Alphabet's revenue jumps 24.7% on AI demand, but earnings per share miss by 29.6%

Sales on target, profit well short

Alphabet, Google's parent company, reported third-quarter revenue of $127.6 billion, up 24.7% from the same quarter a year earlier. Analysts had expected $127.9 billion, so the top line landed essentially where Wall Street thought it would, as Yahoo Finance reported in its deep dive.

Profit was the problem. Earnings per share, the profit attributed to each share of stock, came in at $2.12 against an expected $3.01. That is a 29.6% shortfall, and the report describes the market's reaction as sharply negative.

One detail does not obviously fit. Operating margin, the share of revenue left after running the business, rose to 31.8% from 30.5% a year earlier. The source does not explain how margin improved while earnings per share missed, so we will not guess.

Where the AI momentum showed up

Management credited AI for the growth across Search, Cloud and YouTube. It pointed to higher usage of AI products and wider adoption of the Gemini 4 series of models, which includes Argon and a newly announced model called Carbon. Argon, the company says, got developers up to speed 3.5x faster than earlier models, while Carbon is described as better at cybersecurity and complex reasoning.

Google Cloud was the standout, with revenue growth of more than 100% year over year. The driver was demand for AI infrastructure, particularly sales of TPU systems. TPUs, or Tensor Processing Units, are Google's specialized computing hardware for AI work. Nearly 90% of Fortune 100 companies used Gemini Enterprise, and more than 100 million seats were sold for AI-driven knowledge work tools.

YouTube added over 30 AI-enabled creator tools and passed 150 million Music and Premium subscribers. Amazon was also added to YouTube Shopping, so creators can tag products directly in their videos.

Search: more engagement, less money per query

Google says AI features such as AI Overviews and AI Mode are leading people to ask longer questions, including ones that mix text with images or other formats. Global Search queries rose noticeably, and engagement is higher.

There is a catch. Management noted that queries in AI Mode earn less than traditional search queries, and that this held back the growth rate of Search revenue. In plain terms, people are using the new experience more, but Google does not yet make as much from it.

That matters because Search is the business that pays for much of the rest. How quickly Google closes that monetization gap is one of the quieter questions behind the numbers.

The bill for building it

Management said rising costs for technical infrastructure and AI investment weighed on profitability. The company also warned of margin headwinds from higher depreciation, which means the cost of expensive equipment is spread across the years it is used, so big purchases keep showing up in the results long after they are made.

CFO Anat Ashkenazi said that infrastructure spending will "continue to weigh on the P&L," meaning the profit and loss statement. This is not presented as a one-quarter blip.

Looking ahead, Alphabet plans to keep rolling out advanced AI models, expand its TPU system business and scale what it calls agentic experiences, which are AI tools built to automate complex tasks for users. It is broadening TPU distribution through partners including Oracle to meet demand from enterprises and AI labs. CEO Sundar Pichai said he expects the next 12 months to bring continued acceleration in AI model development.

What a reader can take from it

The quarter is a snapshot of a trade many companies are making: spend heavily now, and expect the revenue to follow. Alphabet's sales are growing fast, and its Cloud business is more than doubling. Yet the profit line shows what that push costs, and the market clearly noticed.

For ordinary users and small businesses, the direct effect is not yet visible. The source does not say whether Gemini or Google Workspace pricing will change, and any link between this quarter and future prices would be speculation.

What the numbers do show is that the AI tools many people rely on are expensive to run, and their makers are still working out how to earn enough from them. That is a good reason to treat your own AI vendor choices as reversible. This is not investment advice, only a prompt to look at what you pay for and how locked in you are.