From the search bar to the shopping agent: the discovery shift in 5 charts

Product discovery is moving from the search-results page to the AI assistant faster than most roadmaps assume. Here is the shift in five charts, each one a real, sourced data point your strategy should be planning around.

Chart 1: AI shopping is already mainstream

AI shopping is no longer an early-adopter behavior. 61% of consumers say they have used generative-AI tools like ChatGPT for online shopping, and 69% have used AI for online shopping in some form (Capital One Shopping, 2026). The assistant is already in the buying journey.

Chart 2: The traffic surge is real, and it converts

Adobe measured a 693% year-over-year jump in AI-driven traffic to U.S. retail sites over the 2025 holiday, and a further 393% in Q1 2026 (Adobe, April 2026). That traffic converted 31% more than other sources, and revenue per visit from AI referrals grew 254% year over year (Digital Commerce 360, January 2026). The AI shopper is not just more numerous; the visits are more valuable.

Chart 3: Adoption is broad across generations

Adoption is not confined to the youngest shoppers. 46% of Millennials and 44% of Gen Z have used AI to shop, alongside 26% of Boomers (Capital One Shopping, 2026). The behavior already spans the generations with the most spending power.

Chart 4: Agents are becoming infrastructure

This is not only about advice; it is about action. OpenAI and Stripe co-developed the Agentic Commerce Protocol and open-sourced it in 2025, and in January 2026 Google, Shopify, and partners launched the Universal Commerce Protocol, standards that let agents read catalogs and complete purchases (Fast Company, 2026). The trajectory is steep: Gartner projects agents will drive 20% of digital-commerce transactions by 2030, and commercetools cites U.S. agentic-commerce projections reaching the hundreds of billions of dollars by 2030 (commercetools, 2026). The shopper is being joined by an agent acting on their behalf.

Chart 5: Readiness is uneven, and accuracy is the gate

Demand is here, but supply-side readiness is not uniform, and shoppers verify what they are told (Yext, 2026). A CatalogSignal CEI Benchmark of 100 brands across 10 verticals and four AI providers (more than 100,000 AI shopping queries) found 47.7% mean funnel accuracy, 9.3% mean hallucination, 10.6% mean commercial harm, and a provider-level accuracy range from 12.5% to 53.2%. (Directional panel metrics, not shopper-level rates; no individual brands reported.) The traffic is arriving faster than catalogs are getting ready for it.

The takeaway for CDOs

Five charts, one conclusion: discovery is being intermediated by machines, the high-intent traffic is already material, and the differentiator is whether your catalog is legible and trustworthy to those machines. The brands that treat AI catalog readiness as a measured, ongoing program, not a one-time SEO chore, are the ones that will be on the shortlist when the agent decides.

Where does your catalog sit on chart 5? A baseline Commerce Eligibility Index™ assessment scores your AI readiness with product-level evidence. Request one at catalogsignal.com.

Sources