A groundbreaking study by Pew Research Center has provided concrete evidence of what publishers and SEO professionals have long suspected: Google AI Overviews is significantly eroding the web ecosystem by dramatically reducing traffic to websites. The findings reveal a troubling shift in user behavior that threatens the foundation of how content creators monetize their work.
Study Methodology: Real User Data Tells the Story
Pew Research Center conducted an extensive analysis tracking the browsing patterns of over 900 consenting adults throughout March 2025. Using sophisticated online browsing trackers, researchers gathered data from 68,879 unique Google search queries, with 12,593 queries displaying AI-powered summaries.
This methodology provides unprecedented insight into actual user behavior rather than relying on theoretical projections or Google’s own internal metrics.
The Numbers Don’t Lie: AI Overviews Slash Website Clicks
The research delivers stark confirmation of publisher concerns about declining referral traffic from Google’s AI-enhanced search results. When users encounter AI summaries, they demonstrate significantly different browsing patterns compared to traditional search experiences.
Click-Through Rates Plummet with AI Summaries
The data reveals a dramatic disparity in user engagement:
- Only 8% of users who saw AI summaries clicked any link to visit a website
- 15% of users clicked through when seeing standard search results only
- This represents nearly a 50% reduction in website visits when AI summaries appear
Citation Links Remain Largely Ignored
Perhaps most concerning for content creators, users almost completely ignore the source links embedded within AI summaries. A mere 1% of users clicked citation links in AI overviews to visit the original websites that provided the information.
User Behavior Shifts Away from Web Exploration
Contrary to Google CEO Sundar Pichai’s recent assertions that AI search features encourage greater content consumption, the Pew research demonstrates the opposite effect. Users are actually engaging less with web content when presented with AI-generated answers.
Sessions End More Frequently
The study reveals that AI summaries prompt users to abandon their search sessions entirely:
“Google users are more likely to end their browsing session entirely after visiting a search page with an AI summary than on pages without a summary. This happened on 26% of pages with an AI summary, compared with 16% of pages with only traditional search results.”
Traditional Results Get Overlooked
When AI summaries appear, users become less likely to explore conventional search listings below them:
“Users who encountered an AI summary clicked on a traditional search result link in 8% of all visits. Those who did not encounter an AI summary clicked on a search result nearly twice as often (15% of visits).”
The Google Ecosystem Trap
A particularly alarming finding shows that approximately 66% of all Google searches result in users either browsing elsewhere within Google’s properties or leaving entirely without visiting any external websites. This creates a closed-loop system where Google retains users while the broader web ecosystem loses potential visitors and revenue.
The research states: “…the largest share of Google searches in our study resulted in the user either browsing somewhere else on Google or leaving the site entirely without clicking a link in the search results. Around two-thirds of all searches resulted in one of these actions.”
Dominant Platforms Strengthen Their Hold
The study confirms another persistent concern among publishers: the increasing dominance of a select few websites in both AI summaries and traditional search results. Three platforms consistently monopolize visibility:
The Big Three: Wikipedia, YouTube, and Reddit
These platforms collectively accounted for:
- 15% of sources cited in AI summaries
- 17% of sources in standard search results
The research notes specific patterns in how these platforms appear:
- Wikipedia links occur more frequently in AI summaries
- YouTube links appear more often in traditional search results
- Reddit maintains strong presence across both formats
This concentration of traffic among a handful of major platforms leaves smaller publishers and content creators fighting for increasingly scarce visibility and traffic.
What This Means for Publishers and SEOs
The Pew Research findings validate longstanding concerns within the digital publishing industry. Content creators invest significant resources in producing high-quality, informative content, only to see Google’s AI systems synthesize that information into summaries that eliminate the need for users to visit the original sources.
The Traffic Diversion Reality
Publishers face a challenging paradox: their content fuels AI summaries, but those same summaries prevent users from visiting their websites where they can generate revenue through advertising, subscriptions, or product sales.
SEO Strategy Implications
These findings suggest that traditional SEO approaches focused solely on ranking in search results may need fundamental reassessment. If users increasingly consume information without clicking through to websites, content creators must explore alternative discovery and engagement channels.
Looking Forward: Adapting to the New Search Landscape
While Google continues to assert that AI overviews benefit the broader web ecosystem, the Pew research provides objective evidence to the contrary. The data suggests that AI-powered search features are fundamentally altering how users interact with online information, potentially threatening the economic model that has sustained digital publishing for decades.
Content creators and SEO professionals must acknowledge these realities while developing strategies that account for reduced search referral traffic. This might include diversifying traffic sources, building direct audience relationships, and creating content formats that encourage deeper engagement when users do arrive at their websites.
The conversation between Google and the publishing industry will undoubtedly continue, but now it can proceed with concrete data about user behavior rather than speculation. The challenge lies in finding solutions that serve user needs while preserving the economic viability of the web ecosystem that creates the content users seek.
