TAILWIND MONETIZING2025-04-30

Microsoft Corp. Q3-2025

Information Technology · Systems Software

Revenue
$70.1B
YoY Growth
13.0%
EPS
$3.46
Gross Margin
69.0%
Revenue by Segment
Productivity and Business Processes
+10%$29.9B
Intelligent Cloud
+21%$26.8B
More Personal Computing
+6%$13.4B

Operator Signal: TAILWIND

Microsoft Cloud hit $42.4B, up 22% in constant currency. Azure grew 35% cc with 16 points from AI. Non-AI Azure actually outperformed expectations — enterprise migration demand is accelerating alongside AI. Commercial RPO at $315B, up 34%.

Founder Implication

If you sell into enterprise IT, budgets are expanding not contracting. Microsoft is seeing accelerating cloud migration demand alongside AI — your customers are spending more, not reallocating. But capacity constraints mean if you depend on Azure GPU access, expect queues through at least June.

AI Intelligence

MONETIZING

AI contributed 16 points of Azure growth, token processing is up 5x YoY to 100 trillion, and multiple AI products (GitHub Copilot, M365 Copilot, Dragon Copilot, Foundry) have measurable paying customer bases with accelerating adoption metrics.

Mentions: 47AI CapEx: $21.4B
MONETIZING
Microsoft Cloud surpassed $42 billion in revenue, up 22% in constant currency. Cloud and AI are the essential inputs for every business.Satya Nadella
MONETIZING
We processed over 100 trillion tokens this quarter, up 5x year over year, including a record 50 trillion tokens last month alone.Satya Nadella
MONETIZING
Our cost per token is more than halved. We have increased AI performance by nearly 30% ISO power.Satya Nadella
MONETIZING
Foundry is the agent and AI app factory. It's now used by developers at over 70,000 enterprises and digital natives.Satya Nadella
DEPLOYING
Over 10,000 organizations have used our new agent service to build, deploy, and scale their agents.Satya Nadella
MONETIZING
We now have over 15 million GitHub Copilot users up over 4X year-over-year.Satya Nadella

Competitor Intelligence

OpenAI5x mentioned
BULLISH
Commercial bookings driven by an Azure commitment from OpenAI
DeepSeek2x mentioned
NEUTRAL
DeepSeq is sort of added to it and done good work as well
VMware1x mentioned
BULLISH
We remain the cloud of choice for customers' mission-critical VMware, SAP, and Oracle workloads
SAP1x mentioned
NEUTRAL
We remain the cloud of choice for customers' mission-critical VMware, SAP, and Oracle workloads
Oracle1x mentioned
NEUTRAL
We remain the cloud of choice for customers' mission-critical VMware, SAP, and Oracle workloads
ServiceNow1x mentioned
BULLISH
Customers like Coca-Cola and ServiceNow expanding their footprints on Azure
Databricks1x mentioned
BULLISH
Whether it is Databricks or even Snowflake on Azure are growing
Snowflake1x mentioned
BULLISH
Whether it is Databricks or even Snowflake on Azure are growing
Meta1x mentioned
NEUTRAL
New models from Cohere, DeepSeq, Meta, Mistral, Stability, to Foundry
Adobe1x mentioned
NEUTRAL
We offer a growing number of AI apps from partners like Adobe, Canva, and Zoom
Autodesk1x mentioned
BULLISH
Leading partners like Autodesk, PTC, and Siemens have built solutions on our stack
Siemens1x mentioned
BULLISH
Leading partners like Autodesk, PTC, and Siemens have built solutions on our stack

What They Actually Said

It was a record quarter driven by continued strength of Microsoft Cloud, which surpassed $42 billion in revenue, up 22% in constant currency.

Satya Nadella, CEO
growth-signal

We processed over 100 trillion tokens this quarter, up 5x year over year, including a record 50 trillion tokens last month alone.

Satya Nadella, CEO
aigrowth-signal

Our cost per token is more than halved. Model capabilities are doubling in performance every six months thanks to multiple compounding scaling laws.

Satya Nadella, CEO
aipricing

Hundreds of thousands of customers across geographies and industries now use Copilot up 3x year over year. Our overall deal size continues to grow.

Satya Nadella, CEO
aigrowth-signal

The real outperformance in Azure this quarter was in our non-AI business. The only real upside on AI was we were able to deliver supply early to customers.

Amy Hood, CFO
growth-signalguidance

Our margins on the AI side of the business are better than they were at this point by far than when we went through the server-to-cloud transition.

Amy Hood, CFO
aimargin-expansion

For every Moore's Law change, there's probably a 10x improvement because of software. That's what's happening with these models.

Satya Nadella, CEO
ai

We now expect to have some AI capacity constraints beyond June. Demand is growing a bit faster.

Amy Hood, CFO
aiguidanceheadwind

Microsoft Fabric has more than 21,000 paid customers, up 80% year-over-year. Real-time intelligence is the fastest growing workload in Fabric.

Satya Nadella, CEO
growth-signal

We now have over 15 million GitHub Copilot users up over 4X year-over-year.

Satya Nadella, CEO
aigrowth-signal

Forward Guidance

MAINTAINED
Productivity and Business Processes $32.05-32.35B, Intelligent Cloud $28.75-29.05B, More Personal Computing $12.35-12.85B(Q4 FY2025)
MAINTAINED
We expect Q4 revenue growth to be between 34% and 35% in constant currency(Q4 FY2025)
LOWERED
Microsoft Cloud gross margin percentage should be roughly 67% down year-over-year, primarily driven by the impact of scaling our AI infrastructure(Q4 FY2025)
MAINTAINED
We expect COGS of $23.6 to $23.8 billion or growth of 19 to 20 percent in constant currency(Q4 FY2025)
MAINTAINED
Operating expense of $18 to $18.1 billion or growth of approximately 5 percent in constant currency(Q4 FY2025)
MAINTAINED
We expect Q4 capital expenditures to increase on a sequential basis. H2 CapEx in total remains unchanged from our January H2 guidance.(Q4 FY2025)
MAINTAINED
We expect CapEx to grow. It will grow at a lower rate than FY25 and will include a greater mix of short-lived assets more directly correlated to revenue.(FY2026)
MAINTAINED
Even with ongoing AI investments as we scale, we continue to expect full-year FY25 operating margins to be up slightly year over year.(FY2025)
MAINTAINED
M365 commercial cloud revenue growth should be approximately 14% in constant currency, relatively stable compared to the prior quarter(Q4 FY2025)
RAISED
Demand is growing a bit faster. We now expect to have some AI capacity constraints beyond June.(Q4 FY2025+)

Who Ran This Call

DynamicBalanced
CEO Share48%
Satya NadellaCEO
CONFIDENT3177 words
Amy HoodCFO
CONFIDENT3458 words
Jonathan NielsenVP Investor Relations
NEUTRAL465 words
Satya Nadella: Repeatedly used efficiency framing: 'cost per token more than halved,' '10x improvement from software,' 'software is the most malleable resource.' Framed every challenge as an efficiency opportunity.
Amy Hood: Directly corrected analyst assumptions: 'Your question implies something we didn't mean to imply.' Repeatedly grounded AI optimism in non-AI fundamentals: 'The real outperformance was in our non-AI business.'
Dynamic Signal

Amy Hood notably took the lead on correcting analyst narratives about AI being the primary growth driver, steering toward non-AI Azure strength — suggests deliberate messaging strategy to diversify the bull case beyond AI alone.

Evasion Detection

Evasion Score
25%
How stable is Microsoft revenue in a recession?
Mark MortlerSatya Nadella
Redirect40%
Redirected to general cloud efficiency pitch: 'software is the most malleable resource we have to fight any type of inflationary pressure.' No specifics on revenue volatility or recession sensitivity.
How much is non-AI Azure being pulled forward by AI Azure? Could test-time compute change the AI Azure growth curve?
Alex ZukinAmy Hood
Topic Switch35%
Pivoted to 'it's getting harder and harder to separate what an AI workload is from a non-AI workload' — legitimate point but did not directly address either specific question about pull-forward or test-time compute curve impact.

Q&A Intelligence

Questions: 7
Most contested: Azure growth decomposition (AI vs non-AI drivers)
Keith WeissMorgan Stanley
PARTIAL

What's going on with your data center strategy? Are there shifts? Can you address comments about potential GPU oversupply risk?

Signal: Amy: 'We had hoped to be in balance by end of Q4. We saw increased demand, so we will stay a little tight as we exit the year.'
data-center-strategycapexgpu-oversupply
Brent TillJefferies
DIRECT

Can you dig into accelerating demand for cloud migrations?

Signal: Four growth pockets: classic migration, data workloads (Postgres/Cosmos/Fabric), cloud-native compute, and AI workload adjacency ratio.
cloud-migrationazure-growth
Mark MortlerBernstein Research
EVASIVE

How do you think about revenue stability and volatility if we get into a recession?

macro-resiliencerecession-risk
Carl KirstedUBS
DIRECT

Can you unpack the 16-point AI growth lift to Azure? Was it ChatGPT inference surge or enterprise AI adoption? Could it be higher in June?

Signal: Non-AI was the real outperformance. AI upside was only from delivering supply early to customers.
azure-ai-growthdemand-drivers
Cash RanganGoldman Sachs
PARTIAL

Can you slow CapEx growth while still accelerating Azure including AI? Mark to market on capital efficiency?

Signal: AI margins better than at same point in server-to-cloud transition. Efficiency compounding from hardware + software + model diversity.
capex-efficiencycapital-allocation
Mark MurphyJPMorgan
DIRECT

How are DeepSeek-style software efficiencies affecting AI experimentation pace? Could GPU useful life assumptions be extended?

Signal: For every Moore's Law change there is a 10x improvement from software. No change to depreciation policy yet — 'we like a long history before changes.'
model-efficiencyasset-depreciationai-demand
Kirk MaturinEvercore ISI
DIRECT

What drove non-AI Azure upside? How is this cycle different from prior optimization periods?

Signal: Enterprise segment accelerated. Scale motions improved but still have work to do.
azure-non-ai-growthenterprise-demand
Alex ZukinWolf Research
PARTIAL

How much non-AI Azure pull-in from AI? Could test-time compute change the AI Azure growth curve over next few quarters?

Signal: It's getting harder to separate AI from non-AI workloads. Digital natives do AI and non-AI work in the same cloud.
ai-non-ai-relationshiptest-time-computeazure-growth
Topics Analysts Didn't Ask About
Copilot monetization ARPU specificsLinkedIn hiring market weakness trajectoryWindows 10 end-of-support migration timelineTariff impact on hardware/devices beyond inventoryOpenAI relationship economics and equity method lossesGaming margin expansion strategy detailsQuantum computing investment timeline

Audio Analysis

Duration: 56.5 minSpeakers: 7Sentiment: +0.46Chapters: 26
Negative Sentiment Moments (10)
[B] The non-GAAP financial measures provided should not be considered as a substitute for or superior to the measures of financial performance prepared in accordance with GAAP.
[D] The Talent Solutions business continues to be impacted by weakness in the hiring market.
[D] Our losses on investments accounted for under the equity method were slightly higher than expected.
[D] In our on-premises server business, we again expect revenue to decline in the mid-single digits with the ongoing customer shift to cloud offerings.
[D] Windows OEM and devices revenue should decline in the mid to high single digits, We expect Windows OEM revenue to decline in the low to mid-single digits, assuming OEM inventory levels come down through the quarter as noted earlier.

The Number

$77.7B in revenue, up 18% YoY. Every segment grew. Azure up 40%. GAAP EPS $3.72, up 13%. Non-GAAP EPS $4.13, up 23% when you strip out the $3.1B OpenAI investment losses. Microsoft Cloud hit $49.1B with commercial backlog up 51% to $392B. No guidance specifics in the press release — they save that for the call.

What They Actually Said

"Our planet-scale cloud and AI factory, together with Copilots across high value domains, is driving broad diffusion and real-world impact."

— Satya Nadella, CEO [ai] [growth]

Translation: Copilot is landing in enterprise. "Broad diffusion" is exec-speak for "customers are actually using this, not just piloting."

"It's why we continue to increase our investments in AI across both capital and talent to meet the massive opportunity ahead."

— Satya Nadella, CEO [ai] [guidance]

They're accelerating spend, not pulling back. Capex was $19.4B this quarter — 30% more than a year ago.

"We delivered a strong start to the fiscal year, exceeding expectations across revenue, operating income, and earnings per share."

— Amy Hood, CFO [guidance] [growth]

Clean beat across the board. When the CFO leads with "exceeding expectations," the sandbag game is working.

Competitor Intelligence

OpenAI — 4 mentions | Sentiment: CAUTIOUS

- "Non-GAAP results exclude the impact from investments in OpenAI"
- "Losses from investments in OpenAI resulted in a decrease in net income of $3.1 billion"

OpenAI is costing Microsoft $3.1B in losses this quarter alone — up from $523M a year ago. The investment is getting more expensive, not less. Microsoft needs OpenAI to win, but the financial drag is real and accelerating.

Operator Signal: TAILWIND

Azure at 40% growth with $392B in commercial backlog isn't just a good quarter — it's a structural shift. Enterprises are signing multi-year cloud commitments at a pace we haven't seen since the pandemic acceleration. This isn't cyclical spending. It's platform lock-in.

The mechanism: Microsoft is bundling AI (Copilot) into existing enterprise agreements. You don't buy Copilot separately — you upgrade your M365 license. That's why M365 Commercial cloud grew 17% and consumer grew 26%. The AI upsell is working because there's no separate procurement process. It rides the existing contract.

If you're a founder: Enterprise IT budgets flowing to Microsoft are not discretionary — they're infrastructure. If you sell into companies with Microsoft EA agreements, your buyer's wallet just got bigger because Microsoft expanded the budget conversation. Attach to their platform, don't compete with it. The integration play wins here.

What to Watch

  1. How much of Azure's 40% growth is AI workloads vs. traditional cloud migration? Microsoft hasn't broken this out cleanly. If AI is 15+ points of that growth, the traditional cloud business is decelerating.
  1. Will OpenAI investment losses keep growing? $3.1B this quarter vs. $523M last year. At this trajectory, OpenAI could drag $15B+ in annual losses. When does the market care?
  1. Is the $392B backlog converting to revenue on schedule? Backlog growth (51%) is outpacing revenue growth (18%). That's either a great leading indicator or a sign of delayed deployments.