The Automation Curve in Agentic Commerce

McKinsey's six-level automation curve - from human-driven shopping through agent-assisted, agent-guided, supervised execution, autonomous execution, to multi-agent coordination - provides empirical validation for the AXD Institute's autonomy gradient framework. The projection that AI agents could mediate $3 trillion to $5 trillion of global consumer commerce by 2030 quantifies the economic stakes of agentic experience design.

What is McKinsey's six-level automation curve for agentic commerce?

McKinsey's framework defines six levels of commercial automation, each representing a distinct relationship between human control and agent autonomy. Level 1 is fully human-driven shopping. Level 2 is agent-assisted, where AI provides recommendations but the human makes every decision. Level 3 is agent-guided, where the agent narrows choices and the human confirms.

Level 4 is supervised execution, where the agent transacts within pre-approved parameters and the human monitors. Level 5 is autonomous execution, where the agent acts independently within delegated authority. Level 6 is multi-agent coordination, where multiple agents collaborate across domains without human intervention.

How does this validate the AXD autonomy gradient?

The AXD Institute's autonomy gradient framework - published in September 2024 - argued that the transition from human-controlled to agent-autonomous commerce is not binary but graduated. Each level requires different trust architecture, different delegation design, and different failure recovery mechanisms.

McKinsey's empirical research confirms this graduated model. Their finding that most consumers currently operate at Levels 2-3 while the technology enables Levels 4-5 identifies the trust gap that Agentic Experience Design exists to close. The design challenge is not building more capable agents but calibrating trust at each autonomy level.

What are the three converging forces McKinsey identifies?

McKinsey identifies three forces driving the agentic commerce transition. First, decision-grade AI capability - large language models that can evaluate products, compare options, and execute transactions with sufficient accuracy to be commercially useful.

Second, open protocol infrastructure. The proliferation of protocols - MCP, A2A, AP2, ACP, UCP - provides the technical foundation for agent-to-service communication. Third, the upstream migration of intent - consumers increasingly express what they want as outcomes rather than navigating interfaces to find it.

• Decision-grade AI: LLMs capable of evaluating, comparing, and transacting with commercial accuracy

• Protocol infrastructure: MCP, A2A, AP2, ACP, and UCP provide agent-to-service communication standards

• Intent migration: consumers express outcomes rather than navigating interfaces

What does the $3-5 trillion projection mean for organisations?

McKinsey's projection that AI agents could mediate $3 trillion to $5 trillion of global consumer commerce by 2030 is not a prediction about technology adoption. It is a statement about commercial architecture. Organisations that are not machine-readable - whose catalogues, policies, and value propositions cannot be evaluated by autonomous agents - will be structurally excluded from this market.

The implication is that agent-readiness is not a competitive advantage but a baseline requirement. The AXD Institute's position - that every commercial organisation needs an agentic experience strategy - is now supported by the largest management consultancy in the world.