AI-driven referrals are no longer an experiment: visitors sent from AI platforms convert 60% better and generate 53% more revenue per visit than other traffic, according to Adobe Analytics. At the same time, Adobe finds 39% of U.S. retail homepages remain difficult for large language models (LLMs) to read — a direct revenue leak for brands that haven’t made their product information machine-readable.

What Adobe measured

Adobe’s analysis, drawn from more than 1 trillion visits to U.S. retail sites, reports that AI-originated sessions converted at a rate 60% higher than non-AI sessions and delivered 53% more revenue per visit. AI referrals also show stronger engagement: 14% higher engagement scores, 59% more time on site, 33% lower bounce rates and 28% more cart additions compared with other visitors.

On volume, Adobe says AI-driven referrals to U.S. retail sites rose 62% year over year in July 2026 and are up 1,219% since October 2024, when consumer-facing generative AI platforms became broadly available. Adobe’s customer footprint includes more than 200 of the Top 2000 online retailers using its web analytics in 2025.

Where retailers lose visibility

To explain why AI referrals outperform, Adobe introduced an “AI Content Visibility Checker,” a diagnostic that evaluates whether pages can be parsed by LLMs. After expanding its cohort in July, Adobe reports an overall LLM visibility rate of 61% for retail homepages — meaning 39% of homepages are not easily machine-readable.

Category-level machine-readability scores reported by Adobe are: Apparel 76%, Electronics 70%, Cosmetics 68%, Sporting Goods 67%, Furniture & Home 64%, General Merchandise 63% and Grocery 59%. Adobe attributes higher scores in apparel, electronics and cosmetics to more consistent, structured product content.

Why this matters to marketing and commerce teams

These findings tie an analytics signal to immediate commercial outcomes: when AI tools can find and interpret product details — price, availability, amenities — shoppers arrive better informed and convert at higher rates. Conversely, content that LLMs can’t parse effectively means lost visibility to a growing referral channel and, by Adobe’s metrics, fewer conversions and less revenue.

That does not mean Adobe’s numbers map exactly to every merchant. The data reflects Adobe’s analytics footprint and an expanded but unspecified cohort; teams should validate the trend against their own logs and attribution before shifting budget. Still, the direction is clear: machine-readability matters in both discovery and conversion.

Practical steps for teams

Marketing, merchandising and engineering teams can start with three priorities:

  • Audit LLM readability: Run a diagnostic across landing and product pages to identify where product names, pricing, availability and attributes can’t be reliably parsed by models.
  • Structure content for machines: Use consistent, product-level descriptions, clear HTML markup and Schema.org structured data so LLMs and agents can extract the fields shoppers care about.
  • Track AI referrals: Tag and instrument traffic identified as originating from AI tools in your analytics stack to measure conversion behaviour, average order value and lifetime impact.

Prioritize categories with the lowest visibility — grocery, general merchandise and furniture/home — while using higher-scoring categories such as apparel and electronics as templates for consistent product-level content.

What to watch next

Expect the retrieval layer — how AI tools find and rank product content — to matter more as agentic shopping features and platform policies evolve. Vendors and platforms will change how they parse, surface and attribute content; that will affect which pages get traffic and which retailers capture AI-driven demand.

For now, the immediate next step is tactical: treat machine-readability as a measurable conversion lever. Audit, structure and instrument your product information, then compare Adobe’s signal with your own analytics to decide where to invest. Retailers that do this will be best positioned to capture higher-value AI referrals as the channel grows.