Media Briefing: Less clicks, more problems: What Google’s AI Mode means for publishers

On Wednesday, Google announced that it will integrate ads into the AI Mode experience, beginning with search and shopping ads for U.S. users in AI Mode on desktop and mobile. It’s just another sign that Google’s vision for the future of search is within its generative AI experiences, which it will now be monetizing.

Publishers are still reeling from the news, and the path forward is still being determined. AI Mode only launched in beta two months ago, meaning heads of audience and SEO haven’t had much time to test the feature.

Their main concern: Once AI Mode rolls out widely, it will accelerate the slow erosion of click-through rates publishers have seen already on Google Search, thanks in large part to AI Overviews, Google’s generative AI search feature that summarizes answers to queries, above search results links to websites. 

“This is the future of Google Search,” Google VP and head of search Elizabeth Reid said about AI Mode, during the company’s annual I/O developer conference keynote on Tuesday. “You’re starting to see this come to life already with AI Overviews and AI Mode takes this to the next level.”

AI Mode will likely lead to fewer click-throughs from Google

The main concern publishers have with AI Mode is that it will erode referral traffic and clicks-throughs from Google search, jeopardizing their ability to monetize with ads on their sites. That’s because AI Mode answers users’ questions directly with AI-generated summaries — eliminating the need to click through to other sites. Publishers have been (rightly) uneasy about search traffic erosion since Google’s AI Overviews launched a year ago. But some believe its AI Mode that will prove the biggest threat.

Publishers including Dotdash Meredith and other large digital publishers have seen traffic impacted by the rollout of AI Overviews, for example. One publishing exec told Digiday, under the condition of anonymity, that they were seeing a 1-4% decline in page views since the debut of AI Overviews.

“The proportion of clicks coming from AI Mode will be less than on search,” said Tom Critchlow, Raptive’s evp of audience growth. “The number of queries people turn to Google for will increase but clicks out will decrease.”

While AI Mode does attribute the information it uses in its responses with links to sites, those links are less visible than the ones on Google’s traditional search results page, execs told Digiday. Six publishing execs interviewed for this story said AI Mode will chip away at publishers’ Google Search referral traffic.

“While Google says it’s working on UI updates to encourage clicks and provide new opportunities for sites to rank, the reality is that comprehensive AI responses often satisfy user intent directly on the search page, leading to more ‘zero-click’ searches,” Paul Hood, AI media consultant and former News UK exec, said in an email.

Google claims that AI Mode’s “fan-out technique,” which breaks down search queries into subtopics, includes links to websites for users to dive deeper, said Reid during the I/O conference. “We believe AI would be the most powerful engine for discovery that the web has ever seen, helping people discover even more of what the web has to offer, and finding credible, hyper-relevant content,” she said.

Google isn’t sharing AI Mode click-throughs, user adoption unclear

Google is holding click-through data from its AI search features close to the vest. Digiday has previously reported that not having click-through data from AI Overviews makes it difficult for publishers to measure its impact on their Google referral traffic.

AI Mode is no different. Publishers don’t have access to AI Mode click-through data in their Google analytics dashboards, making it impossible to know how many site visitors are coming from the generative AI search experience, and difficult to see what content is showing up there. A Google spokesperson said the company’s guidance to website owners has remained the same, which is to “follow best practices for making sure search engines can understand your content, and focus on creating content that is helpful for people (not made to rank well in search engines).”

For now, limited user adoption offers a bit of breathing room for publishing execs still digesting the arrival of AI Mode.

“It’s ultimately going to come down to the user and if they are satisfied with the result provided, or want to continue their journey via more content [or] information,” said Wes Bonner, svp of marketing and audience development and head of social at Bustle Digital Group.

AI Mode leans heavier on publisher content — but offers less in return

Certain sites are more vulnerable to the rollout of AI Mode, such as food and recipe sites and news publishers. 

Critchlow said some food bloggers hadn’t seen their recipes appear in AI Overviews yet. But in Raptive’s tests with AI Mode, when prompted, the AI search engine generated full recipes from those food bloggers’ sites, without clear links to those sites, he explained.

A news publishing exec previously told Digiday that AI Mode was able to provide information on a report their news outlet had covered, just 10 minutes after it had published a story with the details of the report. However, hard news queries don’t seem to trigger AI Overviews.

AI Mode also answers longer, more complex queries, Google CEO Sundar Pichai said during the I/O event. Users’ queries in AI Mode are two to three times the length of traditional searches, he said.

Hood believes this poses a greater risk for publishers. As the feature offers a richer, more conversational search experience, publishers may have a tougher time attracting users to their sites for more information, and hurting their digital advertising businesses.

So what’s a publisher to do when a significant portion of their traffic comes from Google Search?

“Try to diversify away from Google as a traffic source,” said Critchlow.

Easier said than done. But it does seem to be a growing trend among publishers. The news publishing exec said their company was increasingly focused on tracking metrics like subscription conversions, rather than raw traffic, to measure engagement success. Dotdash Meredith’s approach is to actively reduce its reliance on Google referral traffic, CEO Neil Vogel said in a May earnings call.

“The real question for publishers is how consumers of news will engage with this new, soon-to-be personalized way of searching,” said a third publishing exec, under the condition of anonymity.

It does remain to be seen. But in the meantime, none of the six publishing execs interviewed for this story had figured out any sort of strategy for AI Mode optimization yet. 

As one publishing exec who requested anonymity put it, “We have no data on it at all.”

What we’ve heard

“I reckon the impact to us on sessions is between about… 1.5-4% lower sessions, compared to if AIOs had never appeared. The impact on revenue… is a bit less, because [the search] terms [with higher click-through rates]… are less affected [by AI Overviews]. It’s enough to be a headwind that we’re aware of, but not catastrophic at this point.”

— A publishing exec on the decline in Google Search referral traffic as a result of the introduction of AI Overviews. 

Numbers to know

5%: The increase in U.S. digital advertising revenue for U.K.-based publisher Future, after advertisers delayed or pulled campaigns in the latter part of Q2 due to macroeconomic uncertainty.

1.5 billion: The monthly users of Google’s AI Overviews in over 200 countries.

10%: The growth in Google usage for the types of queries that trigger AI Overviews.

1.2 million: The number of subscribers to The Economist.

What we’ve covered

What’s in and out in the era of Google’s AI Overviews

• Publishers have had little control within the Google ecosystem, a fact illustrated by Google’s admission that it can crawl and index content without publishers’ permission.
• In the era of its generative AI search features, things like driving content and click-through rates are out, things like grim traffic charts shared in publisher Slack channels and worrying about “GEO” are in.

Read the definitive list here.

How publishers are testing agentic AI

• Publishers are starting to explore how agentic AI could help them work smarter, faster and more efficiently.
• Companies like Hearst, Thomson Reuters and DPG Media are testing agentic AI in areas like ad sales and other workflows.

Read more about what publishers are doing with agentic AI here.

Video is making podcasts a premium buy

• As podcast consumption shifts to video, advertisers are treating the format as more premium than ever.
• Ad revenue at podcast production companies, such as Audacy, Wondery and Pave Studios, increased year over year between the first quarters of 2024 and 2025.

Read more about advertisers’ eyeing video podcasts here.

Lessons from publishers after one year of Google’s AI Overviews

• One year after the launch of Google’s generative AI search feature, AI Overviews, publishers are recalibrating and rethinking SEO strategy.
• Publishers have struggled to measure the impact of AI Overviews on search traffic. But they’re looking for ways to anticipate and adapt to a changing search landscape.

Read more about what publishers have learned about AIOs here.

Creator start to creep into the main stage at the Upfronts 

• Companies like TelevisaUnivision, Fox and Amazon featured creators during their Upfront presentations to the ad market.
• Traditional media sellers are trying to reposition their offerings not as just airtime but as cultural currency.

Read more about the creator economy’s role in the TV and streaming upfront marketplace here.

What we’re reading

News Corp introduces mandatory AI training for journalists

News Corp will run mandatory AI “bootcamps” for journalists across its Australian newsrooms in an effort to push adoption of the technology, Capital Brief reported. 

Local newspapers’ AI-generated content hallucinates fake books for summer reading lists 

The Chicago Sun-Times and the Philadelphia Inquirer published syndicated AI-generated content that created fake books titles for summer reading lists, The Washington Post reported.

New CJR reports on how platforms and publishers are navigating the AI era

A new report by the Columbia Journalism Review — featuring interviews with news executives, editors, current and former platform executives, and AI experts — highlights the high-stakes fight over copyright and IP rights in the AI era.

A Q&A with Google’s vp for news Richard Gingras ahead of his retirement

Nieman Lab interviewed Google’s longtime global vice president for news, Richard Gingras, and his thoughts on Google’s role in supporting news businesses and the new AI search landscape.

The New York Times’ opinion section doubles in size

The New York Times’ opinion section has doubled in size under the leadership of Kathleen Kingsbury, at a time when other publishers like The Washington Post and The Los Angeles Times shrinks theirs, the Intelligencer reported.