Machine Customer Data — 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 |
Machine customers require structured data across four categories to make autonomous purchasing decisions. Product identity data (SKU, GTIN, brand, manufacturer, category) enables discovery and matching. Specification data (dimensions, weight, materials, certifications in standardised units) enables comparison. Commercial data (real-time pricing, availability, shipping options, return policies) enables transaction evaluation. Trust data (merchant verification, review aggregates, fulfilment reliab
Structuring a product catalog for AI agents requires three layers. The schema layer defines standardised attribute taxonomies using schema.org Product vocabulary extended with category-specific fields. Every attribute must have explicit data types, units, and validation rules. The access layer provides real-time APIs (RESTful endpoints with OpenAPI specifications) that expose catalog data programmatically with sub-second latency. The quality layer implements automated validation pipelines that c
Regular product data is designed for human consumption: marketing descriptions, lifestyle images, persuasive copy, and visual layouts that guide human decision-making. Machine-readable data is designed for agent consumption: structured attributes in standardised formats, explicit data types, quantitative specifications, and programmatic access points. The critical difference is parsability. A human can interpret '2-3 day shipping' but an agent needs structured data: {deliveryMinDays: 2, delivery
Agent authentication requires a delegation-aware approach that verifies both the agent's identity and its authorisation from a human principal. Implement three tiers: anonymous access for catalog discovery (rate-limited, read-only), authenticated access for detailed comparison (agent identity verified, higher rate limits), and authorised access for transactions (both agent identity and human principal delegation verified). Use API keys for agent platform identification, OAuth 2.0 for delegation
Data accuracy for machine customers requires continuous automated quality assurance. Implement schema validation on every data update - checking required fields, valid data types, unit consistency, and cross-field logic. Build consistency monitors that compare structured data against visual product pages - discrepancies trigger trust penalties from agent evaluation systems. Establish freshness SLAs for each data type (pricing: real-time, inventory: 5 minutes, specifications: 24 hours) with autom
Adopt schema.org Product vocabulary as your foundation and extend it with category-specific attributes -