Google just fired back at critics who blame its AI-powered search features for tanking website visits. The tech giant’s search boss Liz Reid dropped a blog post this week saying click rates stayed “relatively stable” over the past year, directly challenging studies that show AI Overviews cut traffic by half.
Publishers have been screaming about vanishing visitors since Google rolled out AI Overviews in May 2024. Now Reid says those complaints don’t match Google’s internal numbers.
Different Numbers Tell Different Stories
The clash centers on two very different data sets. Google’s internal metrics show what Reid calls “relatively stable” click volume to websites year-over-year. But a March 2025 Pew Research Center study tells a darker story.
When Pew tracked 900 US adults’ browsing habits, they found users click traditional search links just 8% of the time when AI summaries appear. Without those summaries? Click rates jump to 15%, nearly double.
The gap gets worse when you look at clicks on links within AI summaries themselves. Users click those just 1% of the time, according to Pew’s data.
AI Summaries Change How People Search and Leave
Google users behave differently when they see AI-generated answers. The Pew study found people end their browsing sessions 26% of the time after viewing pages with AI summaries. Compare that to 16% for traditional search results.
Reid acknowledges some searches end without clicks. She points to “quick answer” queries like “when is the next full moon” where people get what they need from the AI summary. But she argues this mirrors other Google features like Knowledge Graph panels or sports scores.
Meanwhile, website owners report steep traffic drops. European content creators even filed an EU antitrust complaint against Google’s AI Overviews earlier this year over traffic losses.
Google Says Quality Clicks Actually Increased
Reid’s defense hinges on what Google calls “quality clicks”, visits where users don’t immediately bounce back. She claims Google sends “slightly more quality clicks to websites than a year ago.”
The search executive argues AI summaries help users find what they really want. People might use AI answers to “get the lay of the land” before clicking through to specific sites for deeper information.
Google also says AI Overviews trigger more searches overall. Reid notes people ask “new questions that are often longer and more complex” when AI features are available.
Third-Party Studies Called “Flawed”
Reid dismissed outside research showing traffic declines as having “flawed methodologies.” She specifically targeted studies using “isolated examples” or measuring “traffic changes that occurred prior to the roll out of AI features.”
But website operators aren’t buying Google’s explanation. Helen Havlak from The Verge told NPR that when people see AI summaries, “they visit sites for information less often”
Industry data backs up website owner concerns. Of the top 50 news domains, 37 suffered year-over-year traffic drops after AI Overviews launched, according to traffic analysis firms.
Different Sites See Different Impact
Reid admits the web sees “shifting traffic to different sites” with some losing visitors while others gain. She says sites with “forums, videos, podcasts, and posts” where people find “authentic voices and first-hand perspectives” benefit from current trends.
Google’s data shows certain types of searches trigger AI summaries more often. Questions starting with “who, what, when, or why” produce AI summaries 60% of the time. Longer searches with 10 or more words generate summaries 53% of the time, compared to just 8% for simple one or two-word queries.
Stakes Keep Rising for Web Content Creators
The battle over AI Overviews reflects bigger questions about how search engines should balance innovation with supporting the web ecosystem they depend on. Google processes billions of searches daily and sends massive traffic volumes to websites.
Reid emphasized Google’s “distinct approach” of building AI features “to highlight the web” rather than replace it. She pointed to prominent links, source citations, and inline attribution in AI responses.
But content creators see an existential threat. They argue AI summaries provide enough information to satisfy users without requiring clicks to original sources. That cuts the traffic they need to support content creation and website operations.
The debate will likely intensify as more AI-powered search features roll out. Google insists it wants to support the web ecosystem while improving search experiences. Website owners want proof that support translates into actual traffic they can monetize.
For now, both sides are sticking to their numbers and their very different interpretations of what those numbers mean.
RELATED POST:
