McKinsey Projects $5 Trillion Global Agentic Commerce Market as AI Agents Become Retail Gatekeepers

Explore how AI agents are redefining retail. McKinsey predicts $1T in US revenue by 2030 as agentic commerce replaces traditional e-commerce models.

By: AXL Media

Published: Apr 10, 2026, 4:03 PM EDT

Source: Information for this report was sourced from McKinsey & Company

The Seismic Shift from E-Commerce to Autonomous Agentic Flows

Agentic commerce represents a rethinking of shopping where the boundaries between platforms and services dissolve into an integrated, intent-driven flow. Unlike the previous e-commerce revolution, which took decades to mature, the agentic era is expected to move significantly faster because agents can "ride on the rails" of existing digital infrastructure. By 2030, the US B2C market alone could see $1 trillion in revenue orchestrated by these autonomous proxies. This transition moves the consumer journey away from manual clicks and scrolls toward a world where AI anticipates needs, negotiates deals, and manages complex logistics, such as relocating a family across the country, with minimal human intervention.

Emerging Technical Infrastructure and Interoperability Protocols

The infrastructure enabling this shift is being built through several critical protocols designed for machine-to-machine communication. The Model Context Protocol (MCP) allows agents to share context and intent across different tools, while the Agent-to-Agent (A2A) Protocol enables autonomous negotiation between a consumer’s agent and a retailer’s system. Furthermore, Google’s Agent Payments Protocol (AP2) and the Agentic Commerce Protocol (ACP), co-developed by OpenAI and Stripe, are creating secure, cryptographically signed mandates for verifiable purchases. These tools allow AI models like Claude 4.5 to handle tasks with "time horizons" exceeding 30 hours, moving from simple queries to multi-day coordination of services and goods.

Three Primary Interaction Models Defining the Marketplace

McKinsey identifies three key models for agentic interaction: agent-to-site, agent-to-agent, and brokered agent-to-site. In the first, an agent scans traditional merchant platforms directly; in the second, agents transact autonomously with other agents to negotiate bundles or discounts. The brokered model involves intermediary systems, such as OpenTable or Expedia, facilitating multi-platform interactions. This "de-verticalization" of shopping means that consumers may no longer visit specific destinations like Amazon or travel sites, but instead initiate their intent through a personal concierge embedded in everyday tools like Slack or ChatGPT.