The AI checkout standard that debuted at NRF in January is now default infrastructure. Walmart was just the headline act.

Agentic checkout stopped being a demo this year. At the National Retail Federation show in January 2026, Google unveiled the Universal Commerce Protocol (UCP), an open standard that lets AI agents discover products, build a cart, and complete a purchase inside a conversation. It did not launch it alone: UCP was built with Shopify, Etsy, Wayfair, Target, and Walmart, and endorsed by more than twenty other partners. Walmart was the flagship retail example, pairing its and Sam’s Club’s assortment with Google’s Gemini so a shopper can get a recommendation and pay with Google Pay or PayPal without leaving the chat. Six months on, the standard has spread from a launch announcement to the default wiring of online retail, which is the part operators need to act on.

The protocol matters more than any single retailer’s deal, so it is worth being precise about it. UCP is a shared language that lets an AI agent talk to a store’s catalog and checkout. Before it, every “buy inside an assistant” integration was a one-off. With an open protocol, any agent that speaks UCP can read any compliant catalog, assemble a cart, and trigger a checkout the merchant still controls. That is why the January announcement was an architecture story, not a Walmart story: it set the standard the rest of the year would build on.

And build on it they did. The clearest proof that this crossed from announcement to infrastructure is Shopify. Shopify co-authored UCP and has since turned it on by default across its stores, extending agent-accessible commerce to its entire merchant base rather than a hand-picked pilot. When the platform that powers a large share of independent commerce flips a standard on by default, the question for operators stops being “will this channel exist” and becomes “is my catalog ready for it.”

Here is the shift in plain terms. The old funnel was a sequence of pages: a shopper searches, clicks to a product page, browses, adds to cart, enters payment, completes checkout, and your analytics stack tags every step. The agentic path compresses that into one exchange. The shopper asks an assistant for the best air purifier under $200 for a small room, the agent recommends a product, the shopper says buy it, and the payment clears inside the conversation. The conversion path did not get shorter. It moved off your website.

That relocation is where the real operator decisions live.

Catalog structure is now a distribution decision, not a merchandising one. An AI agent does not look at your product photography or read your brand story. It parses structured data: titles, attributes, pricing, inventory status, specifications. The listing that ranks well for a human skimming search results and the listing an agent confidently recommends are not the same listing. Ambiguous titles, missing attributes, and inconsistent pricing are the noise that makes an agent pick a competitor instead. For anyone selling on Walmart Marketplace or running a Shopify store with UCP on, that catalog is potentially in the agent flow already, and its data quality decides whether it shows up well.

Attribution breaks in ways most teams have not planned for. If a purchase completes inside an assistant using Google Pay, that sale may never appear in your store analytics, your Google Analytics property, or your paid-media attribution. The funnel happened off-site. For teams running meaningful ad spend, that is a measurement gap that widens as agentic volume grows, and the fix is still being worked out across the industry. The move now is to map where the gap will hit before it distorts your reporting, not after.

The urgency is real but not a fire drill. Consumer adoption of AI shopping agents is still early, and transaction volume through these flows is not yet material for most merchants, though adjacent agentic channels are already reporting outsized conversion. But the standard is set, the payment rails work, the largest US retailer is a launch partner, and the dominant independent-commerce platform has it on by default. The operators who spend the next few months making their product data agent-readable are positioning for a channel that could turn material faster than mobile commerce did.

What to watch: whether Walmart and Google publish real transaction volume (still unreleased), how fast other retailers and platforms adopt UCP now that Shopify has defaulted it on, and whether a clean attribution standard emerges for off-site agentic sales.

Key Facts

• Standard: Google’s Universal Commerce Protocol (UCP), an open standard for AI agents to discover, cart, and check out inside a conversation.
• Announced: NRF 2026, January 11-12, 2026. Co-built with Shopify, Etsy, Wayfair, Target, and Walmart; endorsed by 20+ more partners.
• Walmart’s role: Walmart and Sam’s Club assortment accessible inside Google Gemini via UCP, with checkout through Google Pay or PayPal.
• Shopify: co-authored UCP and enabled it by default across its stores on June 17, 2026, extending agent-accessible commerce to its full merchant base.
• Operator takeaway: structured product data (clean titles, attributes, pricing, inventory) now determines whether agents recommend your products.
• Open issue: off-site agentic purchases can bypass standard analytics and attribution.

Frequently Asked Questions

When did this actually launch?

Google announced UCP and the Walmart-Gemini integration at the NRF 2026 conference on January 11-12, 2026. It has rolled out across the industry through the first half of 2026, including Shopify enabling UCP by default across its stores.

Does this affect me if I sell on Walmart Marketplace?

Potentially yes. Walmart’s integration covers its third-party marketplace, so your products may be discoverable and purchasable through Gemini conversations. Audit your listings for clean, structured data: unambiguous titles, complete attributes, accurate pricing, and inventory status, since those are the signals agents use to decide what to recommend.

Can Shopify merchants use UCP today?

Yes. Shopify co-authored UCP and has enabled it by default across its stores, so agent-accessible commerce is available to its merchant base rather than requiring a custom connector. Merchants should confirm their setup and, most importantly, make sure their catalog data is structured for agent readability.

How does agentic checkout affect my analytics and attribution?

If a shopper discovers and buys your product entirely inside an assistant using Google Pay or PayPal, that transaction may not show up in your store analytics, Google Analytics, or paid-media attribution, because the purchase happens off-site. Map this gap now and watch for attribution solutions as volume grows.

How much volume is flowing through agentic checkout right now?

Google and its retail partners have not released transaction volume. Adoption of AI shopping agents is still early, so the practical move is to get catalog data agent-ready and monitor the channel rather than reallocate significant budget to it today.

Sources

• Walmart, “Walmart and Google Turn AI Discovery Into Effortless Shopping Experiences” (Jan 11, 2026)
• CNBC, “Walmart teams up with Google’s Gemini” (Jan 11, 2026)
• Google, Sundar Pichai’s NRF 2026 remarks
• Retail Dive, “Walmart teams up with Google’s Gemini for AI-assisted shopping”
• Shopify, “Building the Universal Commerce Protocol”

Related

• Shopify enables Universal Commerce Protocol by default for 4M merchants — the platform side of the same standard
• Agentforce Commerce GA: AI Traffic Converts 8x Social — another agentic-commerce channel going live