How much traffic do AI assistants actually send to retailers?

Adobe's recap of the 2025 holiday season, drawn from "over 1 trillion visits to U.S. retail sites" plus a companion survey of "more than 1,000 U.S. respondents," reports that traffic to retail sites referred by generative AI tools "increased by 693.4% compared to the year prior." Vivek Pandya writes for Adobe Digital Insights that the surge was not confined to retail: "travel traffic increased 539%, financial services 266%, tech and software 120%, and media and entertainment 92%," with retail showing "the strongest year-over-year growth in generative AI traffic among all industries" — AI-driven traffic to retail sites rose 769% in November and 673% in December. Adobe reports that this AI-referred traffic did not merely arrive in greater volume; it converted better too, with AI referrals "converting 31% more than other traffic sources — nearly doubling YoY," and AI-driven revenue per visit (RPV) "up 254% this holiday season year-to-date."

Adobe attributes the stronger performance to relevance and intent-matching rather than volume alone. Shoppers arriving from generative AI assistants were "33% less likely to leave immediately (a 14% improvement since the beginning of 2025)," spent "45% more time on-site," and viewed "13% more pages per visit" — evidence, Adobe argues, of "stronger content relevance and user intent" in what an AI assistant hands off to a retailer. Adobe's survey layers trust onto this behavioral data: 47% of consumers reported trusting AI, 64% of AI shoppers reported satisfaction with the links they received, more than 55% said they actively click on them, and 65% reported more purchase confidence after AI's help, with 68% less likely to return what they bought — a pattern Adobe connects to holiday returns running down 1.2% YoY.

This cuts against the clickthrough story on the publisher side of the same shift. How much are AI Overviews actually reducing organic search clicks? and Will AI Overviews reduce search referral traffic to publishers? both describe AI summaries suppressing the traffic that used to flow to publisher and search-result pages. Adobe's holiday data describes the other side of that same funnel: traffic an AI assistant does send onward, specifically to retailers, arrives converting disproportionately well. Read together, the two bodies of data suggest AI intermediaries are not simply shrinking the open web's traffic — they are also concentrating and pre-qualifying what traffic does pass through. On trust, Adobe's transaction-linked satisfaction and purchase-confidence figures corroborate survey findings like How much do shoppers trust AI shopping assistants? and Does naming a payments brand increase trust in AI transactions?, but with actual conversion and return-rate outcomes rather than stated intent.

The excerpt does not say how Adobe's analytics attribute a given visit to "generative AI" referral — whether by referrer header, UTM tagging, or some other method — nor does it break the aggregate 693.4% figure down by individual AI tool; ChatGPT, Gemini, and Perplexity are named only as a general trend. The 1,000-respondent survey results are self-reported trust and satisfaction, not measured behavior, and the whole report is Adobe's own vendor analysis of a single holiday season rather than an independent or peer-reviewed study. It establishes that AI-referred traffic converted better and grew faster during this one season; it does not establish that generative AI referral causes higher purchase intent rather than simply reflecting shoppers who were already further along in deciding what to buy before they asked an AI assistant.