1. Confusing Paid Seats with Active Use
Microsoft announced that Microsoft 365 Copilot surpassed 30 million paid seats in its fiscal 2026 results, which sounds like a home run. But a paid seat is not the same as an actively engaged user. Research shows a notable gap can exist between the number
of licenses a company buys and how many employees use the tool daily. One study found that about one in three paid Microsoft 365 Copilot seats went unused in a typical week. For investors, the critical metric isn't just the seat count, but the intensity of use. Are employees merely trying it out, or are they embedding it into their core workflows? The latter signals long-term product stickiness and a greater willingness for customers to renew, while the former could indicate a future churn problem once the initial hype fades. Real value is created when usage becomes a daily habit, not just a line item on an enterprise agreement.
2. Ignoring the Massive Cost of AI Revenue
Sky-high AI revenue growth is exciting, but it comes at a staggering cost. Microsoft’s capital expenditures—mostly for the data centers and specialized chips needed to power AI—hit $115.9 billion in fiscal year 2026. The company expects to spend over $50 billion in the first quarter of fiscal 2027 alone. This massive investment directly pressures profit margins. Microsoft Cloud's gross margin has already seen contraction, dipping due to the heavy spending on AI infrastructure. An 'AI dollar' is simply more expensive to generate than a traditional software dollar because of the immense power consumption and rapid hardware depreciation involved. Investors focused solely on revenue growth without scrutinizing the corresponding rise in cost of revenue and the impact on gross and operating margins are only getting half the picture.
3. Misinterpreting 'Azure AI' Growth
When Microsoft reports strong growth in 'Azure and other cloud services'—like the 41% jump in fiscal 2026—it’s easy to think of it as a single, monolithic success. However, this figure is a blend of many different services. A huge, and somewhat opaque, component of this is the revenue from OpenAI. For fiscal year 2026, Microsoft disclosed booking $24.1 billion from its partnership with OpenAI, which includes revenue-sharing payments and OpenAI's own massive spending on Azure for computing power. This single customer appears to account for a substantial portion of Azure's total revenue. While this is real revenue, its concentration and unique structure make it different from the organic growth coming from thousands of other enterprise customers. Investors need to ask how much of the growth is from this single, symbiotic relationship versus broader market adoption of Azure's native AI tools.
4. Overlooking the Shift to Consumption-Based Billing
For decades, Microsoft built its empire on predictable, per-user, per-month subscriptions. AI is changing that. The company is increasingly moving to a 'per-seat-plus-consumption' model. This means a company might pay a base fee for Copilot access, but then incur additional charges as its AI 'agents' perform more complex or lengthy tasks. This model can accelerate revenue when usage is high, as seen with GitHub Copilot's growth after it adopted usage-based billing. However, it also introduces volatility. Revenue is no longer a simple calculation of seats multiplied by price; it's now tied to the variable intensity of customer usage. For investors, this makes forecasting more difficult and ties Microsoft's success even more closely to its ability to prove tangible, ongoing value that justifies the metered spending.
5. Equating Headline Announcements with Financial Reality
Microsoft's financial reporting is becoming more complex as it adapts to the AI era. The company announced a major overhaul of its reporting segments starting in fiscal year 2027, collapsing the previous three into just two: 'Agents and Infra' and 'Devices and Consumer'. The goal is to provide a clearer view of its AI-driven model. However, these changes can also obscure trend lines and make year-over-year comparisons tricky for a time. For example, some revenue from developer tools that was once under Azure has been reclassified. An investor who isn't aware of these structural shifts might misinterpret growth in one area without understanding the accounting change behind it. The lesson is to always read the fine print in earnings reports and understand how the company is presenting its numbers, not just what the headline figures are.
