The Universal Commerce Protocol: Google's Plan to Let AI Agents Buy Things for You

Google, Shopify, and more than 20 major partners are building the infrastructure for a future where AI agents handle the entire shopping journey. The implications for brand visibility are enormous.

There is a protocol quietly reshaping the future of online commerce, and most marketers haven't heard of it yet.

In January, Google launched the Universal Commerce Protocol (UCP) at the National Retail Federation conference. Co-developed with Shopify and endorsed by more than 20 partners including Walmart, Target, Etsy, Wayfair, Mastercard, Visa, Stripe, American Express, Best Buy, The Home Depot, and Macy's, UCP is an open-source standard that creates a shared language between AI agents and commerce systems.

In plain terms: UCP is the plumbing that allows an AI agent to find a product, check inventory, apply a discount, process payment, and complete a purchase on your behalf, all without you ever visiting a retailer's website.

This is not a concept. It is already live. US shoppers can now buy items from Etsy and Wayfair directly within Google's AI Mode in Search and the Gemini app. Shopify, Target, and Walmart integrations are coming soon.

How UCP actually works

The traditional online shopping experience involves a lot of friction: searching across multiple retailers, navigating different product pages, comparing prices, managing separate checkout processes. Each retailer has its own technical setup, its own way of structuring product data, its own payment flow.

UCP eliminates this fragmentation by standardizing how commerce data is organized and exchanged. Products, prices, carts, checkout flows, payment processing, and order status all follow a shared format. When a retailer adopts UCP, any AI agent that speaks the protocol can interact with that retailer's inventory and checkout system without custom integration work.

The protocol is built in layers. At its core, it defines basic transaction primitives like checkout sessions, line items, and order status. On top of that, modular capabilities handle specific functions like product discovery, cart management, and fulfillment. Extensions add specialized features like discount codes, loyalty credentials, and subscription billing. Businesses implement only the capabilities they need, and agents negotiate only what they can handle.

Shopify, which co-developed the protocol, describes the design philosophy this way: "Monolithic protocols eventually collapse under complexity. Thoughtfully layered protocols survive and thrive by separating responsibilities, defining clear APIs, and enabling composition."

Critically, the retailer remains the merchant of record throughout the entire transaction. They keep their customer data, their pricing logic, and their fulfillment process. UCP doesn't disintermediate the retailer from the transaction. It disintermediates the retailer's website from the transaction.

That distinction matters enormously.

What this looks like in practice

Here is the experience Google is building: A shopper opens AI Mode in Search or the Gemini app and types something like "I need a carry-on suitcase for a two-week trip to Europe." The AI doesn't just return a list of links. It researches options, compares features, checks real-time inventory and pricing across participating retailers, and presents a curated recommendation.

If the shopper likes what they see, they tap a buy button right there in the AI interface. Google Pay handles the payment using credentials already saved in Google Wallet. The retailer processes the order. The shopper never leaves the conversation.

For returning customers, the protocol allows retailers to personalize the experience: loyalty pricing, member discounts, product recommendations based on purchase history. For new customers, retailers can offer enrollment incentives at the point of sale.

Google CEO Sundar Pichai, speaking at NRF, framed UCP as a way to keep the customer relationship intact while removing friction from the buying process. The retailer still owns the relationship. The AI just handles the logistics.

The brand visibility problem

This is where things get interesting for marketers.

In a UCP-powered world, the AI agent becomes the primary decision-making layer between the consumer and the purchase. The consumer expresses a need. The agent researches, compares, and recommends. The consumer confirms. The transaction completes.

At no point in this flow does the consumer browse a retailer's website, scroll through a category page, see a banner ad, or encounter any of the traditional touchpoints where brands compete for attention. The entire discovery and evaluation process happens inside the AI.

This means that brand visibility is no longer about where you rank on a search results page or how well your product page converts. It's about whether the AI agent surfaces your product at all.

Andy Reid, chief innovation officer and AI advocate, raised this concern publicly after the UCP announcement: "If UCP allows Gemini to collapse the entire shopping journey into a single 'Pay' button, does this accelerate the move toward a 'Default Economy' where only one brand is surfaced as the definitive answer? How do smaller brands survive if the protocol favors the single most coherent 'default' rather than a marketplace of choices?"

It's a fair question. When a human browses a category page, they might see 20 or 30 options. When an AI agent makes a recommendation, it might surface two or three. The competitive surface area shrinks dramatically.

What influences the AI's recommendations

So how does an AI agent decide which products to recommend? This is the question every brand should be asking, and the answers are still taking shape.

Based on what we know about how AI models operate and what Google has signaled through its product design, several factors appear to matter:

Structured data quality. UCP relies on standardized product information. Google has announced dozens of new data attributes in Merchant Center specifically designed for conversational commerce, going beyond traditional keywords to include things like answers to common product questions, compatible accessories, and substitutes. Brands with richer, more complete structured data will be more discoverable to AI agents.

Information consistency. AI models assign confidence based on corroboration. Christian Ward, EVP and Chief Data Officer at Yext, put it directly: "AI will think, 'I found this information in 26 places, so I'm pretty confident it's true.' Versus, 'I found it in four places, so I'm not so confident.'" Consistent brand information across your website, third-party listings, reviews, and structured data sources increases the likelihood that an AI agent will recommend you.

Reviews and reputation signals. AI models weigh sentiment and credibility. Strong, consistent reviews across multiple platforms are a signal that a product or brand is trustworthy and relevant.

Brand salience across the web. The more frequently and consistently your brand appears in relevant contexts across the web, the more likely AI models are to surface it in recommendations. This is the digital version of word-of-mouth: if enough credible sources mention your brand in connection with a topic, the AI treats that as a signal of authority.

The bigger picture

UCP is not just a Google initiative. It's compatible with existing industry protocols including Agent2Agent (A2A), the Agent Payments Protocol (AP2), and Model Context Protocol (MCP). It's open source and available on GitHub. Google is explicitly positioning it as an industry standard, not a proprietary platform.

But the implications of industry-wide adoption are significant. As eMarketer's Jeremy Goldman noted: "Universal, agent-mediated commerce reduces the importance of owned storefronts at the moment of purchase. When transactions occur inside Google's environment, brands lose control over UX, merchandising, and post-purchase engagement. Products risk becoming interchangeable inputs, ranked by availability, price, and fulfillment reliability rather than narrative or identity."

That's the tension at the heart of agentic commerce. The protocol is designed to benefit everyone: consumers get less friction, retailers get more reach, AI platforms get to monetize the experience. But the shift in power is real. When the AI agent is the one making (or heavily influencing) the purchase decision, the brands that thrive will be the ones that understand how to be visible and credible in that context.

We are still in the early innings. UCP checkout is currently limited to a handful of US retailers on Google surfaces. Multi-item carts, loyalty program integration, and post-purchase workflows are still on the roadmap. Global expansion to markets like India, Indonesia, and Latin America is planned but not yet live.

But the architecture is being built now. And the brands that start optimizing for AI agent visibility today will have a significant advantage over those that wait until the infrastructure is fully deployed.

The shopping journey is being compressed into a conversation. The question is whether your brand is part of it.

James Calder is the editor of The Search Signal, covering AI-powered search, generative engine optimization, and the future of brand discovery.