Microsoft CEO Satya Nadella says AI success is "more about getting intense users and intense usage" than seat counts

Microsoft is posting record profits and strong cloud growth, but just like Google, the company is saying little about how its generative AI business is actually performing.

Microsoft pulled in $82.89 billion in revenue during its third fiscal quarter (ending in March), up 18 percent year over year. Microsoft Cloud hit $54.5 billion, up 29 percent. Azure led the pack with 40 percent growth (39 percent adjusted for currency), and the company expects that pace to hold through next quarter.

Microsoft 365 Copilot now has more than 20 million paying users, up from 15 million in January. Nadella said Copilot has reached the same weekly usage level as Outlook, which he sees as a sign that AI use is becoming "a habit."

Like Google, Microsoft plans to spend $190 billion on capital expenditures in 2026, well above analyst estimates. Even with that buildout, Nadella said capacity will stay tight at least through the end of 2026.

Next quarter's revenue and margin forecast came in below expectations, and the stock dropped more than 5 percent today. Microsoft, like Google, also won't break out its AI numbers: there's no clear picture of how much of Azure's growth actually comes from AI, how profitable Copilot is on its own, or how much direct revenue OpenAI brings in as a customer.

Microsoft shifts to a usage-based licensing model

CEO Satya Nadella already has an answer to the bigger worry that AI could eat into traditional software sales if companies need fewer employees and fewer licenses. On the earnings call, he said, "Any per-user business of ours, whether it's productivity, coding, security, will become a per-user and usage business."

Microsoft has already moved GitHub Copilot to a license-and-usage model. If companies with smaller headcounts using more AI turn out to be just as productive as today's larger teams, lower license revenue could be offset by higher usage fees.

But if AI spending soon matches the cost of several full-time hires at midsize companies, the pressure to prove AI's dollar value will only grow.

Nadella sees it the same way. "At the end of the day, it's going to come from some eval and outcome that a business has, where these agents that are working on behalf of users or with users has created value," he said. That's what drives usage, and that's what Microsoft is chasing. As Nadella put it, Microsoft's AI business "is more about getting intense users and intense usage, and that's what we're focused on."

So far, few companies have pulled that off, partly because measuring the impact of computer-based work is genuinely hard. Local effects, like how fast individual workers move through tasks, are easy to track. Whether and how those gains show up in a company's overall results is a much tougher question.