Source-led article

ChatGPT Access Linked to 9% Drop in Traditional Search Queries Among US Desktop Users

AI News India//3 min read
A graph showing a decline in traditional search engine use corresponding with increased ChatGPT access.
A graph showing a decline in traditional search engine use corresponding with increased ChatGPT access.
Featured image from the source article

New research analyzing US desktop browsing data has revealed a significant shift in online search behavior, with broadened access to OpenAI’s ChatGPT linked to a 9% reduction in traditional search queries. This decline was particularly pronounced for informational content, raising questions about the future of content creation and digital advertising.

The study, conducted by researchers at Bocconi University using Comscore clickstream data, compared how often ChatGPT and Google referred users to external websites. While Google continues to dominate referral traffic at 31.1% of sessions, ChatGPT now accounts for 5.2%. However, the expansion of ChatGPT access to paid subscribers (October 2024), free users (December 2024), and anonymous users (February 2025) directly correlated with a measurable drop in traditional search.

Key facts

Metric Google Referrals ChatGPT Referrals
Session Referral Rate 1% 2%
Monthly Referral Rate N/A 5% to 6.5%
Traditional Search Drop N/A 4% (after 20 weeks, 17%)

Shifting Referral Patterns

The research highlights a distinct difference in the types of websites users visit when referred by ChatGPT compared to Google. ChatGPT tends to direct users towards non-profit organizations, subscription-based services, freemium platforms, and specialized sites such as reference and knowledge platforms, tools, SaaS platforms, and academic or developer sites. This contrasts with Google’s referral traffic, which often leads to popular destinations like YouTube, Reddit, and Wikipedia.

Ad-supported sites saw a notable decrease of 27.6 percentage points in referral traffic from ChatGPT. This suggests a potential long-term impact on publishers heavily reliant on advertising revenue generated through traditional search engine referrals.

Impact on Search Query Types

The study found that the decline in traditional search queries was not uniform across all content types. Informational searches were the most affected, with academic research referrals dropping by 32.8% and general references by 26.5%. Conversely, transactional and recreational searches showed minimal change, indicating that users might still prefer traditional search for specific purchasing decisions or entertainment.

The researchers cautiously interpret their findings, emphasizing that their analysis measures changes in traffic allocation rather than consumer surplus, publisher revenue, or long-term content production. However, the observed trends suggest a significant evolution in how users seek and consume information online.

Implications for Content Creators and Digital Marketers

For Indian content creators, digital marketers, and businesses, these findings underscore the growing influence of AI chatbots on user behavior. The shift away from ad-supported sites and towards specialized content platforms for AI-driven referrals means that strategies focused solely on traditional SEO for ad-revenue generation may need re-evaluation.

Understanding how users interact with AI summaries and referred links becomes crucial. As previous data from the Pew Research Center indicated, users often click links less frequently when an AI summary is present. This new research expands on that, showing that even when users do click through from ChatGPT, they are directed to a different ecosystem of websites. Content creators may need to adapt their strategies to cater to these new referral patterns, potentially focusing on high-value, specialized content that aligns with the types of sites ChatGPT tends to promote.

The study also prompts a critical question for the industry: will content creators continue to produce at the same scale if AI search leverages their work but generates fewer direct visits? This complex issue requires further investigation, particularly with mobile, international, and revenue data to fully understand the economic impact on the content ecosystem.

Source: Search Engine Journal (https://www.searchenginejournal.com/chatgpt-access-tied-to-9-drop-in-traditional-search/582167/)