More earnings analysis: https://betafinch.com
Groups: MAG7 (https://betafinch.com/groups/MAG7), AI_LEADERS (https://betafinch.com/groups/AI_LEADERS)
──────────
MSFT Q4 FY26 script.

ALEX: Welcome to Beta Finch, your AI-powered earnings breakdown. I'm Alex.

JORDAN: And I'm Jordan. Today we're digging into Microsoft's fiscal fourth quarter and full-year FY2026 numbers.

ALEX: Before we get into it — this podcast is AI-generated content for educational and entertainment purposes only. Nothing we discuss should be considered investment advice. Always do your own research and consult a qualified financial advisor before making any investment decisions.

JORDAN: Solid disclaimer out of the way, let's talk numbers, because this was a genuinely strong quarter for Microsoft.

ALEX: Big picture first: full fiscal year revenue crossed $331 billion, up 18%, and that growth actually accelerated through the year. Microsoft Cloud alone did over $214 billion, up 27%, and Azure crossed the $100 billion mark, up 41% for the year.

JORDAN: And for the quarter itself, revenue was $90 billion, up 18%. EPS came in at $4.74, up 23% when you adjust for the OpenAI investment impact. There were some one-time items too — a $3.2 billion gain from their Anthropic investment, some savings from a voluntary retirement program, offset partly by Xbox severance and impairment charges.

ALEX: Azure was really the headline. Growth hit 43% this quarter, and CFO Amy Hood said they expect it to accelerate into the mid-40s next quarter. That's remarkable for a business already this large.

JORDAN: Right, and what stood out to me is *why* it accelerated — it wasn't just demand, which is still outstripping supply, by the way. It was efficiency. They shaved dock-to-live time for new GPUs by nearly 50% over the year, and any capacity they free up gets monetized almost immediately because demand is so far ahead of supply.

ALEX: That supply-demand gap is clearly the story of this whole AI infrastructure buildout. They added 31 new data centers this quarter alone, 88 for the year, and they're on track to roughly double total capacity in just two years.

JORDAN: Which explains the capex number — $41 billion this quarter, and guidance for next fiscal year points to over $175 billion, partly due to an accounting change where more data center leases shift from finance leases to operating leases. That's not new spending, just a reclassification, but it's worth knowing so you don't misread the capex jump.

ALEX: Let's talk Copilot, because the seat growth was eye-catching — over 30 million paid Microsoft 365 Copilot seats, and net seat adds more than doubled quarter over quarter.

JORDAN: What I found more interesting than the seat count was the usage data. Conversations per user nearly doubled year-over-year, and time from deployment to high usage across a customer's workforce dropped from months to just days. Satya Nadella also mentioned the number of customers with 50,000-plus seats grew over 7x year-over-year. NHS England alone is rolling Copilot out to over 500,000 clinicians and staff.

ALEX: And they're clearly trying to move the business model beyond simple per-seat pricing. The new E7 suite bundles Copilot, E5, Entra, and Agent 365 together, and they're layering in usage-based billing on top of seats — same playbook they used with GitHub Copilot, where consumption revenue jumped after a June pricing change and Copilot revenue accelerated over 60% quarter-over-quarter.

JORDAN: That model-choice strategy came up a lot in the Q&A too. An analyst asked Satya directly about open and custom models potentially cutting into Microsoft's own frontier-model partnerships, like OpenAI.

ALEX: And his answer was essentially: enterprises want to control their own "learning machine" — keep the harness, memory, and context separate from any single model so they're never locked in. Microsoft's building itself as the neutral platform where that choice happens, whether it's OpenAI, Anthr

This episode includes AI-generated content.