Skip to main content
Agentic commerce is online buying where an AI agent performs the discovery, comparison, and sometimes the checkout on a shopper’s behalf, rather than the shopper browsing a storefront directly. The shopper states an intent — “find me a waterproof hiking boot under £150 in a wide fit” — and the agent queries catalogues, compares options, and returns a shortlist or completes the order.

​
Why it changes ecommerce SEO

In a browser-led purchase, a store competes on the storefront: imagery, layout, reviews, persuasion. In an agent-led purchase, the agent never sees the storefront as a human does. It consumes structured, machine-readable product data. The consequences:
  • Structured product data becomes the product page. Price, availability, GTIN, variants, shipping, and returns must be present as machine-readable markup, not only as rendered text or images.
  • Accuracy outranks presentation. An agent filtering on “in stock, wide fit, under £150” will exclude a product whose data does not say so, regardless of how good the page looks.
  • Specification completeness wins comparisons. Missing attributes cannot be compared, so incomplete products drop out of consideration sets rather than losing on merit.
  • Persuasive copy has little effect. The agent is matching constraints, not being convinced.

​
What it asks of a store

  • Valid Product structured data on every product page, including offers, price, priceCurrency, and availability
  • Distinct, individually addressable variants rather than a single page with options hidden behind scripts
  • Product attributes stated as data — material, dimensions, fit, compatibility — not only in prose or images
  • Feeds and markup kept in sync with real inventory, since stale availability data causes silent exclusion
  • Body content readable without JavaScript execution

​
Current state

Agentic commerce is early. Standards are still forming across payment protocols, agent identity, and checkout handoff, and support differs by platform and by agent. The durable preparation is the same either way: complete, accurate, machine-readable product data — which also serves conventional search and shopping surfaces.