Product Pages for Agents — an AXD Institute resource on agentic experience design, agentic commerce, trust architecture, and human agent interaction. Founded by Tony Wood..
| Dimension | Traditional UX | Agentic Experience Design (AXD) |
|---|---|---|
| Primary material | Attention and affordance | Trust and delegation |
| User state | Present, navigating | Absent, delegating |
| Design output | Screens and interfaces | Outcomes and constraints |
| Temporal model | Session-based | Relationship-based |
| Success metric | Task completion | Trust calibration |
Redesigning product pages for AI agents requires shifting from human persuasion to machine parseability. Implement comprehensive schema.org Product JSON-LD with all attributes (SKU, GTIN, specifications, pricing, availability). Separate factual attributes from marketing copy - agents parse structured fields but may ignore prose. Publish standardised comparison attributes in quantitative formats. Embed verifiable trust signals (verified reviews, return policies, fulfilment metrics) as structured
AI agents require structured data across five categories on product pages: identity data (name, SKU, GTIN, brand, manufacturer), specification data (dimensions, weight, materials, certifications in standardised units), commercial data (price, currency, availability, shipping options, return policy), trust data (aggregateRating, reviewCount, verified seller status, fulfilment reliability metrics), and variant data (every colour, size, and configuration as a distinct structured entity). All data s
AI agents evaluate trust through verifiable signals rather than persuasive design. They assess competence trust (can this merchant fulfil correctly?) through fulfilment reliability metrics and return rates. They evaluate integrity trust (is the product description accurate?) by cross-referencing structured attributes against review sentiment. They check benevolence trust (does the merchant act in the buyer's interest?) through return policy generosity and dispute resolution fairness. Predictabil
Yes - but the purpose changes. Visual product pages will continue to serve human shoppers who browse directly, and they remain important for brand building and high-consideration purchases. However, you must now design two layers: the visual layer for human visitors and the structured data layer for agent visitors. These layers must be consistent - if your visual page claims '4.8 star rating' but your structured data shows 4.2, agents will flag the discrepancy as a trust failure. Think of struct
Traditional ecommerce APIs serve internal systems (inventory management, order processing, analytics). Agent-accessible product APIs serve external AI agents making purchasing decisions. The key differences: agent APIs must expose comparison-critical attributes in standardised formats, not just internal data structures. They must include trust signals (reliability metrics, verification status) that internal APIs never needed. They must support agent authentication and mandate verification - conf
Implement comprehensive schema.org Product JSON-LD on every product page - include name, SKU, GTIN, brand, description, offers (price, currency, availability), aggregateRating, and review data as required by Validate your structured data against Google's Rich Results Test and schema.org validators weekly - broken markup is invisible to humans but makes your products invisible to Design your trust signals to be verifiable rather than merely claimed -