It might not be long before you start to see AI in your favorite stores

It might not be long before you start to see AI in your favorite stores

AI is already moving real retail revenue, not just generating buzz. Adobe’s data shows AI traffic to U.S. retail sites jumped 393% year over year in Q1 2026, and AI-referred visitors in March converted 42% better than non-AI traffic.

What is happening

The shift started with shoppers using generative AI for research, recommendations, and deal-finding before they ever reached a store site. Adobe first reported that traffic from generative AI sources to U.S. retail websites rose 1,200% in early 2025, with those visitors showing higher engagement and lower bounce rates than non-AI traffic.

By April 2026, that behavior had strengthened: AI traffic to retailers was up 393% in Q1, and Adobe said the gap in conversion had flipped from AI underperforming to AI outperforming by a meaningful margin. That matters because it suggests AI is no longer just helping people browse; it is helping them buy.

Why retailers care

AI shopping assistants reduce friction. They help customers narrow choices faster, compare products more naturally, and get to a decision with fewer clicks than traditional search and filters.

That translates into better business metrics: more engaged visits, higher conversion, and often higher basket value through smarter recommendations and upsells. Amazon’s Rufus is a clear example, with Amazon saying users of the assistant are 60% more likely to complete a purchase.

Online to in-store

Retailers are now pushing AI from the website into the physical store. The next step is using AI as an in-store assistant for associates and shoppers, helping with stock checks, product recommendations, and guided selling on the sales floor.

That matters because the same AI behavior that boosts online conversions can also improve in-store service, especially when shoppers want fast answers about availability, fit, compatibility, or alternatives. In practice, AI becomes a sales layer across both channels instead of a single chatbot on a website.

Back to blog