Agentic AI Projected to Account for 15–25% of US E-Commerce by 2030

Bain and Morgan Stanley: $385 Billion by 2030

Two of the most respected names in business forecasting have converged on a striking projection: agentic AI could account for 15–25% of total US e-commerce by 2030. Bain & Company's forecast puts the share at up to 25%, while Morgan Stanley estimates that AI agents will influence approximately $385 billion in US ecommerce transactions by the end of the decade. Together, these projections signal that AI-mediated commerce is not a niche trend—it is a structural shift comparable in scale to the transition from brick-and-mortar to online retail.

What Agentic Commerce Actually Means

Agentic commerce refers to a model where AI agents—autonomous software systems acting on behalf of consumers—participate in some or all stages of the purchasing process. This ranges from product discovery (an AI assistant recommending products based on a conversation) to full transaction completion (an AI agent researching, comparing, negotiating, and purchasing on behalf of a user).

The "agentic" distinction is important. Unlike current recommendation engines that suggest products and let the human decide, agentic AI can make decisions autonomously within parameters set by the user. A consumer might authorize their AI agent to "buy the best-rated wireless earbuds under $100 with same-day delivery" and trust the agent to execute the entire transaction without further input.

The Protocol Wars: UCP and ACP

Two competing—but potentially complementary—protocols are emerging to standardize how AI agents interact with merchants:

Universal Commerce Protocol (UCP)

Developed by Google in collaboration with Shopify, UCP aims to create a universal standard for AI agent-to-merchant communication. It defines how agents discover products, access pricing and inventory data, manage carts, and complete checkouts. UCP has attracted support from Amazon, Microsoft, Meta, Walmart, Visa, and Mastercard, suggesting it could become the dominant standard.

Agentic Commerce Protocol (ACP)

OpenAI has developed ACP as its standard for AI-to-merchant interactions within the ChatGPT ecosystem. While UCP aims for universal applicability across platforms, ACP is designed specifically for OpenAI's agent infrastructure. The two protocols may eventually converge or coexist, much as different payment standards coexist in the current ecommerce landscape.

What Must Change for Merchants

The shift to agentic commerce requires merchants to rethink several foundational elements of their business:

Loyalty Programs Must Become Agent-Readable

Today's loyalty programs are designed for human comprehension: points balances, tier levels, reward catalogs. In an agentic commerce model, AI agents need to evaluate loyalty benefits programmatically. This means loyalty data must be expressed in structured, machine-readable formats that agents can compare across retailers. A loyalty program that only exists as a mobile app interface will be invisible to an AI agent choosing between competing retailers.

Product Data Must Be Structured for AI Consumption

The quality and structure of product data will become the single most important factor in AI-driven commerce. Products need comprehensive, accurate, and well-structured data that AI agents can parse and evaluate. This goes far beyond basic product titles and descriptions—it includes specifications, compatibility data, use-case information, comparison attributes, and policy information (returns, shipping, warranties).

Pricing and Promotions Must Be Machine-Accessible

AI agents need to access real-time pricing, active promotions, and discount eligibility without navigating a website. Merchants will need to expose this data through APIs or structured feeds that agents can query programmatically.

Expert Perspectives

Industry leaders have emphasized that the transition to agentic commerce is already underway. The CPO of Talon.One, a promotions platform, noted that merchants need to begin restructuring their promotional data now to be ready for AI agent consumption. Promotions that require human interpretation—complex coupon codes, conditional discounts buried in marketing copy—will be difficult for agents to evaluate, putting those merchants at a disadvantage.

The CEO of Kantar's commerce practice described the shift as the most significant structural change in retail since the advent of ecommerce itself. Retailers that adapt their data infrastructure, loyalty programs, and product presentation for AI consumption will capture disproportionate share as agentic commerce scales.

Action Plan for Shopify Merchants

1. Audit your product data quality. Ensure every product has comprehensive specifications, high-quality images, structured attributes, and complete policy information.
2. Implement structured data markup. Use JSON-LD and schema.org markup to make product data machine-readable.
3. Evaluate your loyalty program's machine readability. Can an AI agent determine a customer's loyalty tier, available rewards, and applicable discounts programmatically?
4. Monitor UCP and ACP developments. These protocols will define the rules of engagement for agentic commerce. Early awareness allows early preparation.
5. Test your AI discoverability. Ask ChatGPT, Gemini, and Perplexity to recommend products in your category. Do your products appear? If not, your data needs improvement.