The referral traffic crisis is keeping many publishers up at night. They claim Google’s AI Overviews, its AI-generated summaries on the search page, is resulting in fewer clickthroughs to their sites. But there are ways publishers can build SEO strategies to try to insulate themselves from the changes.

“I do think the conversation has a lot of doomerism, rather than being focused on identifying opportunities and trying to find new ways to rise to this challenge,” Hyatt said. “This is an opportunity to get things right for your business. As publishers, we should always have been moving towards the ultimate goal, essentially, of bringing folks into our ecosystem.”

Here are the key takeaways from the conversation.

What works for traditional SEO will work for AI

Traditional SEO best practices still apply in the AI era, Hyatt said. He referred to Google’s Core Web Vitals and schema to ensure a good user experience, but also to rank well in search and crawled by AI engines to get surfaced on those platforms. That means The Wall Street Journal focuses on layout, structured content and fast page load speed.

“What works for SEO broadly, will work for those platforms like ChatGPT,” Hyatt said. (The Wall Street Journal’s parent company News Corp. has an AI licensing deal with ChatGPT’s owner OpenAI.)

AI platforms use query fan-out to answer users’ questions, spinning one search query into multiple different queries – and crawling multiple different pages.

“You’re up against hundreds of other websites. So if you want to be crawlable, you want to be fast,” Hyatt said. “If you’re trying to do your best to appear… [in] one of the new AI search engines, then you really want to look at all those broader keywords and topics and and see how you can bring that into your answer, so that you can meet some of those new ways of information gathering.”

But beyond that, The Wall Street Journal is still trying to “figure out” optimizing for AI engines, Hyatt added.

Gather data on what’s getting crawled

The first step any publisher’s SEO team should take is gathering data, Hyatt said. “Look at your log files and see how the crawlers are engaging with your website,” he said. 

Then combine that data with SEO tools like SEMrush, Ahrefs and Profound to paint a picture of what impact this is having on site traffic, and how to optimize around certain queries or topics. In other words, finding a way to meet AI crawlers where they’re looking for content, he said.

“It will never be as good as first-party data, but it could certainly give you a much stronger understanding of your brand’s place in the market, whether it’s for AI Overviews, AI Mode [or other AI platforms]. It will help you get closer to the answer of, what’s our publisher’s place in this ecosystem?” he said. 

For example, if an AI engine is crawling a publisher’s tech section at a much greater rate than any other section, it could mean testing a different strategy for that vertical, depending on a publisher’s goals (trying to or not trying to appear in AI answer engines). 

No more “commodity content”

The Wall Street Journal is leaning away from “commodity content,” or the “stuff you can find anywhere,” Hyatt said. That means staying away from evergreen content. If somebody is looking for a stock price, that’s a low engagement and low intent query, for example. 

It’s difficult to rank first in search results with this content – because of the competitiveness, but also because generative AI features are providing answers to these queries, he said. 

“One of the things that Google is trying to do with its algorithm is direct people to places where they can get first party information that is directly related to what they care about,” Hyatt said. “For those basic queries, the AI platforms want to keep you on there, and they want to answer that question for themselves.”

Focus on the most valuable traffic

Without first party data from Google on how AI Overviews is impacting publishers’ traffic, it’s difficult to paint a clear picture of what’s really happening – even at The Wall Street Journal. 

Liz Reid, Google’s head of search, wrote in a blog post that AI Overviews is driving “more queries and higher quality clicks,” with no data provided to publishers to back up that claim. Declines in traffic could be due to AI Overviews, but core algorithm updates can also impact search referrals, Hyatt said. (He declined to share The Wall Street Journal’s search referral traffic trends during the conversation.)

The Wall Street Journal is focusing on the most valuable clicks, and those are coming from Google search referrals, not Google Discover – despite the growth in Google Discover traffic publishers’ are seeing. That’s why search traffic remains a key part of the reader journey to sign up for a WSJ subscription or newsletter, Hyatt said.

“What we know is that Google organic search is highly intentional – somebody actually went to Google, typed in a specific query,” Hyatt said. “Google Discover is not as valuable as a referrer in the sense that it’s more like social, in how people may come across something that they’re interested in. It’s a personalized algorithm, but it’s not something you intentionally seek out.”

Hyatt said WSJ is focused on the most valuable traffic, and squeezing more out of those platforms.

“[We think] about new ways to connect with readers and bring them in and keep them on platform. That’s the mission,” Hyatt said.