Publishers are recalibrating their SEO strategies one year after Google launched its generative AI search feature AI Overviews.
A publisher is rethinking their SEO strategy and moving away from optimizing for long-tail queries. Another publisher is reinvesting more in exclusives, scoops, and breaking news – prioritizing the urgency of stories over evergreen ones. A third is actively trying to reduce its dependence on Google for traffic.
These changes are being made by publishers after learning what works and what doesn’t since Google’s AI Overviews were launched.
An SEO manager from a lifestyle publisher who spoke under the condition of anonymity said that “the old rules of optimizing content and SEO don’t apply anymore.” “It is getting harder and more difficult to keep up with [search engine results page] the changes that Google keeps throwing our way… It’s a major shift that’s occurring.”
Publishing companies have struggled to measure AI Overviews’ impact on search traffic. As the feature is included in more searches and Google’s AI Mode is in beta, publishers are already looking for ways they can adapt and anticipate a changing landscape. Publishers are not all on the same page. Some publishers are actively pivoting their SEO strategy for a generative AI future, while others bet that their focus on building a direct relationship with their audience via subscriptions and email newsletters will payoff in the long-run and make them less reliant on referral traffic.
Here’s what publishers have learned so far about AI Overviews:
For many years, SEO at publishing companies focused on identifying high performing keywords and creating content around them in order to rank their pages highly in search engines. Some publishers are now refocusing their efforts on keywords that do not trigger AI Overviews in order to preserve visibility and click-through.
This is because the data so far shows that publishers receive lower CTRs when their content is surfaced on the AI Overview compared to other times.
Mail Director of SEO and Editorial E-Commerce Carly Steven told delegates attending the WANIFRA World News Media Congress, in Krakow Last week it was found that the Daily Mail, even though it appeared as the top link on AI Overviews had a lower clickthrough rate of 43.9% on desktop and 32.5% on mobile.
Using keywords to determine which topics and categories are triggering AI Overviews has become a more popular method to figure out what keywords are impacted, and how to adjust SEO strategies accordingly.
Some publishers believe that there is little reason to optimize for AI Overviews, if they are receiving fewer clicks. They’re now focusing on fewer keywords than before. The previous SEO manager stated that “targeting those keywords [that trigger AIOs] isn’t needed.” It’s not worth [anymore] our time to try to rank in Google Search when we can get much more traffic by optimising for Google Discover.
As an example, there is little incentive to go after searches like “How much lemon juice is in it?” when Google’s search page will answer that question with an AI Overviews summary. Other keyword-based searches include those that contain “how long” and “why is.”
Publishers report that certain types of search queries can circumvent AI Overviews.
These keyword are less likely (about 4.8%) to trigger AI Overviews. According to a report from marketing agency Amsive. According to the report, when they trigger an AI Overview they see an average 18.7% increase clickthroughs. This is likely due to users searching for specific events, brand names, and brand loyalty.
This may be the reason why smaller independent publishers, without the same brand power as The New York Times for example, have experienced a rise in clickthroughs. Since AI Overviews began to dominate Google’s search results, traffic has dropped . Marc McCollum is chief growth officer of Raptive, which oversees more than 6,000 independent publishers. He says that AI Overviews are hurting smaller recipe sites and blogs. He said that users can find recipes summarized in the AI Overviews section. This reduces the incentive to click through to a website to get the recipe.
AI Overviews primarily trigger when a user types longer, more complex informational queries such as “What is hurricane?” or “What’s difference between a tornado and a hurricane?” Users can find answers to these questions in an AI Overviews summation without clicking through to an actual article. AI Overviews grew the most in content related to science and health (up around 20%) between January and March 2025. According to a recent Semrush report
The SEO manager stated that they were not seeing the [click-throughs] results from these [long-tail, informational] keyword searches as we used to. [from the ones] That triggers AI Overviews. Although they did not provide any data to support this, they did mention that the impact was so noticeable that their team no longer targeted those keywords.
Google AI Overviews are still a measurement blackhole
Google doesn’t share click-through rates for AI Overviews and doesn’t separate that data from search referral traffic on its analytics dashboards.
Publishers are using third-party tools such as Semrush and Moz to track the impact AI Overviews. Ziff Davis CEO Vivek Sharma said that AI Overviews are triggered by “a little more than 20%” of the top searches from a year earlier.
About a third of the times, we get cited [in AI Overviews]and presumably, citations are a good thing in AIOs. It comes with links. Shah said, “I think [is]is something to watch.” In a recent earnings conference, CEO Neil Vogel revealed that
AI Overviews are displayed for about a quarter of search results relating to Dotdash Meredith content. Dotdash Meredith aims to reduce its dependence on Google referral traffic which accounts for around a third of publisher’s traffic. These calculations are not foolproof.
One news publishing executive who requested anonymity said, “We’re relying on third party tools and that can be hit-or-miss.” Although their news outlet had not seen much of an increase in Google search traffic, or clickthroughs, since the rollout AI Overviews, it was noted that it was difficult to compare apples to apples due to the fluctuations in news cycles.
Publishers are concerned that AI Mode is the true threat to referral traffic.
News providers in general appear to be more protected from Google’s generative AI features so far. Hard news queries don’t seem to trigger AI Overviews. According to a recent analysis by Semrush & Datos, real-time information categories such as news and sport were the least affected by Google Search’s increase in AI Overviews.
However, this may not last for long. Publishers are growing increasingly concerned about how Google’s AI Mode could accelerate the erosion of CTRs they’ve seen with AI Overviews.
AI Mode, powered by Google’s Gemini2.0 AI model, is an experimental search feature that is similar to ChatGPT Search and Perplexity. It allows users to ask follow-up queries without leaving the page. The feature is still in beta, but some publishers who have access to Google’s Search Labs incubator test it. Google has recently begun testing it on its homepage.
A news publishing executive said that recent tests conducted in Google’s AI Mode revealed the search engine could ingest breaking-news content and summarize this information “much faster” than AI Overviews.
They entered some queries into AI Mode regarding a breaking story they hadn’t yet published, and the search engines couldn’t answer any questions about it. AI Mode had “everything available” to it ten minutes after the news organization published its story.
A news publishing executive said that AI Mode will be the real threat to publishers referral traffic once it is widely rolled out. “AIOs are just a test balloon to collect data for AI Mode,” said they.
Google didn’t respond to a comment request before publication time.
