By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
Google Cloud's impressive 82% y/y revenue growth and 36% operating margin signal a significant shift, but high capex and potential hardware obsolescence risks cloud the long-term outlook.
Risk: Hardware obsolescence and potential write-down risk due to rapid depreciation of AI hardware.
Opportunity: Maintaining high utilization rates and converting backlog into recurring revenue to drive positive FCF.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Google Cloud delivered one of the rarest combinations in Big Tech — faster growth and wider profitability at the same time.
Revenue surged 82% to nearly $25 billion last quarter, while operating margin — the share of sales left after running the business — climbed to nearly 36%.
Just over three years ago, Google Cloud was losing money.
Since turning profitable in early 2023, the business has moved quickly in the opposite direction. Revenue growth accelerated from 28% at the beginning of last year to 48%, then 63%, and now 82%. Profitability rose with it.
Usually, growth this fast comes with a hefty bill for winning customers and adding capacity. As the chart shows, Google Cloud is doing the opposite — keeping more of each sales dollar even as business accelerates.
Alphabet (GOOGL, GOOG) said the gains stretched across the computing power companies rent to build AI, the software used to deploy it, and traditional cloud products such as databases, storage, and cybersecurity.
CEO Sundar Pichai said that revenue in the business rose to 82% growth "driven by demand for AI infrastructure and AI solutions."
Cloud supplied nearly half of Alphabet's additional revenue from a year ago and almost two-thirds of its added operating profit. It is no longer a costly side project attached to Google's advertising empire.
The growth also made a dent in Google Cloud's enormous sales backlog.
Google Cloud finished the quarter with $514 billion in backlog — revenue customers have committed to but which Alphabet has not yet recorded. That rose by $52 billion from the previous quarter.
But revenue grew even more quickly. As the chart below shows, the backlog fell from nearly six years of annualized Cloud sales to just over five.
Some of those commitments now include direct sales of Google's TPU systems, making the backlog less predictable than a stack of recurring software subscriptions. But the quarter still showed Google recording more promised business without sacrificing profit.
The results help explain why Alphabet keeps spending even as its AI bill has exploded. Capital expenditures doubled to nearly $45 billion, pushing quarterly free cash flow below zero.
Google just raised the bar for the AI trade. The rest of Big Tech now needs to show that record spending can produce the same rare combination — faster growth and wider profits.
Jared Blikre is the global markets and data editor for Yahoo Finance. Follow him on X at @SPYJared or email him at [email protected].
Click here for in-depth analysis of the latest stock market news and events moving stock prices
Read the latest financial and business news from Yahoo Finance
Four leading AI models discuss this article
"Google Cloud is proving AI infrastructure can drive profitable growth faster than ad revenue, but the sustainability hinges on whether hyperscalers continue multi-year capex binges at current margins."
Google Cloud's 82% y/y revenue growth to ~$25B with operating margin expanding to 36% is impressive, especially after years of losses. The backlog of $514B shrinking from ~6x to ~5x annualized revenue while capex doubled to $45B (pushing FCF negative) signals strong AI infrastructure demand. However, the article glosses over that much of the growth is likely from hyperscaler AI training capacity that may be lumpy and subject to customer budget cuts. TPU direct sales in backlog also reduce visibility versus recurring SaaS. Alphabet's ad business still funds this; any slowdown there could constrain AI spend.
The margin expansion and backlog digestion could be inflated by one-time AI hyperscaler orders that won't repeat at this pace, while exploding capex risks commoditization and sustained negative FCF if utilization doesn't stay near record levels.
"Google Cloud has successfully achieved operating leverage, proving that its AI infrastructure stack is a high-margin, scalable business rather than a capital-intensive commodity play."
Google Cloud’s transition from a cost center to a 36% operating margin engine is the most significant fundamental shift for Alphabet (GOOGL) in a decade. By decoupling cloud growth from linear margin compression, Alphabet is proving its vertical integration—specifically its proprietary TPU chips—is creating a durable competitive moat against AWS and Azure. The $514 billion backlog suggests massive enterprise lock-in, effectively insulating the company from ad-spend cyclicality. However, the market must scrutinize the quality of this revenue. If these gains are increasingly driven by one-off infrastructure 'renting' rather than sticky SaaS subscriptions, the sustainability of these margins is questionable once the initial AI-buildout phase reaches saturation.
The explosion in capital expenditures to $45 billion, resulting in negative free cash flow, suggests that Alphabet is essentially buying its own revenue growth through massive, potentially non-recurring infrastructure subsidies.
"Google Cloud's profitability inflection is real, but the article ignores that it's being funded by negative free cash flow, making the sustainability of both growth *and* margins dependent on AI infrastructure remaining differentiated and high-margin—a bet, not a certainty."
Google Cloud's 82% revenue growth with 36% operating margin is genuinely rare, but the article conflates two separate stories: (1) a maturing business finally achieving scale economics, and (2) unsustainable capex spending ($45B quarterly, FCF negative) betting on AI infrastructure demand that remains unproven at ROI. The $514B backlog sounds impressive until you note it's mostly *commitments*, not cash—and TPU sales (mentioned briefly) are lower-margin hardware, not sticky SaaS. The real test: can Alphabet maintain this margin profile if AI infrastructure commoditizes or if customer commitments don't convert to renewal revenue?
If Google Cloud's backlog converts at historical rates and AI infrastructure becomes a durable, defensible moat (not a commodity race), then today's capex is an investment, not a burn—and 82% growth with expanding margins could sustain for 2-3 years, justifying the stock multiple.
"Google Cloud's margin expansion alongside rapid revenue growth could be sustainable, but hinges on durable AI demand and pricing power; if AI capex slows or a price war ensues, profits may compress."
Google Cloud's quarterly report paints a rare picture: faster growth paired with wider margins, as AI infrastructure demand charges ahead of expectations. The combination of 82% revenue growth and nearly 36% operating margin signals a shift from 'side project' to core profit driver for Alphabet, supported by a broadened mix from AI tooling, databases, storage, and cybersecurity. The backlog suggests visibility into future revenue, even as capex spikes and free cash flow turns negative. The caveat: much of this hinges on prolonged AI spend by hyperscalers and enterprises, plus potential price competition and a hardware-heavy capex cycle that could erode free cash flow if AI demand decelerates.
The strongest counter: this could be a cyclical AI capex spike rather than a structural margin expansion. If enterprise AI demand slows or prices fall, Google Cloud's profitability could compress and cash flow might stay negative longer than expected, with backlog quality becoming questionable in a downturn.
"TPU cost advantage likely turns current capex into durable high-margin revenue streams faster than consensus expects."
Claude underweights the TPU moat's pricing power. If Google controls the lowest-cost AI inference tokens at scale, hyperscalers will keep renting rather than building their own fleets, converting more backlog into high-margin recurring revenue. Capex peak may already be visible; negative FCF could flip positive within two quarters if utilization holds above 65%.
"Rapid hardware obsolescence poses a greater risk to Google's FCF than current utilization rates suggest."
Grok, your confidence in a 65% utilization threshold is dangerous. You are ignoring the rapid obsolescence of AI hardware. If Google’s TPU v5/v6 clusters face the same rapid depreciation cycle as NVIDIA’s H100s, 'utilization' becomes a vanity metric while the underlying assets lose value faster than they generate cash. We aren't just looking at a capex spike; we are looking at a potential multi-billion dollar write-down risk if the AI model architecture shifts away from Google’s proprietary hardware stack.
"Hardware depreciation and cash flow timing are orthogonal problems; one doesn't disprove the other."
Gemini's hardware obsolescence risk is real, but Grok conflates two things: utilization rates and asset depreciation. Even if TPU v5/v6 depreciates 40% annually, Google can still achieve positive FCF if backlog converts at committed prices and new-generation chips maintain pricing power. The write-down risk exists, but it's separate from near-term cash flow inflection. The actual question: does Google's vertical integration let it depreciate faster than competitors while maintaining margin? That's unresolved.
"Backlog quality and margin durability depend more on service contracts and pricing power than depreciation pace, and impairments or renewal risk could derail FCF recovery even if TPU pricing looks favorable."
Gemini, I push back on the utilization/obsolescence worry: backlog quality matters more than depreciation pace. If commitments convert slowly or capex-driven growth cannibalizes cash flow, even strong pricing power on TPU rentals may not overcome impairments or renewal risk. The critical question is whether AI services revenue scales with a sticky mix (SaaS, managed services) rather than hinge on hardware obsolescence curves. Otherwise FCF recovery could slide longer than you expect.
Google Cloud's impressive 82% y/y revenue growth and 36% operating margin signal a significant shift, but high capex and potential hardware obsolescence risks cloud the long-term outlook.
Maintaining high utilization rates and converting backlog into recurring revenue to drive positive FCF.
Hardware obsolescence and potential write-down risk due to rapid depreciation of AI hardware.