% growth, heading to the mid-40s. What is driving this, and are you still capacity-constrained? A: (Amy Hood, CFO) Yes, demand continues to exceed available supply. The acceleration was driven by significant efficiency gains across our CPU and GPU fleet, allowing us to get more out of existing infrastructure. Process improvements also reduced the lead time for bringing new capacity online. Because of the supply-demand imbalance, these efficiency gains were quickly monetized within the quarter. This dynamic, along with stronger-than-expected GitHub Copilot consumption after its business model change, drove the strong results.

Q: How does Microsoft protect itself if there is an oversupply of data centers or chips, and how do you manage through hardware price increases without impacting margins? A: (Amy Hood, CFO) Our CapEx is heavily weighted toward short-lived assets (CPUs/GPUs), which have shorter lead times. If demand changes, we can simply slow down that component. Investments in land and buildings are a smaller percentage and have flexible timing. Our diverse book of business by geo, segment, and workload, along with a large first-party app business that uses the same capacity, provides flexibility. On pricing, we focus on efficiency to deliver customer value, and the cloud still offers a better ROI than on-premises purchases facing the same price increases.

Q: M365 Copilot had a very strong quarter with over 30 million paid seats. How are you seeing customers move from pilots to broader deployments, and what is the main driver of monetization? A: (Satya Nadella, Chairman and CEO) The product itself is evolving rapidly into a “super app” combining chat, Cowork, Autopilots, and code. Time from license purchase to high usage has collapsed from months to days, and usage intensity is now on par with Outlook and Teams. The enterprise wiring is key, with Copilot integrated into governance (Agent 365), IT/SecOps, and business processes (CRM, ERP). The business model is also expanding from per-seat to “seat-plus consumption,” expanding the TAM significantly as we deliver more enterprise-wide value.

Q: Satya, with the introduction of Project Perception, what does this moment mean for your cybersecurity business and for trust in Microsoft? A: (Satya Nadella, Chairman and CEO) The entire physics of cybersecurity has changed. We are taking a model-forward approach with Perception, creating an agentic system of red, blue, and green teams that continuously operate to find, triage, and fix vulnerabilities. This system leverages all our signals (identity, network, app security). Crucially, it uses a multi-model approach for cost and resilience, using a smaller, efficient model (MAI-Cyber-1-Flash) for 90% of tasks and a Frontier model for the rest. This ensures operations can continue even if one model fails, which is critical for trust and resilience.

Q: Amy, can you put the CapEx and monetization pieces together? How does your view on ROI for these investments compare to a year ago, and what levers can you still pull? A: (Amy Hood, CFO) My confidence in the ROI has increased due to TAM expansion and the number of margin levers available. These levers include continued improvements in first-party silicon (Maia), model diversification (using the most efficient model for the task), and the broad portfolio of high-value applications (knowledge work, coding, security, agents). All of these contribute to better cost structures and increased confidence in the return on invested capital. It is a daily “grind” of getting a little better, which we are quite good at.

Q: Satya, you mentioned the new “Frontier Core” engineering organization. Can you elaborate on how this will help customers turn AI into measurable outcomes? A: (Satya Nadella, Chairman and CEO) The most valuable data is inside each customer’s tenant. Frontier Core embeds 6,000 industry and engineering experts with customers to co-design and continuously improve AI systems. We’ve tested this model on over 330 projects, like helping Novo Nordisk analyze clinical data and partnering with LSEG to embed AI into their workspace. This is about turning customers’ workflows, domain knowledge, and judgment into AI systems that learn and improve with every usage, ensuring they don’t outsource their core IP.

Q: Amy, you mentioned a change in the estimated useful life of data centers. Can you explain the impact of this change on CapEx and financials? A: (Amy Hood, CFO) Effective FY27, we are extending the useful life of data centers and office buildings from 15 to 25 years. This change only affects the timing of future depreciation, with a minimal benefit to FY27 operating income. The greater impact is on CapEx, as more future data center leases will shift from finance leases (included in CapEx) to operating leases (not included). This adjusts our calendar year 2026 CapEx expectation to approximately $175 billion, though our underlying investment plans remain unchanged.

For the complete transcript of the earnings call, please refer to the full earnings call transcript.