Similarweb Tracks ChatGPT Ad and Referral Data

Referral traffic from ChatGPT to brand websites has surged 150% following a May 7 update that increased the prominence of links within AI-generated answers [2]. This shift in user behavior highlights the growing influence of AI platforms on web traffic, as brands recommended by ChatGPT are now 2.5 times more likely to receive a site visit within one week compared to those that are not [3].

Tracking the AI advertising shift

The rise in traffic coincides with OpenAI’s expansion of its advertising ecosystem. Since launching sponsored placements for US Free and Go users on February 9, 2026, more than 600 advertisers have begun bidding on high-intent queries, such as software comparisons and service recommendations [1]. These ads appear as sponsored cards below the AI’s response, featuring a headline, favicon, and a link to a destination page [1].

Because OpenAI does not provide a centralized ad library, businesses are increasingly using manual monitoring and third-party tools to track competitor activity [1]. Effective competitive intelligence requires capturing four specific data points: ad title, description, destination URL, and impression share [1]. Analysts suggest that tracking impression share—the percentage of total ad impressions a competitor captures on a specific prompt—is the most accurate way to determine which brands are dominating a category [1].

Impact on user engagement

Data indicates that traffic originating from AI recommendations often leads to deeper user engagement. Users who discover a brand through a ChatGPT-influenced search view an average of 12 pages and remain on the site for nearly 12 minutes, significantly higher than the 6.5 pages and 5.6 minutes observed for users arriving through other channels [3].

However, the volatility of AI recommendations remains a challenge for marketers. Because AI tools can return different brand suggestions across repeated versions of the same query, relying solely on AI visibility is considered a high-risk strategy [3]. Approximately 56% of traffic to recommended brands is driven by users performing a follow-up branded search, suggesting that traditional search engine optimization remains a critical component of an AI-driven marketing strategy [3].

What to watch

• Ad inventory expansion: Monitor whether OpenAI introduces a searchable ad library, which would reduce the current reliance on manual prompt-based tracking [1].
• Recommendation stability: Watch for changes in how AI models prioritize brand suggestions, as frequent shifts in results can disrupt long-term traffic patterns [3].
• Conversion metrics: Track whether the current 150% increase in referral traffic translates into sustained customer acquisition or if it remains a temporary byproduct of the new interface [2].

The integration of sponsored placements and prominent organic links has turned ChatGPT into a significant, albeit unpredictable, driver of web traffic. As brands compete for visibility in AI answers, the ability to map buyer intent and monitor competitor bidding will likely become a standard requirement for digital marketing teams.