Brands to exercise caution as AI overviews alter online behavior

AI summaries have been found to alter online behavior, according to WARC, but some experts don’t think AI-based search will completely replace traditional search engines in the near future.

The Pew Research Center, a nonprofit organization, first published results of its study in May, but has subsequently released more information about the data. It includes the following takeaways:

Google users who encountered an AI summary clicked on a traditional search result link in 8% of all visits; but those who did not encounter an AI summary clicked on a search result nearly twice as often (15% of visits).

Read more: Sundar Pichai says Google Search's AI Overviews now has over 2 billion monthly users

Google users are more likely to end their browsing sessions when they visit pages that have an AI overview (on 26% of pages with an AI summary) versus 16% of pages with traditional results.

The most frequently cited sources – accounting for 15% of AI summaries – were Wikipedia, YouTube and Reddit.

Google searches that contain more words lead to an AI summary more often: 53% of the time when the search term contains 10 or more words, 8% when it’s just one or two words.

The study, based on an analysis of 900 US adults who agreed to share their online browsing activity in March, specifically looked at Google overviews: a top-line finding is that most people will end their browsing session when they encounter a summary than when they don’t.

Experts debate impact of AI search

Datos, a market intelligence firm, has found that the percentage of traffic going to browser-based AI search (excluding mobile and apps) has more than doubled since June 2024, WARC added. It stands at 5.6% of US search traffic, but it’s still very far off the 94% of traffic that goes to traditional search engines.

Eli Goodman, the chief executive and co-founder of Datos, told the Wall Street Journal that the growth in AI searches could mark “a sea change in online behavior”.

The numbers, he added, are more striking when you look at early adopters of LLMs such as ChatGPT or Perplexity. “Traditional search engines’ share of the desktop browser traffic from these early adopters fell significantly in the same period, to 61% in June from around 76% a year earlier,” reported the WSJ.

Read more: European publishers challenge Google over AI search summaries

But Goodman said that brands should proceed with caution, despite what appears to be rapid change, and that’s because 90% AI searches are “informational or productivity-based”, such as help me solve this problem or answer this question.

Andrew Lipsman, founder of consulting firm Media, Ads + Commerce, believes AI overviews and search are becoming another responsibility for marketers rather than a replacement for search engines.

He added that the spread of advertising in AI search replies “may become the next turning point” and will require new spend from brands that want to remain competitive.

Why it matters

The findings suggest that AI development is rapidly changing online behavior, perhaps faster than previously thought, said WARC. Publishers have long complained that they are under threat because the combination of AI search and overviews has been found to depress website traffic significantly. Some have reported a drop of as much as 70%.

Read more: Google expands AI-driven ads in India, betting on generative search to power growth

Moreover, LLMs reveal just one answer and do not list a series of links in response to a search – and that means there are fewer opportunities for brands to appear in front of consumers.

Brands that have heavily relied on web traffic and traditional SEO are going to have to rethink their tactics for reaching online communities and users, not to mention how they might work with content creators.