Who’s Winning in ChatGPT Referral Traffic?
The traditional search journey has changed forever.
From the rollout of Google’s AI Overviews to new shopping features being integrated into ChatGPT, people are increasingly using and relying on AI agents to fulfil their informational and transactional queries.
For digital marketers, this represents a significant loss of control, as their ability to influence customers through advertising is diminished. A recent eMarketer report estimated that AI agents could cause a 38% decrease in ad exposure during discovery and 47% during consideration. Furthermore, marketers still largely remain in the dark on their website’s AI agent visibility and how they compare against close competitors, relying solely on their AI agent referral sessions from Google Analytics.
At Gener8, our Consumer Browsing data feed provides a comprehensive view of the GenAI intent landscape. For this article we analysed traffic from ChatGPT citations to identify which websites are seeing the largest share of referral traffic, the prevalence of product detail pages (PDPs) in ChatGPT recommendations, and what role ChatGPT plays in shaping the shopping journey.
Which websites recieve the most ChatGPT citation traffic?
We analysed UK ChatGPT referral traffic using data from our nationally representative Consumer Browsing panel of 25,000 users across mobile and web. By tracking visits tagged with the ‘utm_source=chatgpt.com’ parameter, we were able to identify when users navigated from ChatGPT to other websites by clicking on a citation link.
Unsurprisingly, Wikipedia received the largest share of ChatGPT referral traffic among all UK domains. This aligns with expectations, as ChatGPT frequently includes references or citations to support factual statements and detailed responses. These links serve to validate or add credibility to its answers—and Wikipedia is often treated as a go-to truthset for general knowledge and verification.
Taking a bird’s-eye view using IAB-based domain classification, we found that the categories receiving the highest share of ChatGPT referral traffic are News & Weather, Shopping, and Technology & Computing.
Looking at the Shopping category, it’s clear that Amazon dominates - capturing over a quarter (26.4%) of all referral traffic.
As search behaviour evolves, AI search platforms and agents are expected to reduce ad opportunities during the consideration stage of the customer journey. For brands, tracking AI agent referral traffic—both against close competitors and across the broader vertical—will become increasingly vital. New visibility KPIs like "Share of GenAI" are likely to emerge, complementing or even replacing traditional "Share of Search" benchmarks.
Do people use ChatGPT in their shopping journey?
When we looked more closely at ChatGPT’s referral traffic to Amazon, we found that 89% of visits go directly to Product Detail Pages (PDPs). This share has consistently stayed within the 80–90% range since the beginning of the year, but it’s likely to increase significantly in the coming months with ChatGPT having rolled out enhanced shopping results.
Following its May 5, 2025 update, ChatGPT can now showcase relevant product options in visually engaging carousels, offer extra product details, and direct users to websites where they can explore further or complete a purchase (see below).
After the ChatGPT product carousel feature launched, we observed multi-PDP browsing behaviour in our Consumer Browsing data.
For instance, on 7th May we saw that a user, seemingly evaluating their wardrobe options, clicked out from ChatGPT to PDP pages of Amazon, Etsy and Ikea.
As AI agents reduce the need for customers to navigate through websites to find the product they want, marketers must adapt to a ‘zero-click internet’, where products are presented directly within AI interfaces. In this environment, it's crucial for brands to ensure these AI-generated summaries are accurate and favorable—especially since many customers may not investigate further. Optimising product feeds will be essential, especially as ChatGPT plans to introduce a feature allowing merchants to submit their product feeds directly to the platform.
Understand ChatGPT’s impact on the purchase journey
Customer journeys are clearly being redefined by AI agents, yet most brands remain largely unaware of how these changes are affecting them.
As highlighted in our previous Path to Purchase article, Gener8’s interconnected datasets provide comprehensive, unfiltered analysis of user journeys across multiple data sources - a unique capability in the market. Notably, this means that we can also identify customer purchase journeys that included the use of ChatGPT.
For instance, in the smartwatch user journey above we could see:
- The numerous Google searches, article and eRetailer visits, which saw the search query evolve from evaluation intent to a specific product-focused search over time
- The use of ChatGPT to seemingly validate thinking before proceeding to purchase
- The visit to a purchase confirmation page to end the purchase journey
How can I access this data?
We utilised Gener8's Consumer Browsing dataset alongside our UK Nat Rep panel and IAB Domain Classification to reveal insights behind ChatGPT citation traffic.
Gener8 Labs’ complete data and insights solution empowers media and marketing businesses to find actionable consumer and market insights, using our unique, consented, first party panel data sets that are all connected around one user ID.
Discover how you can power your decisions and gain a competitive edge from our behavioural truth set by contacting us today!
