How Google will change with AI, according to SEO experts

“It’s a big change. These are disruptive moments,” says Sundar Pichai, CEO of Google in a recent podcast. He is referring to the arrival of artificial intelligence to search engines. For a few weeks now, Google has been testing AI Overviews in the U.S. for some searches: these are AI-generated paragraphs that try to respond to a user’s search inquiry. “Let Google do the searching for you,” the company says.

The change has not reached Spain or Europe, and there is no timeline for when AI Overviews will be available. But there is a group of Spaniards who are waiting anxiously to see how the changes will work: SEO experts, the workers who try to improve their clients’ page ranking. How will they do their job if Google no longer offers links and directly gives answers? Will their job disappear?

EL PAÍS asked these questions to a handful of digital consultants specialized in SEO. Their collective response was: “Relax, it’s no big deal.” “SEO has died so many times,” says David Carrasco, a SEO consultant. The arrival of AI search engines will not completely eliminate links. In AI Overviews, for example, Google lists the links from which it has sourced its information, adds other links later and also creates a new tab called “web,” where users only links, without any type of AI.

“This has happened before,” says Betlem Cardona, SEO director at an agency. “Everyone believes that we are going to experience a revolution, we throw our hands on our heads and in the end, it’s not the case.” Since the appearance of ChatGPT, in November 2022, the murmur about the end of the traditional search engine has been constant. OpenAI — the creators of ChatGPT — teamed up with Microsoft to, among other things, improve its search engine, Bing. At the time, Microsoft CEO Satya Nadella said of Google: “I want people to know that we made them dance.” In February, there were rumors that ChatGPT was developing its own search engine, which was supposed to be launched in May, but it is still in the works. There are already search engines on the market that work with AI, such as Perplexity, but they have a small market share.

So far, there have been none of the heralded big changes. Pichai even gently responded to Nadella in May: “I think one of the ways you can do the wrong thing is by listening to the noise out there and playing someone else’s dance music,” he said. For millions of people, the Google search engine is their gateway to the internet. The changes on Google are not happening quickly, although in February, the consulting firm Gartner predicted that traditional search engine volume will drop 25% in 2026, as AI chatbots and other virtual agents gain ground.

“They are scared of what others can do, but it has been more than a year and a half since ChatGPT, a year and two months since Bing Copilot, a year and three months since Perplexity and none of them have gained a significant share of Google,” says Juan González Villa, SEO consultant. “Bing made an investment and so far it has barely achieved anything, although 1% of a very large market is a lot. But that doesn’t scare Google that much,” he adds.

Changes of this magnitude take longer. “On a day-to-day basis, this will go more slowly,” continues González Villa. “Google will do it carefully so as to not cannibalize itself. It lives primarily on advertising revenue when people click on links,” he explains. Google has already announced that it will include adds inside AI Overviews. Google does not provide AI Overviews for all search inquiries. In the tests carried out by EL PAÍS and this group of SEO experts, it is more likely that AI Overviews will be used to answer strange searches or questions that could not previously be answered by links (what is the relationship between chimpanzees and astronauts?) and less likely for lucrative Google searches, such as “best coffee maker.”

It will also be more difficult for AI to respond to news or breaking news. AI needs time to compile the overviews and will have a hard time summarizing a soccer match or a bombing blitz while it is happening or just over. The media is, however, one of the groups most concerned by the shift to AI search. Social media is sending less traffic to media outlets, and Google is their last hope. But if AI Overviews mean that users no longer need to click on an article to find out the response to a simple inquiry (for example, “who is that person”), it will have a big impact on the media. On June 7, Business Insider CEO Barbara Peng sent a message to the entire newsroom: “Traffic from Google has been especially volatile across the publishing landscape over the last month,” she wrote. “The way in which people find and access information is changing rapidly [...] When our content is summarized and served up in this way, we don’t make any money to support our journalism.”

These changes have already begun to trigger controversy. In the early days of AI Overviews, users found that Google gave wrong or ridiculous results to some questions. But in a more serious case, Perplexity openly summarized an exclusive news story that Forbes had published behind its paywall. Perplexity has tried to clarify what happened, but regardless, it is the latest case that shows how AI companies grow thanks to the content of others.

The careful approach

Google, as an integral part of the internet, has to be careful. “We cannot assume that in a year there will be no SEO. Google will be careful because 70% of the referred web traffic comes from Google. Imagine if it stopped referring it, such a radical change from Google would not be sustainable,” says González Valle.

But Google has no doubt that AI will bring imminent and important changes. “I was in Google training,” explains Beatriz Tejada, a digital marketing consultant who attended a course at Google’s headquarters in Madrid in May. “In class they suddenly told us [that the traditional method] ‘is dead, it no longer works.’ But I think that right now we are in the transition period.”

Some searches are based on keywords. Now those words will be converted into a series of values that will allow the algorithm to calculate how to best answer a question. “Before it was about looking for a keyword, you did a keyword study based on the client’s niche. Now it’s more about semantically looking for a series of words that are related to the keyword,” says Tejada. This is a way to facilitate responses in AI search.

“We are checking what are the results of the questions asked by our users, the target audience of your clients, and what type of results are activated by artificial intelligence and analyzing the sources,” says Cardona. “This way you have a strategic line to work on, in order to find out what content you have to develop for answers, because we are discovering many questions that we did not know about.”

According to Pichai, another change is that users don’t apathetically accept the AI Summaries: they look for more information in the links. Cardona agrees: “User behavior as Google search results are constructed has a very pinball machine-like behavior. They don’t click on the first three results, but go down, scroll through all pages, click here, go back up, scroll. They have a very different behavior because there are results that are more attractive,” she says.

All these details mean that the real changes depend on how users accept the changes from Google, which is the only one that has all the metrics. “With each update we have known how to be a little more cautious, more skeptical,” says David Carrasco, who admits there is a degree of concern. “If the search engine no longer works as it does now, perhaps five years from now Google will no longer be the main search engine, then we will invent something else to do, we will have to evolve.”

Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition