Universal Cart is coming to the UK. Here's the product-data homework to do before it arrives.

Google's Universal Cart is already live in the US, letting people buy from brands like Nike and Wayfair without leaving Search or Gemini. Google has said the UK comes later, which is exactly the problem. "Later" feels like permission to wait, and waiting is how retailers get caught out. When the buying moment moves off your website, the only thing representing you is your product data, and most UK feeds are nowhere near ready for an AI agent to read them cold.

This is a plain-English look at what Universal Cart is, why the UK gap is an opportunity rather than a reprieve, and the specific homework worth doing on your product data now, before the feature lands. Universal Cart is one piece of the wider Google AI search shift, and this is the product-data half of getting ready for it.

What Universal Cart actually is

Universal Cart is Google's answer to a simple question: why should you have to visit five different websites to buy five different things? Google describes it as a shopping cart that works across retailers and across its own services, so a shopper can add items from different shops and check out with Google Pay in a few taps, without ever landing on your site. Some journeys still hand the shopper over to the retailer to complete the purchase, but the discovery, the comparison and increasingly the buying itself happen inside Google.

It is live in the US now, with named partners including Nike, Sephora, Target, Ulta Beauty, Walmart and Wayfair, plus Shopify merchants such as Fenty and Steve Madden. Payment options go beyond Google Pay to include Affirm and Klarna, so the checkout is being built to remove as much friction as Google can.

The important shift is not the cart. It is where the decision gets made. When a shopper buys inside Search or Gemini, they are choosing between products based on the data Google can read about each one, not the experience of visiting your website. Your homepage charm, your brand story, your carefully designed product page, none of it is in the room. Your feed is.

Why "later" is the trap, not the reprieve

Google has been specific about the order of the rollout. Canada and Australia come in the coming months. The UK is described simply as "later". It is tempting to read that as breathing room, and it is the opposite.

Here is why. The retailers who win inside a new Google feature are almost never the ones who react when it launches in their market. They are the ones whose data was already clean when the switch was flipped. As Syed Ali, COO of agency Honcho, puts it:

"Agentic commerce is Google stepping directly into the role of the shop floor. It is no longer just helping people find products, it is helping them decide and buy. That changes everything about how brands need to think about visibility, offers and conversion."

Syed Ali, COO, Honcho

The UK gap is the window to get your shop floor in order before the customers arrive. Once Universal Cart is live here, fixing a thin feed is a scramble. Doing it now is a project.

When the buying moment leaves your website

Think about what an AI agent knows about your product when it assembles a Universal Cart. It does not browse your site the way a person does. It reads structured data: your product feed, your schema markup, your prices, your images, your stock status. If a detail is missing, wrong or ambiguous, the agent does not email you to check. It either guesses, or it moves on to a competitor whose data is unambiguous.

That is the real change for UK retailers. For years, a weak feed cost you a bit of Shopping performance and you could make up the ground with a strong website and good ads. In an agentic checkout, the feed is not one signal among many. It is the whole shopfront. Being invisible to the agent is being invisible full stop.

The product-data homework worth doing now

None of this requires new technology. It requires your existing product data to be complete, accurate and machine-readable. Here is the homework, in priority order.

• Feed completeness. Every product should have a full, accurate title, description, price, availability, brand and GTIN. Gaps and "N/A" fields are where agents lose confidence. This is the foundation everything else sits on.
• Structured data (schema) on your product pages. Product, price, availability and review schema act as the translation layer that lets Google and non-Google AI read your pages reliably. Cheap builds routinely skip it, and it is one of the lowest-cost AI-readiness moves you can make. We go deeper in our guide to Product, Review and FAQ schema.
• Price accuracy and competitiveness. Agents compare ruthlessly on data they trust. A price that is stale in your feed, or quietly higher than a comparable retailer, is a reason to be left off the shortlist. Your feed price and your site price must match, always.
• Images that work for machines and people. Clear, correctly attributed primary images, plus secondary images showing the product in real use. Visual search increasingly rewards context over sterile pack shots.
• Stock and delivery accuracy. Availability and shipping data have to be live and correct. An agent that surfaces an out-of-stock product once learns not to trust your feed.
• Conversational attributes. Google's newer feed fields, things like a product Q&A, related products and variant grouping, are exactly what an AI reaches for when it answers a shopper's question. Most UK retailers have done nothing with them yet, which is precisely why they are an advantage. We cover them in our piece on Conversational Attributes in Merchant Centre.

If that list looks like a lot, the honest reframing is this: it is the same data hygiene that already decides your Shopping performance today. Universal Cart just raises the stakes on getting it right. If you are also weighing whether to opt in at all, we cover that decision in the stay-or-go question every UK retailer faces.

How to know if you are actually ready

The uncomfortable test is to imagine an AI agent has been asked to buy the kind of product you sell, and the only thing it knows about you is your feed. Would it understand what you sell, trust the details, and put you on the shortlist? For a lot of UK retailers the honest answer is "I don't know", and not knowing is the problem worth solving before the feature arrives, not after.

What we do, and where to start

We help UK retailers get their product data ready for AI-led shopping before it becomes the default, so you are on the shortlist rather than scrambling to get back on it.

The sensible first step is a Technical Audit, which checks your feed completeness, schema markup, price and image data, and the conversational attributes most retailers have ignored, then gives you a prioritised list of what to fix. If you want the wider picture of how findable you are to AI right now, an AI Visibility Report tests where you actually show up when a shopper asks an AI for a product like yours, and who gets recommended instead.

Neither is a leap. Both are a way of doing the homework while you still have the room to do it properly. If you would rather just talk it through, tell us the challenge and we will point you at the shortest route.

Frequently asked questions

Is Universal Cart live in the UK yet?

No. As of July 2026 it is live in the US, with Canada and Australia named next and the UK described by Google as coming "later". That gap is the window to prepare.

Do I need a new website for this?

Usually not. Universal Cart reads your product data, so the work is in your feed, your prices, your images and your schema markup, not necessarily a rebuild. A technical audit tells you which of those need attention.

What is the single most important thing to fix first?

Feed completeness and accuracy. If your titles, prices, availability and GTINs are wrong or missing, nothing else you do can rescue you, because the agent has nothing reliable to work from.