Meta's AI Ad Machine Pulls Ahead as Google Search Stumbles

For two decades, Google and Facebook built the internet's advertising economy around a simple bet: people click links, and companies pay to be the link they click. That bet is starting to look shakier for one of them.

Meta's advertising revenue climbed 26% in the third quarter to $63.1 billion, according to CNBC. Google's Search and other unit grew just 12%, to $63.4 billion. That 14-point gap is the widest it's been in Meta's favor since 2021, and it leaves the two ad businesses roughly the same size for the first time in years — though Google still has YouTube padding its results with an extra $11 billion in quarterly revenue.

The split shows up starkly on Wall Street. Alphabet had its worst trading day ever after reporting third-quarter earnings, with shares plunging more than 12% and erasing roughly $500 billion in market value. Meta isn't having an easy run either — its stock fell 14% in October, its worst month since 2022, following a 27% surge in September. But the underlying stories behind those swings are different. Investors are nervous about Meta's AI spending paying off eventually. They're nervous about Google's core business changing permanently.

The chatbot problem

The reason is straightforward: AI chatbots are becoming a real alternative to typing a query into Google. When people ask ChatGPT, Anthropic's Claude, or Meta's own newly launched Muse assistant to find something for them, they get an answer instead of a page of links — and they don't click through to the websites that Google's ad business has monetized for twenty years.

Nate Elliott, an analyst at Emarketer, put it to CNBC this way: "It's a double-edged sword, because as AI revolutionizes discovery, people might come across tools they like better than Google. If ChatGPT or Claude or Muse is better at helping them find what they're looking for, people's search habits might change for the first time in a generation. That's potentially catastrophic for Google's ad business."

Google isn't standing still. It's pushing AI Overviews and AI Mode, features that answer queries directly inside Google's own search results rather than sending users elsewhere. The company said search queries hit a record in the quarter — people are still searching, in other words — but revenue growth is slowing anyway, because those AI-generated summaries don't drive the same clicks that generate ad dollars. Analysts at UBS wrote that Meta's strong quarter, combined with the looming threat from Muse and other AI agents, is reviving fears about Google cannibalizing its own search business.

Meanwhile, Meta is benefiting on the other side of the same shift. People are spending more time scrolling through Instagram, and Meta has gotten considerably better at using AI to decide which ad to put in front of which user. Mark Zuckerberg has been arguing for a while that AI-driven targeting is the payoff for Meta's enormous infrastructure spending, and the numbers are starting to back him up.

How Meta got here

Meta didn't choose this AI push for fun. It needed it. Apple's 2021 privacy changes to iOS gutted Meta's ability to track users and target ads on iPhones, and the fallout was severe enough that Meta's revenue fell in 2022 — the only year that's ever happened in the company's history.

Meta was also late to the generative AI race itself, trailing OpenAI and Anthropic badly on frontier model development. But where it has made real progress is folding AI into the unglamorous mechanics of ad sales. A year ago, Meta rolled out an AI business assistant that plugs into its Ads Manager tool to help brands build and optimize campaigns. Meta CFO Susan Li said on last week's earnings call that advertisers who adopted the system's recommendations saw almost 5% more conversions and increased their ad spending by 9% compared with similar advertisers who didn't use it.

Li cited syrup maker Simply Organic Maple as an example: using the assistant, the company saw a 57% jump in purchases and a 36% drop in cost per purchase. Nearly 10 million advertisers used at least one of Meta's AI tools during the quarter, according to Li.

Google is chasing the same opportunity on its own platforms. Chief business officer Philipp Schindler said on Google's earnings call that the company is rolling its Gemini models out "across our entire ads infrastructure," building new tools for advertisers and improving ad quality — including on YouTube.

Betting on something other than ads

Both companies are now pouring enormous sums into AI infrastructure, spending so heavily that both have slipped into negative free cash flow. Google is projecting up to $205 billion in capital expenditures this year; Meta is projecting up to $145 billion. That level of spending has investors asking an obvious question: will it generate enough revenue to justify itself?

Part of the answer, for both companies, is trying to not rely purely on advertising. Google has a real head start here — its cloud computing division just posted quarterly revenue of $31 billion, more than doubling from a year earlier, fueled by demand for its cloud infrastructure and its custom AI chips, known as tensor processing units (TPUs). Ads now make up less than two-thirds of Alphabet's total revenue.

Meta has much further to go. Ads still account for 98% of its sales, and its push into subscription-based AI products and agents is still in its early stages.

Analysts at Wedbush see Google's underlying search business, not its diversification efforts, as the thing that actually matters for its stock right now. "GOOGL shares are unlikely to work until the search narrative changes," they wrote in a report Thursday, "and it will likely require new ad products and proof points that it can monetize in the evolving AI era as it did in the past."

That's the real tension sitting underneath this quarter's numbers. Meta's ad business is proving that AI can make advertising more effective. Google's search business is facing the possibility that AI could make its core product less necessary in the first place. Both companies are spending unprecedented amounts of money to make sure they're on the right side of that distinction — and this quarter was the first time investors got hard numbers suggesting the gap between them might be real.