Google hasn’t figured out how to make AI pay like its search engine

Search queries are growing at their fastest rate in five years, Google parent Alphabet said in its earnings report last week. However, Google’s Search revenue growth slowed to 12 percent year over year in the third quarter, down from roughly 17 percent in the previous quarter. Meanwhile, growth in paid clicks — clicks on ads that generate revenue for Google — slowed from 13 percent to 8 percent.

That’s still substantial growth, but not the direction investors were hoping for from a company planning to spend about $200 billion this year on infrastructure, largely to support AI. Alphabet shares have fallen about 14 percent since it reported earnings last week, as investors question the costs and returns of the company’s AI ambitions.

“This is the first major disruption that Google is facing to its core business in decades.”

Chief Business Officer Philipp Schindler said currency effects were the biggest driver of the revenue-growth slowdown, alongside changes in spending by advertisers in particular industries. But Google’s AI makeover also contributed to the slowdown: AI Mode queries were growing strongly but monetizing at a lower rate than traditional searches, Schindler says. He added that Google feels good about its trajectory and that its “ad experiments have shown promising results.”

It remains unclear whether that reflects a deliberate rollout of ads, as Google says, or a more fundamental problem with advertising in AI search.

YOY growth for search revenue vs. paid clicks in Q1-3 in 2026

“This is the first major disruption that Google is facing to its core business in decades,” Nate Elliott, an eMarketer principal analyst covering AI and advertising, tells The Verge. AI is changing how people find information — and whether they click on the ads that fund Google’s business.

What Google’s core Search business looks like increasingly depends on what you ask. In many cases, AI Overviews add generated summaries to search results; electing AI Mode turns searching into a conversation. But someone searching for Nikes can still get familiar shopping listings and text ads, says Andy Taylor, vice president of research at digital advertising agency Tinuiti. They’re looking to buy shoes rather than get an AI explanation of them.

Google’s argument is that AI is expanding the business. Schindler says most AI Mode queries are “incremental” — additional questions rather than replacements for existing searches. Many are also “upper-funnel,” meaning people are researching potential purchases before they know what they want. The company sees those searches as opportunities to sell more ads over time.

Google is currently expanding ads in AI Overviews and AI Mode, including internationally and across multiple turns of US AI Mode conversations. It has also floated bringing ads to its standalone Gemini app, saying formats developed for AI Mode could carry over, but hasn’t announced when.

The thing is, Google, as well as competitors like OpenAI, are still figuring out how to make ads work in AI — both delivering results for advertisers and avoiding alienating the people using their AI tools.

For now, advertisers are still getting results from Google. Tinuiti’s same-store advertisers increased their Google paid search spending 14 percent year over year in the third quarter while maintaining their return-on-ad-spend targets.

“If people are clicking through less often. That is a challenge that they need to figure out.”

“We’re still seeing a good return, and we’re not seeing performance take a hit,” Taylor says.

The catch is that advertisers can’t isolate how Google’s AI ads are doing. Google doesn’t let them separately target that inventory or break out its performance, Taylor says. Steady overall returns could mean AI ad traffic is still too small to noticeably affect the results, or that it performs similarly to existing inventory. Taylor suspects it’s mostly the former.

That leaves an important question unanswered: How much will AI add to Google’s advertising business, and how much could it replace? Additional queries could bring in new revenue even if each earns less than a traditional search. The risk is that AI Mode also draws people away from searches that Google monetizes more effectively. Google’s claim that most AI Mode queries are additional doesn’t tell us how much revenue it could lose from the searches AI Mode replaces.

“If people are clicking through less often,” Elliott says, “that is a challenge that they need to figure out.”

The promise of conversational advertising is that people share much more about their needs than they would in a few keywords. But platforms still have to identify which parts of a conversation matter and match them with the right ads.

“It’s not a slam dunk that it’s going to turn into more effective advertising,” Elliott says.

An early study of ChatGPT ads suggests richer conversations don’t automatically translate into more retail sales. An October working paper found that US visits from ChatGPT ads generated 64 percent less revenue for retailers per session than visits from Google Search ads: $0.89 versus $2.45. The working paper measures retailers’ sales rather than platform revenue, and its findings about ChatGPT don’t establish how Google’s own AI ads perform.

Taylor cautions that those comparisons depend on what people are searching for. Someone searching Google for Nike shoes may already be ready to buy, while someone asking an AI tool which running shoes to choose is still researching. A higher sales return could reflect that difference in purchase intent rather than more effective ads.

Google’s existing business has given it room to take its time with AI ads, Taylor says. He points to its successful shift to mobile: Early phone traffic was less valuable to advertisers because shopping on a phone was cumbersome, but that changed.

Elliott cautions that finding effective ad formats can take years. Search engines started with banner ads before adopting paid listings; social media went through its own experimentation. Today’s formats may not be the ones that ultimately work.

But Google’s mounting AI spending raises the stakes. Search revenue is still growing by double digits, but the company has yet to show how well its new way of searching can pay.

“They absolutely do not have it all figured out,” Elliott says. “No one has it all figured out.”

Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.

• Rani Molla