Agentic Experience Design (AXD) is the discipline for designing trust-governed relationships between humans and autonomous AI systems. Founded in September 2024 by Tony Wood in Manchester, United Kingdom, AXD addresses how humans delegate, calibrate, observe, interrupt, and recover trust in agentic AI.
| 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 |
Designing for machine customers requires six core shifts: machine readability over visual appeal (structured data, APIs, schema.org), programmatic trust over brand trust (verifiable credentials, performance history), outcome guarantees over feature lists, negotiation interfaces over fixed pricing, interoperability over lock-in, and continuous accountability over one-time conversion. The AXD Institute's Machine Customer Design guide provides the complete methodology.
Human-first commerce presents products through visual interfaces, brand storytelling, and emotional appeals optimised for human attention and conversion. Machine-first commerce presents the same products through structured APIs, machine-readable metadata, and programmatic trust signals optimised for agent discoverability and transaction efficiency. The most effective strategy is a dual-layer architecture serving both through the same underlying systems.
AI-to-AI commerce is the model where both the buyer and the seller are autonomous agents. A buyer agent, delegated purchasing authority by a human principal, negotiates with a seller agent delegated pricing and fulfilment authority by a merchant. This requires trust establishment between agents, machine-speed negotiation, programmatic arbitration, and observability for both human principals.
Focus on four priorities aligned with the Four Pillars of AXD Readiness: (1) Data Architecture - implement schema.org markup, create documented APIs, ensure structured product metadata. (2) Trust Infrastructure - build verifiable credentials, performance history APIs, compliance certifications. (3) Transaction Capability - enable API-first commerce, dynamic pricing, standardised contracts. (4) Organisational Readiness - train teams to think in terms of machine and human customers simultaneously.
Banking and financial services, retail and e-commerce, and B2B enterprise procurement are the most immediately affected. In banking, agents negotiate loans and manage portfolios under regulatory constraints. In retail, agents compare thousands of merchants simultaneously. In B2B, agents evaluate suppliers and negotiate contracts autonomously. Each industry requires specific machine customer design patterns addressed in the AXD Institute's industry analysis.
Gartner predicted that by 2028, 15 billion connected products will have the potential to behave as customers. But the design implications extend far beyond the technology. When the customer is no longer a human navigating an interface, every assumption of traditional customer experience design is challenged. Brand loyalty becomes irrelevant. Visual design becomes invisible. Emotional persuasion becomes noise. Machine customers operate with zero brand loyalty, infinite patience for optimisation, and no susceptibility to emotional marketing. They evaluate structured data, verify trust signals, and optimise for the outcome their human principal specified. The businesses that win in Designing for machine customers requires inverting many of the principles that govern human-centred design. The following six principles form the foundation of machine customer design, drawn from the AXD Institute's research across Machine customers cannot see your website. They parse structured data, APIs, and machine-readable formats. Every product, service, and offer must be described in structured, parseable formats - schema.org markup, JSON-LD, well-documented APIs, and standardised product taxonomies. The Machine customers do not trust brands. They verify trust signals - verifiable credentials, audit trails, compliance certifications, and performance history. Machine customers optimise for outcomes, not features. They need to know what result a product or service will deliver, under what conditions, with what guarantees, and what happens when things go wrong. Outcome specification replaces feature marketing. Machine customers can negotiate at scale. Businesses that offer programmatic negotiation interfaces - dynamic pricing APIs, bundle configuration endpoints, and conditional offer structures - will capture agent traffic that bypasses fixed-price competitors. Machine customers operate across multiple services simultaneously. They favour businesses that support open standards, standard