Pointchecknote
한국어English

Microsoft vs Meta Q2 2026 Earnings: Azure +43% While Meta's Profit Falls 14%

Breakdown of the July 29-30 reports: Microsoft jumped 10% on Azure's 43% growth and 30M paid Copilot seats; Meta fell 10% as expenses rose 55%. Wall Street now demands AI receipts, not AI plans.

공유𝕏 포스트

Bottom line: the July 29-30 earnings split Big Tech in two — Microsoft surged ~10% while Meta sank ~10%. What separated them wasn’t how much they spend on AI, but whether that spending shows up as revenue. Microsoft posted $90.01B in revenue (+17.8%) with Azure up 43%; Meta grew revenue 28% but expenses jumped 55%, dragging net income down 14%.

The numbers, side by side

Metric Microsoft Meta
Quarterly revenue $90.01B (+17.8%) +28% growth
Headline metric Azure +43%, first $100B annual run Expenses +55% ($42B)
Profit Record Net income -14% ($15.8B)
Direct AI revenue 30M paid Copilot seats Indirect (ad efficiency)
Capex Toward $190B by FY27 Raised to $130-145B for 2026
Stock reaction +10% -10%

Why did only Microsoft get rewarded?

Because the market’s question changed. Last year investors asked “how aggressively are you investing in AI?” This quarter they asked “what is that investment earning?”

Microsoft answered in hard numbers: Azure’s 43% growth is external customers paying for AI infrastructure, and 30 million paid Copilot seats are subscription revenue AI generates directly. Meta’s monetization path — better ad targeting lifting ad prices — is real but indirect, so the spending (+55% expenses) showed up crisply while the payback stayed blurry.

How big is Big Tech’s AI bet now?

Amazon, Google, Meta, and Microsoft plan roughly $725B of combined AI infrastructure spending in 2026, up ~77% year over year. Whether that pace is sustainable is the second half’s defining question — and this week gave the first verdict: judgment is now company-by-company.

It’s the same test behind Nvidia’s $250B OpenAI bet and the circular-financing debate: prove real demand. Microsoft’s formula — cut costs with in-house models (its MAI switch cut GPU costs up to 89%) and recoup through subscriptions — just became the industry’s reference answer.

What to watch next

  • July 31: Amazon and Apple report — does the “AI receipts” test apply again?
  • Meta’s rebound condition: showing AI ad gains as a separable number
  • Whether the four-company $725B capex total is a ceiling or a waypoint

FAQ

Q. If Meta’s quarter was bad, why did revenue still grow 28%? Advertising stayed strong. The problem was cost: expenses grew 55% — faster than revenue — so net income fell 14%. The business performed; the AI bill ate the profit.

Q. Why exactly did Microsoft stock jump ~10%? Azure grew 43% and crossed $100B in annual revenue for the first time, and Copilot passed 30 million paid seats. That’s direct, external, recurring AI revenue — the clearest “receipts” any hyperscaler has shown.

Q. What does this mean for the AI bubble debate? It suggests the bubble question resolves company by company. Firms that show AI payback get re-rated upward; firms still in the “trust our plan” phase get discounted. Selection has begun.


Sources: Axios, Euronews, Forbes earnings coverage (July 29-30, 2026); company filings

#Microsoft#Meta#earnings#Azure#AI capex
공유𝕏 포스트

Related posts