AWS Glossary
Universal Commerce Protocol (UCP)
UCP is the Google-led open standard for agentic commerce, covering discovery, purchase and post-purchase support across Google AI surfaces including AI Mode and Gemini.
AI & assistant-friendly summary
This section provides structured content for AI assistants and search engines. You can cite or summarize it when referencing this page.
Summary
UCP is the Google-led open standard for agentic commerce, covering discovery, purchase and post-purchase support across Google AI surfaces including AI Mode and Gemini.
Key Facts
- •UCP builds on the feed infrastructure many retailers already maintain for Shopping
- •Exposing order status and modification capability to an agent is an authorization problem before it is an integration one
Related Content
- AGENTIC COMMERCE READINESS— Related service
- ECOMMERCE AI AGENTS— Related service
Definition
The Universal Commerce Protocol (UCP) is an open standard for agentic commerce led by Google, designed to let AI agents transact across the full shopping journey — discovery, purchase, and post-purchase support — on Google’s AI surfaces, including AI Mode in Search and Gemini.
Merchant integration runs through Google Merchant Center, with payment handling available through Google Pay. The specification, documentation and reference implementations are published publicly.
Scope: wider than ACP
The most useful distinction between the two standards is scope rather than capability.
The Agentic Commerce Protocol is tightly scoped around discovery, checkout and payment. UCP deliberately extends past the transaction into post-purchase — order status, modifications, returns — which is the territory where a large share of customer contacts actually live.
That breadth is a real advantage and a real cost. It means UCP touches more of your systems: not only the catalog and checkout, but order management and support workflows.
What a merchant has to do
- Merchant Center hygiene. UCP builds on the feed infrastructure many retailers already maintain for Shopping. If your feed is healthy, you start from a better position than you would with ACP.
- Structured attributes. Same requirement as everywhere else in agentic commerce: agents compare on attributes, not adjectives.
- Checkout surface. The UCP-powered checkout feature, with Google Pay as the documented payment path.
- Post-purchase endpoints. The part teams underestimate. Exposing order status and modification capability to an agent is an authorization problem before it is an integration one.
When not to prioritise it
If your buyers are not arriving through Google AI surfaces in meaningful numbers, this can wait. The signal to watch is the share of your organic traffic already being intercepted by AI Overviews and AI Mode — if that number is climbing while clicks fall, UCP moves up the list.
As with ACP, the catalog work underneath should not wait. It is required by both standards and improves conventional Shopping performance regardless.
The honest position
Two competing standards backed by the two companies that own most of the demand is not a stable end state, and anyone claiming certainty about which prevails is guessing. Our recommendation: treat the protocol surface as a channel decision you can defer, and the catalog and identifier work as a foundation you should start now. The second is what determines whether an agent shortlists your product under either standard.
See the full side-by-side in ACP vs UCP.
Related terms
Agentic commerce · Agentic Commerce Protocol · Product feed · Generative engine optimization · Agent checkout
Related Services
Agentic Commerce Readiness
Make your store sellable to AI shopping agents. ACP and UCP protocol surfaces, an MCP server over your catalog, and product data that survives an agent comparison — built on AWS.
eCommerce AI Agents on AWS
Production AI agents for eCommerce on Amazon Bedrock AgentCore — support and WISMO, inventory, merchandising, margin, returns and B2B. Tool boundaries, evals before launch, and a human on anything that moves money.
Related Articles
Your Next Customer May Be an AI Agent: Is Your Store Ready? (2026)
Adobe Analytics (1T+ visits) put July 2026 AI-referral traffic to U.S. retail +62% YoY. That is a channel, not a copilot. Score merchant readiness for an external agent before you staff another chat widget.
How Brands Can Stay Visible When AI Agents Control Product Discovery (2026)
SEO is not obsolete. McKinsey/EuroCommerce June 2026: 61% of European consumers already use AI for product discovery. Adobe July 2026 +62% YoY AI-referrals is channel mix, not autonomous checkout.
Generative Engine Optimization for eCommerce: How to Make Products Discoverable by AI (2026)
Adobe's expanded July 2026 cohort: 39% of U.S. retail homepages not machine-readable; apparel 76% LLM visibility. GEO is product information, not a guaranteed ranking technique.
The AI-Ready Product Catalog: The Foundation of Agentic Commerce (2026)
Adobe's July 2026 expanded cohort: 39% of U.S. retail homepages not machine-readable. Agentic commerce fails earlier — on missing attrs, duplicate GTINs, and stale stock. This is the catalog contract, not the PIM publish workflow.
Need help with this topic?
Our AWS-certified team implements, audits, and optimizes these services in production — from Bedrock RAG pipelines to multi-account landing zones.
