- The Universal Commerce Protocol creates a common language between AI agents, merchants, and payment gateways to manage the entire purchase cycle.
- Google Merchant Center and consistent product data are the foundation for agents to discover, recommend, and sell your catalog.
- UCP transforms traditional SEO into search engine optimization, where feed quality and server-side measurement are key.
- Stores that prepare their technical and data infrastructure for UCP will be better positioned in the boom of conversational and agentic commerce.
The way we shop online is undergoing a revolution with the emergence of AI assistants capable not only of recommending products but also of completing purchases for us. In this new landscape, the Universal Commerce Protocol (UCP) has become the key component enabling frictionless connections between AI agents, online stores, and payment gateways.
Google has launched UCP as an open-source standard already being tested by giants like Shopify, Walmart, Etsy, and Target, and which will soon affect small and medium-sized businesses as well. For any e-commerce business that wants to remain visible and sellable in a world of conversational search, understanding how this protocol works and how to prepare for its integration is no longer optional.
What is the Universal Trade Protocol (UCP) and why is it so important?
The Universal Commerce Protocol is an open standard driven by Google in partnership with major players in commerce and payments, including Shopify, Etsy, Wayfair, Target, Walmart, Visa, Mastercard, Stripe, Adyen, American Express, and Best Buy, among others. Its primary mission is clear: to create a common language for AI-powered digital commerce.
Until now, each online store had its own APIs, checkout flows, and rules , requiring the development of one-to-one integrations between merchants, platforms, and assistants. With UCP, AI agents (like Gemini or Google Search's AI Mode) have a standardized way to: discover what a business can do, query its catalog, create a shopping cart, apply discounts, process payments, and manage after-sales service.
UCP doesn't just focus on payment; it encompasses the entire purchase journey : product discovery, cart creation and updates, pricing and shipping calculations, payment processing, order generation, and subsequent status changes (shipping, returns, etc.). In practice, it's the "orchestration layer" that allows AI to manage the entire process from start to finish.
Another key point is that the merchant remains the Merchant of Record , meaning they maintain the contractual relationship with the customer. The protocol doesn't "keep" the sale; it simply standardizes how the AI agent, the store's backend, and payment providers communicate with each other.
How does UCP fit into the ecosystem of AI protocols and payments?
The Universal Commerce Protocol doesn't come alone; it integrates with other existing standards in the AI and payments ecosystem. Specifically, UCP communicates and is interoperable with:
- MCP (Model Context Protocol), which deals with how AI models access tools and data sources.
- A2A (Agent to Agent)designed so that different AI agents can coordinate with each other.
- AP2 (Agent Payment Protocol), the protocol that enables secure agent-based payments.
The combination of UCP with MCP, A2A and AP2 allows you to cover the entire ecommerce funnel : from the initial conversational inquiry to the purchase confirmation and order tracking, including recommendations, product comparison or the application of loyalty programs.
In the area of payments, UCP doesn't replace existing solutions , but rather works with them. For Google's current implementation, purchases initially rely on standard FPANs (card numbers) stored in Google Wallet, although the design is modular and supports payment processors such as Google Pay, Shop Pay, and other compatible PSPs.
How UCP works at a technical level: from the JSON manifest to the checkout session
Although UCP is flexible and extensible, its technical starting point is simple.Every business that wants to be compliant must publish a JSON file in a well-known path of its domain, usually /.well-known/ucpThis file acts as a manifesto of capabilities.
When an AI agent wants to interact with a specific merchant (for example, create a shopping cart or process a payment), the first thing it does is consult that JSON manifest . There it finds, among other data:
- What services does the store support?: checkout, discounts, order management, returns, etc.
- What extensions does it offer?Multi-item shopping carts, points programs, coupons, special shipping options…
- Which processors or “payment handlers” does it acceptGoogle Pay, Shop Pay, proprietary payment gateway…
- Which REST endpoints should the agent use? to create and update payment sessions or orders.
Once the business profile is discovered, the AI agent initiates the transactional flow . Typically, a checkout session is created via a POST call to the merchant's API (defined in the manifest), and from there, it is updated with PUT or PATCH commands: products are added, promotional codes are applied, taxes and shipping costs are recalculated, etc.
The UCP design is intended to adapt to the reality of each business : there may be country restrictions, specific return policies, stock limitations, or complex pricing rules. Therefore, the protocol defines common primitives but allows each backend to maintain its own business logic.
Shopping experience with UCP: what the user sees and what happens in the background
From the buyer's perspective, UCP makes something very simple to explain, yet very powerful, possible : buying without leaving a conversation with an AI assistant. A typical example might be a request like:
“I’m looking for a lightweight cabin-sized suitcase for under €200, with two-day delivery and good reviews.”
The AI agent consults catalogs, prices and stock from various retailers integrated with UCP , compares options, filters by budget, shipping times and ratings, and displays several alternatives ready to buy directly from the conversational interface (AI Search Mode, Gemini app or website, or other agents that adopt the standard).
If the user decides to proceed, the assistant creates a purchase session with the corresponding store, adds the suitcase to the cart, applies any available discounts, and calculates the final cost including shipping and taxes. If the buyer has their information saved in Google Wallet or another payment service, the payment can be authorized in just a few steps.
In the background, the merchant receives a perfectly structured order , complete with an order ID, product details, shipping address, payment method, and any other information necessary for logistics and after-sales service. Order status tracking is also updated via the API contracts defined by UCP.
Advantages of UCP for retailers, marketplaces, and content creators
For retailers, perhaps the biggest advantage is the reach and reduced friction . Their products can appear natively in Google AI experiences (AI Mode in Search, Gemini web and app, as they are activated) at moments when the user's purchase intent is very clear.
Furthermore, the checkout process is seamless, eliminating the need for the user to navigate through multiple tabs , forms, and intermediate steps. By maintaining a consistent conversational flow, cart abandonment decreases, and the purchase decision is made in seconds rather than minutes.
Another key point is control : the store remains the Merchant of Record, maintains ownership of customer data, and manages after-sales service, returns, and loyalty programs. UCP doesn't turn Google into a commercial intermediary, but rather a technical facilitator.
For publishers and content creators, UCP opens a direct monetization channel for recommendations and discovery content within AI environments, without relying solely on traditional affiliate links. If an article or inspirational content is consumed in a conversational interface, the agent itself can close the sale using the protocol.
The role of Google Merchant Center in the Universal Trade Protocol
Google has made it clear that Merchant Center will remain the control panel where product data is prepared and listings and product pages are managed across its various platforms. Ensuring your data is complete, consistent, and up-to-date is a prerequisite for being "readable" by AI agents.
To prepare for UCP, it is essential to thoroughly review your Merchant Center account : product feeds, required and recommended attributes, brand resources, return policies, shipping settings, and (almost) real-time stock availability.
It is also key to activate the appropriate product attribute for native shopping experiencesIn Google's reference implementation, only listings that use the attribute native_commerce They will display the "Buy" button within the conversational flow, so it's a good idea to review which part of the catalog you want to be compatible.
In many cases, having specialized support in feed management and Merchant Center (from agencies to SaaS solutions) can make the difference between having a visible and competitive catalog for agents or being completely off the radar.
Payment requirements and relationship with Google Pay, Wallet and AP2
In its current phase, Google's implementation for UCP primarily uses standard cards (FPANs) that users store in Google Wallet. This allows a shopper to use payment methods they already have set up, reducing friction.
It is not mandatory for the merchant to have the Google Pay button integrated into its own website to participate in these experiences, but it must have an account in the Google Pay Console and Wallet and meet the integration requirements.
If the merchant's payment gateway isn't yet integrated with the Google Pay API , there are specific workflows to complete the onboarding. In cases where direct integration without going through a PSP is required, the encryption and PCI compliance documentation must be handled.
UCP, combined with the Agent Payments Protocol (AP2), ensures that payments are secure and compatible with existing infrastructure. The merchant declares which methods they accept, the AI agent indicates what it can offer, and the payment flow adapts to the set of options supported by both.
Geographic availability: current situation in the United States and Europe
Google is rolling out UCP gradually and in phases . The first public implementation is focused on eligible merchants in the United States, where users can already shop directly from AI Mode in Search and from Gemini.
For Europe, the message is twofold . On the one hand, UCP has been presented as a global standard, not restricted to a single country; on the other, some advanced uses (such as fully automated in-chat shopping) are being tested first in the US market.
Even so, the basic principles on which UCP is based are already applied in the European market : the need for consistent data, reliable feeds, compliance with payment and privacy regulations, and a technical infrastructure capable of displaying catalogs and orders in a standardized way.
For European e-commerce businesses, the time to prepare is now : fine-tuning data, tracking systems, and integrations with Merchant Center will make it easier to be among the first to take advantage of the protocol when it becomes widespread.
From classic SEO to Answer Engine Optimization (AEO) and agent commerce
The arrival of UCP accelerates a transition that was already underway : from SEO focused on ranking URLs in search results, to optimization designed for response engines and agents (AEO: Answer/Agent Engine Optimization).
In this new model, it's not enough for your page to rank well if AI agents end up closing the sale within the conversation. Your product data, feeds, reputation, and trust signals become the primary fuel these systems use to decide what to offer the user.
For ecommerce SEO, this implies several changes in mindset : content is still important (rich product descriptions, guides, comparisons…), but now you also have to think about conversational prompts, AI-assisted journeys, and experiences where the user doesn't necessarily visit your website even if they buy from your store.
Visibility strategies involve strengthening brand authority , implementing robust structured data, maintaining extreme consistency between what your pages say and what you send in your feeds, and technically enabling agent commerce through UCP.
Data consistency: the basic condition for being visible to AI
With UCP, Google and other players no longer just crawl your pages : they also evaluate the reliability of your catalog and your systems. Minor inconsistencies that might have gone unnoticed before can now cause an AI agent to reject your products.
Key signals that AI takes into account include consistent pricing across all channels, up-to-date stock, on-time delivery, well-applied promotions, and clearly defined variants (sizes, colors, packs, etc.).
Modern ecommerce platforms (Shopify, WooCommerce, Magento…) already centralize much of this information, but that does not automatically guarantee that the data meets UCP expectations or that it is synchronized in real time with all channels.
It is essential to thoroughly audit your data flows : which source is considered the "reference system" (ERP, PIM, CMS), how data travels to Merchant Center, what delays there are in updating stock, and whether promotions are correctly reflected across all surfaces.
Trust solutions (seals, verified reviews, certifications) also play an additional role: they reinforce the security perceived by users and provide an additional layer of “algorithmic trust” that agents can use as a signal, although in no case do they replace data consistency.
Impact on tracking, measuring, and attributing conversions
One of the major challenges that UCP brings is measurement . When a purchase is initiated and completed within Google's AI Mode or Gemini, traditional browser-based scripts may not detect it correctly.
To avoid lost or duplicate conversions, merchants need to confirm orders from the server , capturing order ID, amount, currency, and timestamps independently of the payment flow in the interface.
Attribution also becomes more complex : a sale completed within a conversational experience doesn't fit neatly into classic "source/medium" models. Attribution models and dashboards will need to be adjusted to differentiate traffic and purchases originating from the user-centric approach (UCP) versus those occurring on the company's own website or app.
Practical steps to prepare your ecommerce for UCP
Beyond theory, any online store that wants to be UCP-ready should consider a concrete roadmap, which includes at least the following areas:
- Configure and optimize Google Merchant Center: shipments, returns, attributes, feed quality and activation of products eligible for native payments.
- Request access to UCP programs When they become available in your region, join waiting lists and carefully review the Google Developers documentation.
- Publish a UCP profile on
/.well-known/ucp, where supported services, commercial extensions, supported payment processors, and integration endpoints are described. - Enable checkout and order management APIs that function as true programmatic services and do not depend on "fixes" in the front-end.
- Sync orders and statuses with Google through webhooks or other mechanisms that allow real-time status updates.
In parallel, it is advisable to review the content and technical SEO strategy to adapt it to conversational searches, include complete information on shipments and returns, reinforce structured data and work on brand authority in an environment where agents are increasingly making decisions.
UCP, small businesses and intermediate solutions
This might all sound distant to startups and SMEs that sell online and don't have large technical teams. However, UCP's design is intended to allow platforms (Shopify, marketplaces, SaaS solutions) to offer ready-to-use integrations.
There are already examples of major early partners like Walmart and Sam's Club that allow direct purchases from Gemini with real-time catalogs, prices, and availability, and even loyalty account linking. It's reasonable to expect that, in the medium term, popular CMS and e-commerce solutions will incorporate native support for the protocol.
For smaller businesses, the realistic recommendation is to focus on three pillars : clean and complete product data, solid integration with Merchant Center, and measurement systems prepared to capture orders that do not go through the classic site checkout.
As the ecosystem matures, compatibility with AI agents and UCP will become less of an experiment and more of a requirement for entering the digital market, just as having a decent CMS or accepting card payments are today.
The Universal Commerce Protocol aims to become something akin to the “HTTP” of AI-powered commerce : invisible to the end user but absolutely critical for searches, recommendations, payments, and after-sales service to flow seamlessly between AI agents, merchants, payment gateways, and content platforms. Preparing your data, infrastructure, and strategy now for this new landscape will put you several steps ahead when conversational shopping ceases to be a novelty and becomes the norm.