AI agents are beginning to browse, compare and buy on a shopper's behalf. Google's Universal Commerce Protocol and OpenAI's Agentic Commerce Protocol are competing standards for how that transaction happens, and both read structured product data — which puts your Merchant Center feed at the centre of a channel that did not exist two years ago.
Two standards are competing to define how an AI agent completes a purchase. They differ in who owns the checkout, what data they demand, and how much of the customer relationship you keep.
Google's protocol for agent-driven purchasing, and what it asks of your Shopping ads and product feed
Google's and OpenAI's protocols side by side: checkout ownership, data requirements and merchant control
How UCP expanded into cart handling, catalog access and identity linking, and what each one exposes
Agents cannot infer what your feed does not state. The attributes that were merely nice to have for Shopping ads — precise categories, complete identifiers, genuine descriptions — become the difference between being considered and being skipped.
The AI surfaces are already carrying commercial formats, and revenue arriving through them does not attribute like a normal session. Both sides are worth understanding before the volume matters.
Google's Shopping ad format inside AI Mode, and how placement differs from conventional results
A newer AI Mode ad format built around a specific offer, and what advertisers need to prepare
How Shopify is exposing storefronts to AI chat interfaces, and what selling through them involves
Attribution when an agent does the browsing and buying, and the sessions your analytics never sees
Agentic commerce is a new model where AI agents browse, compare, and purchase products on behalf of consumers. Instead of a human clicking through search results, an AI assistant handles the entire shopping journey—from product discovery to checkout.
AI agents evaluate your product feed data directly rather than relying on ad placements. Feed quality, structured data completeness, and pricing accuracy become even more critical. New ad formats like Direct Offers let brands participate in AI-mediated shopping experiences.
Google's UCP builds on existing Merchant Center infrastructure. OpenAI's ACP is an open standard requiring new API endpoints. Most Google Shopping advertisers should prioritize UCP readiness first, then evaluate ACP for broader reach.
Ensure every product has accurate titles, descriptions, prices, images, GTINs, availability, and product categories. Add conversational attributes that help AI agents answer shopper questions. Keep your feed updated in real-time. Use the UCP Product Feed Readiness Checklist for a complete audit.