Why Product Data Is Becoming a Marketing Channel

AI shopping is changing the role of product data. A catalog was once mainly an operational record that fed a retailer’s website, search filters and advertising platforms. It is now becoming the language through which external AI systems decide whether a product is relevant, trustworthy and available—and whether a shopper can buy it without visiting the merchant’s storefront.

The change is visible in Shopify’s October 6 analysis of the 2026 holiday season. Shopify says 65% of shoppers in its Global Holiday Retail Report expect to use AI for at least one shopping task. The company also reports rapid growth in AI-referred orders for several merchants. Those figures are Shopify’s own research and customer examples, but they point to a practical shift: marketing teams increasingly need to communicate with machines as well as people.

Discovery is moving beyond the storefront

Traditional digital marketing assumed that a brand could control most of the discovery journey through search results, advertisements, social posts and its own website. Conversational shopping breaks that sequence. A customer may describe a problem in natural language, receive a short list from an agent and move directly to checkout. The merchant’s homepage, navigation and campaign landing page may never appear.

Adobe’s June 17 Brand Visibility announcement frames this as a new competition for attention across AI interfaces. Adobe says AI traffic to U.S. retail sites grew 1,324% between October 2024 and May 2026. That is a company-reported measure, not an industry census, but it helps explain why marketing platforms are adding metrics such as AI mentions, citation share and prompt-level visibility.

The implication is broader than generative-engine optimization. Being mentioned is useful; being accurately understood is better. An agent needs structured evidence about product attributes, compatibility, materials, availability, delivery, returns and the customer problem a product solves. Weak or contradictory data can remove an otherwise suitable item from consideration before a person sees it.

The catalog now carries brand strategy

Shopify describes its Catalog and Agentic Storefronts as tools that structure product information and distribute it to AI channels. Its Agentic Storefronts announcement says merchants can define product schemas, attributes, policies, frequently asked questions and brand voice while keeping prices and inventory current. These are product claims from Shopify, yet the operating model is important: catalog management, content strategy and channel distribution are converging.

A product title and price are no longer enough. Marketing must work with commerce, merchandising and operations to produce machine-readable claims that can survive comparison. If “sustainable,” “professional grade” or “designed for sensitive skin” matters to the sale, the supporting details need to be specific, consistent and available in authoritative sources. The agent can only present the case the brand has made.

Checkout standards turn visibility into distribution

New protocols are connecting this discovery layer to transactions. Salesforce and Google Cloud said on September 15 that their Universal Commerce Protocol integration would reach general availability in October 2026. Their model lets merchants surface catalog data and enable checkout inside Google Search, AI Mode and Gemini while retaining payments and order management in Commerce Cloud.

Stripe’s earlier Agentic Commerce Protocol announcement described a similar separation of responsibilities: an AI interface can carry purchase context and a scoped payment token, while the merchant accepts the order and continues to handle tax, fulfillment and returns. The strategic point is not which protocol wins. Commerce is becoming available through multiple conversational surfaces, and a merchant cannot treat each one as an isolated campaign.

Marketing performance now depends on operational accuracy

When an agent can recommend and transact, stale inventory or unclear policies become marketing failures. A compelling product description cannot compensate for an unavailable variant, inconsistent shipping promise or missing return rule. Brand visibility therefore depends on the quality of systems that marketers do not usually own.

Leaders should create joint ownership across marketing, merchandising, commerce technology and operations. Start with the products most likely to be researched through natural-language questions. Audit whether their attributes, evidence, policies and availability are consistent across the website, feeds and authoritative third-party sources. Then track not only AI-referred traffic, but also recommendation accuracy, product-feed freshness, conversion, returns and the share of sessions that move from an AI surface to a completed order.

Product data is becoming a marketing channel because it increasingly determines whether a brand enters the AI-mediated buying conversation at all. The businesses best positioned for agentic commerce will not simply publish more content. They will make their product truth structured, current and convincing wherever a customer delegates discovery to a machine.

Header image: Original AI-generated editorial illustration created for WiredBusiness. It represents structured product data flowing through conversational shopping, secure checkout and fulfillment, and does not depict a specific vendor product or interface.