Artificial intelligence (AI) traffic to U.S. retail websites now converts at higher rates than non-AI traffic such as paid search and email marketing, Adobe announced Thursday (April 16).
Vivek Pandya, director of Adobe Digital Insights, said in a blog post Thursday that AI traffic converted 42% better than non-AI traffic in March, marking a “significant reversal” from a year ago, when AI traffic was 38% worse converting.
“Increasing consumer trust is a contributing factor, with Adobe research finding that 66% of respondents believe that AI tools provide accurate results,” Pandya said. “This gives shoppers confidence and encourages more transactional activity.”
According to the post, a conversion is a measure of a visit to a website that results in a purchase.
Two related metrics are also higher for AI traffic. Adobe found in March that people who accessed a retail site from an AI source spent 48% more time on the website and viewed 13% more pages than those who accessed it from a non-AI source.
A December 2025 PYMNTS Intelligence report, “How AI Can Help Consumers Get Started with Everything,” found that purpose-built AI environments are beginning to replace traditional discovery.
Advertisement: SCROLL TO CONTINUE
“Getting attention increasingly depends on whether the truth of a brand’s offer, policy, or product can be interpreted and recommended within a conversational environment,” the report says.
Adobe also announced in a blog post Thursday that the amount of traffic from AI sources to its U.S. retail sites grew 269% year over year in March and 393% in the three months from January to March.
The company found that 39% of consumers say they have used AI for online shopping, and 85% of them say the technology has improved their shopping experience.
“These numbers highlight that AI provides lasting value in the e-commerce space, reducing the time it takes consumers to find what they need or find relevant discounts,” Pandya said in the post.
In the post, Adobe also announced new data showing that 66% of individual product pages on retail websites can be read by large-scale language models (LLMs). This means that the remaining 34% of sites are not optimized for LLM.
“Retailers have thousands of SKUs, and our data shows that much of the content is currently invisible to LLMs,” Pandya said in the post.
