Google Q3 Earnings Miss as Search Ad Growth Slows; Cloud Division Posts Record Revenue

Alphabet's Q3 results disappointed on headline EPS as internet search advertising decelerated, even as Google Cloud delivered strong above-consensus growth.

• Google Q3 EPS missed estimates as search advertising growth slowed materially
• Google Cloud delivered above-consensus revenue growth driven by AI workloads
• Stock fell on results as market weighted search slowdown over cloud acceleration

Google's Q3 miss crystallizes the structural tension inside Alphabet: search advertising — the historical profit engine — is decelerating faster than expected, while cloud computing is growing faster than expected. The net result disappointed investors who had hoped the cloud acceleration would fully offset the search headwind. The market's reaction — selling on the miss — reflects that search cash flow funds everything from Cloud capex to other bets, making its growth rate a critical variable in any Alphabet valuation model.

The search growth slowdown is partially cyclical (advertiser caution in a high-rate environment) and partially structural (AI search summaries reducing click-through rates to advertised results). Both forces are compressing revenue per search query in a way that is difficult to model with precision. The irony is that Google's own AI investments — the Gemini-powered search features — may be cannibalizing the ad revenue that funds them, creating a strategic dilemma without easy resolution.

Google Cloud's record quarter is genuinely good news and validates the infrastructure build. At approximately $12 billion in quarterly revenue and moving toward meaningful operating profit, Cloud is on a trajectory to become a standalone business of significant scale. The Vertex AI platform, which enables enterprise customers to build on Google's foundation models, is competing effectively against AWS Bedrock and Azure OpenAI. If cloud margin expansion continues at the current pace, it will eventually re-anchor the Alphabet valuation thesis away from search dependency — but that transition takes years, not quarters.