Stock Markets June 15, 2026 11:06 AM

AI-driven referrals lengthen sessions and lift spend on U.S. retail sites, Adobe data shows

Customers sent to merchant websites by large language models spend more time, visit more pages and generate higher revenue per visit

By Marcus Reed
Share
Twitter Reddit Facebook LinkedIn

May data from Adobe Analytics indicates U.S. shoppers who arrive at retailers' websites after consulting large language models, including Google’s Gemini and OpenAI’s ChatGPT, spend more time on sites, browse more pages and produce higher revenue per visit than shoppers from non-AI sources. The findings highlight rising AI-driven traffic and a call for brands to make pages readable for LLMs to capture the benefit.

AI-driven referrals lengthen sessions and lift spend on U.S. retail sites, Adobe data shows
Summarize with
ChatGPT Perplexity Claude Grok Gemini

Key Points

  • AI-referred visitors delivered 53% more revenue per visit than shoppers from non-AI sources in May, according to Adobe Analytics.
  • In May, AI-referred visitors converted at a rate 54% higher and spent 53% more time on retailer websites than non-AI referred visitors, while also visiting more pages.
  • AI-driven traffic to retail websites rose 138% year-on-year in May and reached the highest share of total retail visits since Adobe began tracking the metric in October 2024 - impacting retail and e-commerce sectors as well as digital marketing strategies.

U.S. online shoppers referred to retailer sites by large language models are staying longer, viewing more pages and generating larger purchases per visit than consumers coming from non-AI channels, according to May figures from Adobe Analytics.

The data show a marked lift for traffic originating from LLM recommendations - examples named in the report include Google’s Gemini and OpenAI’s ChatGPT - with AI-referred visitors producing 53% more revenue per visit than those arriving from non-AI sources.

Adobe's findings also point to stronger engagement and conversion for those visitors. In May, retail website visitors who were guided to merchant pages by AI converted at a rate 54% higher than online shoppers from non-AI sources, and they spent 53% more time on the sites. The firm reported that AI-referred shoppers also navigated to more retail webpages on average than their non-AI referred counterparts.

The share of retail site visits coming from AI sources surged year-on-year: Adobe Analytics reported a 138% increase in AI-driven traffic to retail websites in May compared with the same month a year earlier. Adobe noted that this represented the highest share of total retail visits from AI since the firm began tracking the metric in October 2024.

Adobe's analysis underscores a practical implication for brands: pages need to be structured so that LLMs can read and recommend product information effectively. Vivek Pandya, director of digital insights at Adobe, said retailers whose products appear in LLM suggestions are better positioned to "drive more personalization" for shoppers who leave AI platforms to complete purchases on the retailers' native websites.


These results suggest that AI recommendation channels are currently delivering both higher engagement and stronger monetization per visit for retailers that capture those referrals. Adobe's month-to-month and year-over-year metrics in May provide a snapshot of how LLMs are influencing shopper behavior online and the potential value of optimizing ecommerce content for AI visibility.

Risks

  • Brands that do not make webpages readable for large language models risk missing higher-converting AI referral traffic - a concern for e-commerce and digital marketing teams.
  • Adobe's AI traffic tracking began in October 2024, so the dataset cited reflects a limited tracking history; this constrains long-term trend visibility for retailers and market analysts.

More from Stock Markets

AMD Acquires MEXT to Add AI-Driven Memory Optimization to Data Center Suite Jun 15, 2026 India's Central Bank Outlaws Deceptive Digital Sales Practices by Lenders Jun 15, 2026 Meta Integrates Muse Spark-Powered AI Into Facebook for Search and Creative Tools Jun 15, 2026 Waste Management of Canada lines up up to C$750M note sale ahead of September maturity Jun 15, 2026 Belgian equities retreat as BEL 20 closes down 0.93% Jun 15, 2026