Agentic Checkout Is Coming for Your Funnel. Get Your Product Data Ready Before It Does.

Google released the Universal Commerce Protocol in January 2026 and has been expanding it since. UCP is an open standard that lets a retailer put a buy button on eligible products inside Google’s AI Mode and Gemini, so a shopper can complete a purchase in the assistant without ever visiting the product page. Checkout happens on Google’s surface while you stay the merchant of record. Google built it with Shopify, Etsy, Wayfair, Target, and Walmart, and more than 20 other partners have endorsed it, including Visa, Mastercard, Stripe, American Express, Best Buy, The Home Depot, and Macy’s. Named rollout brands include Nike, Sephora, Target, Walmart, Wayfair, and Shopify merchants like Fenty and Steve Madden, starting in the US.

This is not a Google side project. It is the same shift Amazon is making with Rufus and Walmart is making by folding Gemini into its shopping experience. The direction is identical across all three: an AI agent sits between the shopper and your listing, reads the available product data, and decides what to recommend and, increasingly, what to buy. The product detail page you have spent years optimizing is no longer guaranteed to be part of the path.

The recommendation

Get your product feed agent-ready now, and treat it as the highest-leverage work you are not currently doing. That means complete and accurate attributes, real-time price and availability, a clear returns policy, and structured review data on every SKU across every channel. An agent cannot recommend what it cannot parse, and it will not buy what it cannot verify. Do not wait for a specific platform to force your hand, because the same data hygiene wins on Google, Amazon, and Walmart at once. And before you enable agent checkout anywhere, read the terms on merchant of record, data sharing, and returns liability so you know what you are trading for the placement.

What Google actually shipped

Two pieces matter. The first is UCP itself, the plumbing that connects an AI agent to a merchant’s catalog, checkout, and payment so a purchase can complete without a browser session on your site. The second is Universal Cart, a cross-merchant cart that follows the shopper across Search, Gemini, YouTube, and Gmail, letting them add items from different retailers while they browse or chat and check out in one place. Underneath both sits the Agent Payments Protocol, Google’s standard for letting an agent pay on a shopper’s behalf within guardrails the shopper sets.

Put together, the shopper describes what they want, the agent assembles options from multiple merchants, and the purchase closes inside Google. Your brand still fulfills and still owns the customer as merchant of record, but the moment of discovery and the moment of decision both move upstream, into a surface you do not control.

Why this matters even if you live on Amazon and Walmart

It is tempting to file this under DTC and move on. That is a mistake. Amazon has spent two years pushing sellers to optimize for Rufus, and Walmart has wired Gemini into its own funnel. The mechanism is the same everywhere: an agent reads structured product data and makes a call. The brands that treat this as one problem, clean machine-readable data everywhere they sell, get compounding returns as each platform turns agent-led buying on. The brands that treat Google, Amazon, and Walmart as three separate optimization projects will be perpetually behind on all three.

There is a second reason. Agent-mediated buying compresses the top of the funnel. When an assistant returns three options instead of a page of thirty, the long tail of weakly differentiated listings gets squeezed out of consideration entirely. Being findable is no longer enough. You have to be the option an agent can confidently justify, which comes down to the signals it can actually read: accurate specs, competitive and current pricing, in-stock status, strong review volume, and a clean returns story.

What changes in your funnel

Three things shift. Discovery moves into the assistant, so the query you used to win on the search results page now gets answered before the shopper ever sees a list of blue links. Conversion moves to whether the agent selects you, which means your price, availability, and attribute completeness become ranking inputs, not just merchandising details. And the customer relationship gets thinner at the moment of sale, because the agent handles the transaction, so your post-purchase flow, your email capture, and your retention motion have to work harder to reconnect with a buyer who never hit your site.

None of that is a reason to sit out. It is a reason to make sure the data an agent reads is better than your competition’s.

What to do now

Start with the feed. Audit every SKU for complete, accurate attributes, and fix the gaps that make a product ambiguous to a machine, sizing, materials, compatibility, use case. Make sure price and inventory sync in real time, because an agent that quotes a stale price or recommends an out-of-stock item once will stop trusting your catalog. Expose your reviews as structured data so ratings and counts travel with the product. Write your returns and shipping policy in plain, machine-readable terms, because an agent weighs friction the way a careful shopper does. Then rebuild your post-purchase capture assuming the buyer never saw your site, with fulfillment inserts, follow-up email, and a reason to come back direct next time.

Do this once, correctly, and it pays off on every surface that adopts agent-led buying, which is now all of the big ones.

The takeaway

Agentic checkout is not a distant trend. The protocol is live, the partners are named, and the same shift is underway inside Amazon and Walmart. The work it demands is unglamorous and entirely within your control: clean, complete, real-time product data and a post-purchase motion that survives a sale you did not host. The brands that get parsed, trusted, and chosen by an agent will quietly take share from the ones that stayed optimized for a search results page that fewer shoppers ever see.

If you want a read on how agent-ready your catalog is across Amazon, Walmart, and your own store, and a prioritized list of what to fix first, book a strategy assessment and we will audit it against your feed.