Referral traffic has shifted over the last 12 months and media executives can’t afford to ignore the implications. AI Overviews intercept a growing share of user queries, resulting in zero-click outcomes. Publishers expect traffic from search engines to decline more than 40% over the next three years, according to the Reuters Institute’s 2026 trends and predictions report.

But, across conversations with media leaders at Axios, Hearst Newspapers, Consumer Reports and Forbes, the prevailing approach is measured rather than panicked. What’s happening is less about SEO replacement than an evolution of existing best practices. SEO still delivers traffic, but with an added GEO layer.

While this last year has introduced a new set of acronyms like AEO, GEO, and AIO (see glossary), search teams are pushing past the alphabet soup and finding that the fundamentals of SEO haven’t changed as much as the hype may suggest. AI answer engines still rely on traditional search databases, making authority and clarity advantages for publishers.

“SEO equals GEO. GEO transcends SEO,” says Gideon Grudo, executive managing editor at Consumer Reports, who oversees the SEO and GEO/AIO team. “All the work that we’ve been doing, that we will be doing for search, we will continue to do, because it will benefit us in the answer engines. Study after study, data point after data point continues to prove that if you’re ranking well in search, you are ranking well in the answer engines.”

“There’s news every day about what matters, and what doesn’t matter, and what might be important, and what’s important in this year, and in the next five years,” says Grudo. “And, there are mainstay SEO search foundational realities that have not changed an inch.”

Re-structuring content for machine-readability

Andy Crestodina, co-founder of Orbit Media Studios, frames it in structural terms. In the traditional search model, a user query goes to a search engine, which returns results. Now, he says user prompts go to AI, which queries search, summarizes results, and surfaces an answer. The abstraction layer, which sits on top of the visual internet and allows us to talk to it, is new. The underlying mechanics are not.

“Traditional SEO still matters,” he says. “If you have a page that’s not discoverable in search, it’s usually for one of three reasons: a technical problem, the page isn’t relevant, or you don’t have any links and you’re not a credible website.”

For publishers, the practical implication is that content needs to be more rigorously structured. Bridget Williams, chief product and strategy officer at Hearst Newspapers, is tracking which content is being scraped, what is most frequently retrieved by AI systems using RAG, and how that retrieval correlates with actual clicks.

“These topics, these URLs are scraped the most by OpenAI. These topics are getting surfaced the most. And then these topics are getting the most people clicking through. What does that all mean?” she asks. “It’s very nascent for us.”

Binti Pawa, VP of audience growth and development at Forbes, says AI optimization means making sure that both search engines and AI engines can retrieve Forbes’ content. She notes that the best practices for traditional SEO and for answer engine optimization overlap. “Your content should be well-organized and clear. Content should be for people and not AI agents or bots. That’s not going to change with any of the strategies.”

Pawa believes that SEO and answer engine optimization are complementary, not competing, and that ignoring either one means leaving visibility on the table. “It’s becoming more important to think about your brand and your brand mentions and the citations, which will return a link, which will return a referral, and track it back to your site,” she says.  

Original reporting is a competitive advantage

Structural SEO work only goes so far without editorial to back it up. As information becomes increasingly commoditized by AI, the most successful publishers are doubling down on original reporting and subject-matter expertise, content that is harder for AI to replicate.

Ben Berkowitz, head of news at Axios, says the Axios strategy is to double down on exclusive reporting and analysis from subject-matter experts. “AI prizes original information, so the depth and quality of reporting really is the optimization.”

One of the strategies that can be effective in the AI era is to publish original content and research, Crestodina says. He has noticed that in the traffic from AI sources to websites, the URLs that get the most traffic are the ones that have data points, statistics, or research studies, he explains.

“Most people don’t fully trust AI, for good reasons. And when they look for something backed by data, they’re likely to want to click through and see it,” he says. “They might want to cite the original source. So, when a prompt indicates the visitor wants hard numbers, and the AI response summarizes those numbers, the reader will click through. Because they want to see: is this legit?”

Expanding beyond clicks to visibility and reputation

Publishers say that standard SEO metrics are still tracked, but they are no longer telling the whole story. For media executives, the focus has shifted to holistic KPIs like brand visibility, citation tracking and conversion metrics.

Where traditional SEO focused on what lived on a publisher’s own site, GEO and AIO requires thinking about the brand’s presence across the entire web: third-party mentions, Reddit comments, social signals, and anywhere else a large language model might go looking for evidence of credibility.

“It’s not just what’s on your site, it’s what exists on all these other platforms. So, we’re starting to think about all these other platforms as well,” Grudo says. “That means I want to think about how we exist outside of our site, and what that means.”

Grudo says his team is also looking at where traffic originates. While bot traffic might be minimal right now, as it continues to grow, will it become important, he wonders. “What is intention like from traffic from bots? What’s crawl like from bots? Is that important? How do we keep an eye on that to determine optimization for it?”

SEO teams are now working more closely with social and audience teams. AI models draw on signals from across the web, including platforms like Reddit and YouTube, to assess a brand’s authority and determine how to describe it to users.

Williams notes that Hearst Newspapers subscription base gives it an advantage, with those direct relationships providing a stable foundation. “We are focusing more and more on direct relationships,” she says. “We have this amazing subscription base of people that really care about local news and we are very differentiated in the market.”

According to Pawa, standard SEO KPIs continue to be important to Forbes. But, Forbes has learned it ranks among the top cited sources within certain AI search environments, intelligence that is now feeding into how the team thinks about audience engagement. “Our KPIs are evolving with that,” she says, adding, “Our strategy is shifting to quality and how do we engage loyal users? Do we know enough about our users? How do we engage better with them to actually move them down the funnel?”

In a recent Digiday article, Karl Wells, The Washington Post’s chief revenue officer, said that users coming from AI platforms show 4-5 times higher subscription conversion rates than those from traditional search and tend to spend more time on site.

Wells’ insight may reflect something broader about how AI is changing discovery. Preliminary research from Stanford and Cornell researchers tentatively suggests that LLM adoption may complement rather than replace traditional search, as using AIs lead to a sustained increase in the number of unique websites people visit. The study found users increasingly integrate LLMs and search engines into multi-step, exploratory workflows for tasks, who visit significantly more distinct domains than those using search alone. While preliminary and awaiting peer review, the study hints LLMs disperse attention across a wider, more diverse range of websites.

Same SEO work, new AI search vocabulary

Publishers navigating this transition most effectively aren’t the ones who pivoted quickest to GEO frameworks or overhauled their content operations quickest. They are the ones treating the new environment as an extension of existing discipline: maintaining a strong SEO foundation, structuring content more rigorously, building original reporting that AI cannot easily replicate, and paying closer attention to how their brands are understood and cited across the web.

Berkowitz says, “It’s funny how many times the industry has pivoted content formats to chase whatever algorithm was most rewarding that year. And always, every time, it comes back to ‘just do good original reporting and write it well.’ The publishers that are doing the best these days are the ones embracing Journalism-with-a-capital-J. Nothing’s going to eat your lunch if you have the information and insight no one else has!”

Grudo believes there’s an overemphasis on GEO as a replacement for SEO. “The data keeps showing that that’s a great mistake,” he says. “We keep seeing the LLM companies themselves say that they are mining search databases for responding to queries. They don’t have their own, so they’re relying on what Google and Bing and Yahoo have.”

For other publishers, Pawa cautions that there is a lot of AI advice out there with little to no proof. She suggests they rely on their own data and let best practices guide their strategies.

---

Practical priorities for AI-driven discovery

Across discussions with executives actively shaping SEO and audience strategy, several consistent priorities are coming into focus:

• Structure content for machine readability
• Clear organization and strong SEO fundamentals still determine whether content is retrieved and surfaced by AI systems.
• Track retrieval and citations, not just clicks
• Teams are beginning to measure what content is scraped, retrieved, and cited by AI, and how that connects to traffic and conversion.
• Treat SEO and AI optimization as complementary
• Answer engines still rely on traditional search infrastructure, so strong search performance carries over into AI visibility.
• Invest in original reporting and data
• Exclusive content, research, and proprietary insights are more likely to be surfaced and drive engagement.
• Build authority beyond your own site
• Brand signals across platforms like social, forums, and third-party mentions influence how AI systems assess and describe credibility.