Sign in

MICROSOFT CORP (MSFT) Q3 2025 Earnings Summary

Executive Summary

  • MSFT delivered a clean beat on both revenue and EPS in Q3 FY25: revenue $70.07B vs S&P Global consensus $68.44B, EPS $3.46 vs $3.22; EBITDA also topped consensus. Strength was broad-based, led by Intelligent Cloud and AI, with Microsoft Cloud revenue at $42.4B (+20% YoY; +22% cc). *
  • Azure and other cloud services grew 33% YoY (+35% cc), with 16 points from AI; CFO noted upside this quarter was primarily in non‑AI services, with AI aided by earlier‑than‑planned capacity delivery.
  • Operating margin reached 46% (+1pt YoY); company gross margin was 69% (down 1pt YoY) reflecting continued AI infrastructure scaling. Commercial bookings rose 18% and RPO hit $315B (+34% YoY).
  • Q4 outlook: P&BP revenue $32.05–$32.35B; Intelligent Cloud $28.75–$29.05B (Azure +34–35% cc); MPC $12.35–$12.85B; COGS $23.6–$23.8B; OpEx $18.0–$18.1B; OI&E ≈ -$1.2B; ETR ≈ 19%. FX expected to add ~1pt to total revenue growth; management flagged AI capacity constraints beyond June.

What Went Well and What Went Wrong

What Went Well

  • Microsoft Cloud revenue $42.4B (+20% YoY; +22% cc) and total revenue/EPS beats amid “continued demand for our differentiated offerings.”
  • Azure +33% YoY (+35% cc); non‑AI services drove outperformance with enterprise strength and improved scale motions; AI capacity came online faster than expected.
  • Bookings/RPO momentum: Commercial bookings +18%; RPO $315B (+34% YoY), with ~40% recognizable over the next 12 months (+17% YoY).
  • Quote: “Cloud and AI are the essential inputs for every business to expand output, reduce costs, and accelerate growth.” — Satya Nadella.

What Went Wrong

  • Gross margin pressure: Company GM% 69% (down 1pt YoY) and Microsoft Cloud GM% 69% (down 3pts YoY) due to AI infrastructure scaling.
  • LinkedIn Talent Solutions faced ongoing hiring market weakness; segment commentary flagged headwinds.
  • Windows OEM/devices: only +3% this quarter; Q4 guidance calls for mid‑ to high‑single digit declines as OEM inventory normalizes and devices decline in the high teens.
  • AI capacity constraints expected beyond June, limiting near‑term AI revenue contribution despite robust demand.

Financial Results

Consolidated Financials vs prior quarters

MetricQ1 2025Q2 2025Q3 2025
Revenue ($USD Billions)$65.585 $69.632 $70.066
Diluted EPS ($)$3.30 $3.23 $3.46
Gross Margin %69% 69% 69%
Operating Margin %47% 45% 46%
Gross Profit ($USD Billions)$45.486 $47.833 $48.147
Operating Income ($USD Billions)$30.552 $31.653 $32.000

Segment Revenue

Segment ($USD Billions)Q1 2025Q2 2025Q3 2025
Productivity & Business Processes$28.317 $29.437 $29.944
Intelligent Cloud$24.092 $25.544 $26.751
More Personal Computing$13.176 $14.651 $13.371
Total Revenue$65.585 $69.632 $70.066

KPIs

KPIQ3 2025
Microsoft Cloud Revenue ($USD Billions)$42.4
Azure & other cloud services YoY growth33% (+35% cc)
Commercial bookings growth YoY+18% (+17% cc)
Commercial RPO ($USD Billions)$315 (+34% YoY; ~40% recognized in 12 months)
Company Gross Margin %69% (down 1pt YoY)
Microsoft Cloud Gross Margin %69% (down 3pts YoY)
Cash from Operations ($USD Billions)$37.0
Capital Expenditures incl. finance leases ($USD Billions)$21.4
Free Cash Flow ($USD Billions)$20.3
Shareholder returns ($USD Billions)$9.7 (dividends + buybacks)

Guidance Changes

MetricPeriodPrevious GuidanceCurrent GuidanceChange
Productivity & Business Processes RevenueQ4 2025N/A (Q2 call guided Q3: $29.4–$29.7B) $32.05–$32.35B New Q4 range issued
Intelligent Cloud RevenueQ4 2025N/A (Q2 call guided Q3: $25.9–$26.2B) $28.75–$29.05B New Q4 range issued
Azure Revenue Growth (cc)Q4 2025N/A (Q2 guided Q3: +31–32% cc) +34–35% cc New Q4 target higher than Q3 guide
More Personal Computing RevenueQ4 2025N/A (Q2 guided Q3: $12.4–$12.8B) $12.35–$12.85B New Q4 range; widened due to OEM variability
COGS ($)Q4 2025N/A (Q2 guided Q3: $21.65–$21.85B) $23.6–$23.8B New Q4 range reflects AI scaling
Operating Expense ($)Q4 2025N/A (Q2 guided Q3: $16.4–$16.5B) $18.0–$18.1B New Q4 range
Other Income & Expense ($)Q4 2025N/A (Q2 guided Q3: ≈ -$1.0B) ≈ -$1.2B Lower OI&E expected
Effective Tax RateQ4 2025N/A (Q2 guided Q3: ≈ 18%) ≈ 19% Higher vs Q3
FX ImpactQ4 2025Prior H2 commentary: FX to decrease Q4 revenue >1pt FX to increase total revenue growth by ~1pt Direction changed

Dividends remain ongoing; total capital returns in Q3 were $9.7B (dividends plus repurchases).

Earnings Call Themes & Trends

TopicPrevious Mentions (Q1 2025)Previous Mentions (Q2 2025)Current Period (Q3 2025)Trend
AI/technology initiativesAI business on track to $10B ARR; Copilot adoption momentum; Azure AI capacity constrained; supply pushouts into H2; focus on inference monetization. AI ARR >$13B; Azure growth included 13 pts from AI; capacity constraints; DeepSeek/efficiency narrative; fungible fleet strategy. Microsoft Cloud $42.4B; Azure +33% with 16 pts from AI; non‑AI outperformance; capacity brought online early; AI capacity constraints beyond June. Rising AI demand; near‑term supply tightness; efficiency improving; non‑AI Azure execution improving.
Supply chain/capacityThird‑party supply pushouts; global DC build; constraints (space/power); H2 acceleration expected. Capacity coming online; constrained; H2 acceleration; CapEx pivot toward short‑lived assets. Early capacity delivery aided AI; still tight exiting FY25; short power in specific locations; Q4 CapEx to increase sequentially. Tight near term; build progressing; location‑specific constraints persist.
Product performanceM365 Copilot fastest suite; Dynamics share gains; gaming momentum (CoD). M365 Commercial cloud +16%; LinkedIn +9%; gaming mixed (content up, hardware down). M365 Commercial cloud +12%; consumer cloud +10%; LinkedIn +7%; Xbox content +8%; search ex‑TAC +21%. Broad strength; consumer/gaming contributing; LinkedIn moderation continues.
Regional trendsBroad adoption across industries/regions highlighted qualitatively. Azure non‑AI growth slightly below expectations; scale motion challenges; enterprise acceleration. Non‑AI services accelerated in enterprise segment; improved scale motions; consistent by geo. Improving enterprise execution globally.
Regulatory/legalSecure Future Initiative progress; audit of Copilot interactions (Purview). Security leading; governance for AI deployments. Security Copilot agents; 84T threat signals; identity scale; governance of AI. Security/AI governance narrative strengthening.
R&D executionEfficiency gains and model optimization; hyperscale software leverage. Software efficiencies extend fleet life; margin progress vs prior cloud transition. Model/tooling advances; Phi, BitNet; fine‑tuning reliability; system optimization across stack. Efficiency compounding; expanding model/tool portfolio.

Management Commentary

  • Strategic message: “From AI infra and platforms to apps, we are innovating across the stack to deliver for our customers.” — Satya Nadella.
  • AI efficiency thesis: “For every Moore’s Law change… there’s probably a 10x improvement because of software.” — Satya Nadella.
  • Execution highlight: “Results exceeded expectations driven by focused execution… revenue from our AI business was above expectations.” — Amy Hood.
  • Capacity outlook: “We now expect to have some AI capacity constraints beyond June.” — Amy Hood.
  • Monetization quality: “We’re not selling raw GPUs… we are literally going to the real demand… enterprise and our own products.” — Satya Nadella.

Q&A Highlights

  • Data center strategy and power/location constraints: Management balances build/lease pace, workload shape, and geo distribution; short power in specific places; long lead‑time decisions.
  • Azure growth drivers: Q3 upside skewed to non‑AI services; AI upside from earlier capacity delivery; enterprise acceleration and improving scale motions.
  • Capital efficiency and margins: Efficiency gains across hardware/software shorten dock‑to‑live times; AI margins better than prior cloud transition at equivalent stage.
  • AI demand/inference economics: Focus on inference monetization with diversified demand (enterprise, Copilot, GitHub); training outlays governed by monetization pace.
  • Guidance clarifications: FX adds ~1pt to Q4 revenue growth; Q4 COGS +19–20% cc; OpEx ~+5% cc; Q4 ETR ~19%.

Estimates Context

MetricQ3 2025 ConsensusQ3 2025 ActualSurprise
EPS ($)3.22*3.46 +0.24
Revenue ($USD Billions)68.44*70.07 +1.63
EBITDA ($USD Billions)37.32*40.74*+3.42

Values retrieved from S&P Global.*

  • Beat magnitude was notable and broad-based; the positive delta versus consensus aligns with stronger non‑AI Azure execution and earlier AI capacity delivery.

Key Takeaways for Investors

  • Azure momentum remains robust (33% YoY; 35% cc), with non‑AI services driving Q3 outperformance and AI adding 16 points; expect Azure growth to accelerate to 34–35% cc in Q4, but watch for AI capacity constraints beyond June.
  • Margin dynamics: Operating margin expanded to 46% despite AI COGS drag; near‑term gross margin pressure is likely to persist as AI scaling continues, but management emphasizes software‑driven efficiency gains.
  • Bookings and backlog are powerful leading indicators (RPO $315B, +34% YoY); roughly 40% recognized within 12 months supports near‑term revenue visibility.
  • Copilot adoption is broadening across segments; ARPU lift from E5 and Copilot continues, while seats in SMB/frontline expand—supportive for P&BP durability.
  • Search and advertising ex‑TAC (+21% YoY) and gaming (+8% content/services) contribute incremental margin expansion in MPC; Q4 MPC guidance incorporates OEM inventory normalization and devices declines.
  • Capital intensity remains elevated near term; Q4 CapEx to increase sequentially, with H2 CapEx plan unchanged; mix shifting toward short‑lived assets in FY26, better correlated to revenue.
  • Trading lens: Expect focus on Azure growth cadence, AI capacity commentary at quarter‑end, FX tailwind (+1pt in Q4), and any signs of continued scale‑motion improvement in non‑AI Azure as catalysts for the stock narrative.

Best AI for Equity Research

Performance on expert-authored financial analysis tasks

Fintool-v490%
Claude Sonnet 4.555.3%
o348.3%
GPT 546.9%
Grok 440.3%
Qwen 3 Max32.7%