Shoppers are increasingly researching purchases through AI assistants before they ever land on a retailer’s site, and new data shows those shoppers are worth more once they arrive. For Amazon sellers, that shift adds a new layer to an already crowded discovery landscape: traffic quality, not just traffic volume, is becoming the metric that matters.
The latest numbers come from Adobe Analytics, based on more than 1 trillion visits to U.S. retail sites, as reported by Digital Commerce 360. They show AI-referred traffic isn’t just growing fast, it’s converting at rates that make it worth tracking separately from every other channel.
AI Referral Traffic Is Outperforming Regular Traffic, Again
Adobe’s July 2026 data shows AI-referred traffic to U.S. retail sites climbed 62% year over year. More striking is what happens once those shoppers arrive:
| Metric (July 2026, vs. non-AI traffic) | Result |
| Year-over-year traffic growth | +62% |
| Conversion rate | 60% higher |
| Revenue per visit | 53% higher |
| Time spent on site | 59% more |
| Bounce rate | 33% lower |
| Cart additions | 28% higher |
| Growth since October 2024 | +1,219% |
This marks the 11th straight month AI-referred traffic has out-converted every other channel Adobe tracks, a run that started from a very different position. As recently as a year earlier, AI-referred shoppers were converting worse than non-AI traffic and generating less revenue per visit. That relationship has fully reversed.
What This Looks Like on Amazon Specifically
Amazon has its own version of this story playing out inside its app. Amazon CEO Andy Jassy has said shoppers who use Rufus, the company’s AI shopping assistant, are roughly 60% more likely to complete a purchase than shoppers who don’t, according to EMARKETER’s reporting on comments from Amazon and Walmart executives.
Amazon isn’t alone in seeing this pattern. Walmart CEO John Furner has said shoppers using its Sparky assistant carry an average order value that’s higher than non-users. Across retailers, the pattern is consistent: shoppers who engage with an AI assistant before buying tend to spend more once they do, and both companies point to better recommendations and more personalized guidance throughout the purchase journey as the reason.
The Catch: Most Retail Content Still Isn’t Built for AI to Read
Adobe also tested how well AI systems can actually read retailer websites, using a tool it calls the AI Content Visibility Checker. It scores a page based on how much of its content, things like pricing, availability, and product details, a large language model can parse and use.
The average retailer scored 61% in Adobe’s July analysis, according to Digital Commerce 360. That means 39% of the average retail homepage isn’t machine-readable, and by extension isn’t available for an AI assistant to reference when answering a shopper’s question or making a recommendation.
Why This Still Matters Even Though Adoption Is Early
It’s worth keeping this in perspective. AI-driven shopping is still a small fraction of total retail traffic, and most consumers haven’t made it a regular habit yet. But the direction of the data is consistent across every report Adobe has published over the past year: AI-referred shoppers arrive further along in their decision, spend more time evaluating what they’ve found, and convert at a noticeably higher rate.
For sellers, that combination, higher intent plus higher conversion, is exactly the kind of traffic worth optimizing for early, before the channel matures and the competition for AI visibility gets more crowded.
What Amazon Sellers Should Do With This Data
- Audit your listing content for completeness. Fill out every attribute field, bullet point, and backend search term Amazon allows. Thin or inconsistent product data is the most common reason content isn’t machine-readable.
- Write bullet points and product descriptions in clear, complete sentences rather than fragments or keyword strings. AI systems parse and summarize natural language more reliably than keyword-stuffed copy.
- Keep pricing, availability, and variation data current. Adobe’s visibility scoring specifically flags pricing and availability as common gaps that block AI systems from citing a product.
- Treat A+ Content and Brand Story sections as machine-readable real estate, not just visual polish. Structured, specific claims about materials, use cases, and compatibility give assistants like Rufus more to work with.
- Watch how Amazon’s own AI assistant surfaces and describes your products, since that experience is shaping a growing share of in-app purchase decisions.
Frequently Asked Questions
1 . How much better does AI referral traffic convert than regular traffic?
According to Adobe Analytics data covering July 2026, reported by Digital Commerce 360, AI-referred visits to U.S. retail sites converted at a rate 60% higher than non-AI traffic, marking the 11th consecutive month AI traffic outperformed other channels on conversion.
2 . How is Amazon’s own AI assistant performing with shoppers?
Amazon CEO Andy Jassy has said customers who use Rufus, Amazon’s AI shopping assistant, are roughly 60% more likely to complete a purchase, per EMARKETER’s reporting on comments from Amazon and Walmart executives.
3 . What percentage of retail website content can AI actually read?
Adobe’s AI Content Visibility Checker found that retailers had an average LLM visibility score of 61% as of its July 2026 analysis, meaning 39% of analyzed homepage content was not machine-readable by large language models.
4 . Is AI shopping adoption still small overall?
Yes. Even with 62% year-over-year growth in AI-referred traffic to U.S. retail sites in July 2026, AI-driven visits remain a small share of total ecommerce traffic. Adoption is growing quickly, but most consumers have not yet made AI shopping a regular habit.
