By Sam Bradley  •  December 11, 2025  •

Ivy Liu

The tools offer the ability match ads to relevant searches — including ones advertisers haven’t bid on. It can customize ad copy based on user intent as well as automatically send users to a brand’s landing page, via search term matching and text and URL optimization. Those features promise improved reach and performance, as well as saving time spent on campaign optimization.

It’s also currently one of the primary ways for brands to get their paid messaging in front of AI Overview or AI Mode users. The tech company has been displaying ads above and below Overview summaries, and in AI Mode results when certain conditions are met — but it’s not a placement advertisers can currently request. Instead, buyers report Google has been selecting ads run through AI Max for inclusion, a little like a prize bingo game.

Google has already experimented by placing shopping and text ads below the first search response shown in AI Mode, in the U.S. for desktop and mobile, and began integrating ads into the first response in the fourth quarter of 2025.

Should those placements become more accessible next year, they’ll be of particular interest to retailers, given the increasing use of AI tools among shoppers.

AI search-driven traffic to retail sites rose 850.7% year-on-year over the Black Friday/Cyber Monday weekend, according to estimates from media agency Wpromote’s Polaris platform. 

Agency media buyers have been experimenting with AI Max since its launch. While some report improvements, others suggest it’s been less useful.

“It’s not delivering what it’s supposed to deliver — that is, long tail queries matched to the right copy and matched to the right landing. I’m not seeing the smartness in the AI as yet,” said one search practitioner, who exchanged candor for anonymity.

A Google spokesperson said in an email that AI Max was the company’s “fastest-growing AI-powered Search ads product.”

”While we offer flexible controls so advertisers can guide the AI, the tool works best when advertisers set clear goals and use the full suite of features to allow it room to find new opportunities beyond a standard keyword list,” they added.

Indie agency Mediassociates has tested AI Max with three clients, each using 5-10% of their search budgets.

”It’s very helpful if you’re for a client that’s looking to get reach,” said Tim Lathrop, vp of platform digital at Mediassociates. “But for especially performance-based clients, it really hasn’t been proven out yet that it’s been effective.”

Google stated earlier this year that non-retail brands activating AI Max in Search would typically see a 14% increase in conversions for the same CPA as their usual search ads. Lathrop said that his team’s testing had found the opposite. “In most cases, unfortunately, it’s two to three times higher than our traditional search campaigns,” he said.

Vitya Vijayan, global head of search and social at M&C Saatchi Performance, told Digiday that in her experience, the feature generated keywords that were so broad — or only tangentially linked to a category — that they weren’t of use to the client. 

Though it’s possible to provide lists of keywords to avoid, doing so adds on time for buyers and brands. ”There’s a lot of optimization that needs to happen… it almost defeats the purpose,” she said.

In particular, some buyers report toggling off AI Max’s AI creative features to take back control of keyword bidding and which landing pages the ads direct users toward.

“The query expansion is not something you can opt in or out of, but the ad copy, the dynamic ad copy and the landing page selection features … disabling that part of AI Max and actually retaining control of that part of the user experience has produced good results,” said Sam Clarke, head of search at Crossmedia.

That’s not to say all experiences have been negative. Indie media shop Roast has been testing AI Max since September. Paid media account manager Patrick Kearney said that while brands already running campaigns against broader keywords hadn’t seen significant improvements, in his experience clients that previously bid on very narrowly defined search queries had found Google’s search matching features useful.

“We’ve been very cautious with AI features… We’ve had mixed successes,” he added.

Other practitioners, exchanging anonymity for candor, praised AI Max’s reporting capabilities. ”With AI Max we can jump in and review the exact queries that triggered or the assets that were generated. This makes performance discussions with clients much easier, and allows us to demonstrate the impact to our clients,” said one agency exec.

In some cases, the mixed signals from testing had caused agencies to step away from AI Max until they see material improvements. “Clients don’t want to sit around … we’ve had many clients that say: let’s move away from AI Max and come back to it,” said Lathrop. He declined to name the clients in question.

With AI Modes, Overviews and Gemini currently terra nova for brands, that reluctance might not last long. Should paid ads within those environments become an accessible prospect next year, brands and buyers might decide that uncertain performance or time required for pro-active management of AI Max tools, are indeed worth it. 

“You have to be willing to part with some tactical and strategic control in certain cases,” said Crossmedia’s Clarke. ”But it’s about where that is going to provide value, versus where that is going to be a hindrance.”

Note: This article was updated to clarify Google’s testing of ads in AI search results.