The Universal Commerce Protocol (UCP) is an open standard that lets AI agents discover, evaluate and buy products across online stores with a single, shared integration. Instead of every store building a separate connection to every AI agent, UCP provides one common language for platforms, agents, stores and payment providers — across the whole shopping journey, from discovery to purchase and post-purchase.
Who is behind UCP?
UCP is led by Google and co-developed with a broad coalition: Shopify, Etsy, Wayfair, Target, Walmart, Amazon, Microsoft, Meta, Salesforce and Stripe in shopping, plus Booking.com, Expedia, Hilton and Marriott (lodging) and DoorDash, Uber Eats, Square and Toast (food). More than 70 companies have joined, including payment giants Visa, Mastercard, PayPal and Klarna. When this many competitors rally around one standard, it signals that agentic commerce is becoming infrastructure, not an experiment.
How UCP works technically
UCP is built on familiar, open building blocks so developers avoid new, proprietary technology:
- REST and JSON-RPC — the transport layer agents use to talk to the store.
- Checkout — handles complex cart logic, dynamic pricing and tax calculation.
- Identity linking — secure account connections based on OAuth 2.0.
- Order management — real-time webhooks for status, tracking and returns.
- Interoperability — Agent Payments Protocol (AP2), Agent2Agent (A2A) and the Model Context Protocol (MCP).
UCP vs. ACP: not the same thing
This is the most common confusion. UCP (Universal Commerce Protocol) is the Google coalition's standard, powering Universal Cart and checkout directly in Google AI Mode and the Gemini app. ACP (Agentic Commerce Protocol) is a separate standard from OpenAI and Stripe that powers Instant Checkout in ChatGPT. Both solve the same problem — letting AI buy on your behalf — but they are two different protocols. A serious online store plans for both, not one.
Why UCP matters for AI visibility
When a customer asks an AI agent to "buy the cheapest merino wool winter jacket with free shipping," ranking on Google is not enough. Your product must be machine-readable, carry unique identifiers and be purchasable directly by the agent. UCP makes your products available to agents — but you still have to be chosen. That comes down to the same principles that drive Answer Engine Optimization (AEO): clear entities, structured data and strong trust signals.
How to prepare with CitationLab
CitationLab measures how your brand and products are mentioned and recommended by AI models — before, during and after agentic commerce takes hold. Monitor citations, measure share of voice against competitors, and see which product questions AI actually answers. Check the AI shopping agents readiness checklist to see how ready your store is.
Summary
UCP is the open standard that makes agentic commerce possible across stores and platforms. It is Google-led, broadly backed and built on open web standards — and it is different from OpenAI's ACP. For online stores the message is simple: make your product data machine-readable, plan for both UCP and ACP, and build the AI visibility that decides whether the agent actually recommends you.
Frequently asked questions
What does UCP stand for?
Who is behind UCP?
Is UCP the same as ChatGPT's shopping feature?
What does a store need to be UCP-ready?
Definitions used in this article
- Universal Commerce Protocol (UCP)
- The Universal Commerce Protocol (UCP) is an open, Google-led standard that lets AI agents discover, evaluate and buy products across online stores through a single shared integration, built on REST and JSON-RPC.
- Agentic Commerce Protocol (ACP)
- The Agentic Commerce Protocol (ACP) is a separate standard from OpenAI and Stripe that powers Instant Checkout in ChatGPT, allowing users to complete purchases without leaving the chat.
- Agentic commerce
- Agentic commerce is the practice of letting AI agents discover, evaluate and purchase products on a user's behalf, rather than the user clicking through a store manually.
- Answer Engine Optimization (AEO)
- Answer Engine Optimization (AEO) is the practice of structuring content, entities and trust signals so that AI answer engines select and recommend a brand or product in their generated responses.
- Model Context Protocol (MCP)
- The Model Context Protocol (MCP) is an open standard that lets AI models connect to external tools and data sources through a consistent interface.
