For two years the agentic-commerce conversation has lived mostly in conference panels and venture decks. Adobe's Q1 2026 Digital Insights data drops a hard number into the middle of it: AI-driven traffic to U.S. retail sites grew 393% year-over-year in the first quarter, Adobe reported this week, and that traffic isn't just window-shopping.

It converts. According to Adobe, AI-sourced traffic converts to purchase at a rate 42% higher than other channels. Visitors arriving from AI sources spend 48% longer on retail sites and view 13% more pages per session. For retailers conditioned to think of AI shopping as a long-tail "future state," that data is the thumb on the scale that says: the future state has started.

What Counts as "AI Traffic"

The Adobe taxonomy is worth understanding. AI traffic in this measurement framework includes referrals from ChatGPT, Perplexity, Gemini, Claude, and a growing set of consumer-facing AI assistants — but it also captures shopping-specific agents like the ones Walmart and Target have publicly partnered on, as well as the agentic checkout layers that Shopify, Stripe, and Visa have rolled out over the past 18 months.

That's a meaningful definitional shift. As recently as Q4 2025, AI referral traffic was a rounding error in most retailer analytics tools — the few visitors that did arrive were tagged as "direct" or fell through into untracked source buckets. Adobe's data suggests that's no longer a viable measurement default.

The 73% That Already Use AI to Shop

If 393% sounds like a top-of-funnel curiosity, a separate study cited by Crescendo is the demand-side gut-check: 73% of consumers report using AI somewhere in their shopping journey. The breakdown — 45% use AI assistants to generate product ideas, 37% to summarize reviews, 32% to compare prices, and 70% express at least some comfort with an AI agent making a purchase on their behalf — closes the gap between "AI is interesting" and "AI is a primary research tool."

Pair that with the NielsenIQ report released Monday, which projects U.S. retail media ad spend will hit $107.6 billion in 2026, and a strategic question emerges that most retailers haven't fully answered: how do you optimize for an audience whose decisions are increasingly mediated by an LLM you don't control?

The Visibility Problem

Therein lies the practical issue. Endcap has spoken with five retail tech leaders over the past three weeks, and the consistent refrain is that measurement of AI traffic is roughly where mobile traffic measurement was in 2011. Most analytics implementations don't reliably attribute AI-sourced sessions. Even when they do, the upstream data — what the assistant said about the brand, what it surfaced, what it omitted — is opaque.

Bain has written about this gap under the framing of "AI agent surface area" — the new battleground for product visibility. Where SEO won the last 20 years on backlink graphs and metadata, agentic commerce will likely be won on structured data, real-time inventory feeds, and the willingness to license catalog access to AI providers on terms that retailers can stomach.

That's the subtext to last week's Endcap coverage of Google's Merchant Center AI product import announcement and the various Shopify, Walmart, and Etsy moves toward agentic catalog exports. Retailers aren't just "trying AI" — they're actively shaping how their products get represented inside someone else's reasoning model.

What This Means for Retailers Right Now

A few things are reasonable to conclude from Adobe's numbers:

The first is that AI traffic is no longer a directional indicator — it's a measurable channel that's growing at hyperscale and converting at premium rates. Retailers that can't separate AI sessions from other sources in their analytics are flying blind.

The second is that the conversion premium suggests AI shoppers arrive with higher intent. They've already done the comparison and review reading; by the time they hit the retailer's site, the decision tree is collapsed. That's the opposite of how social-driven traffic typically behaves.

The third — and this is the part Retail Dive's earlier coverage of agentic risk flagged correctly — is that the same dynamics that make AI a great acquisition channel also concentrate gatekeeping power. Five or six AI providers will likely mediate a meaningful share of retail discovery within 36 months. That's a market-share concentration retailers should be modeling now, not later.

What to Watch

Adobe's Digital Insights team typically updates its retail traffic series quarterly. The Q2 print, due in early August, will be the first sustained read on whether the 393% growth rate is the new normal or a step-change that levels off. Either way, the strategic question — how retailers earn and measure visibility inside AI-mediated discovery — is now central to anyone's 2026 retail tech roadmap.

Sources: