Is Your Product Data Ready for Agentic Commerce?

Something shifted in late 2025, early 2026 that most e-commerce teams have not fully reckoned with yet.

Major retailers, including Gap, Walmart, Anthropologie, and Ralph Lauren, are no longer just optimising for search rankings. They are feeding structured product data (the attributes, the pricing, the availability, the descriptions) directly into AI platforms so that those platforms can discover, recommend, and complete purchases on their behalf, without the shopper ever visiting the retailer’s website.

This is agentic commerce. And it is not a future trend. It is live infrastructure, with real transactions happening right now.

We have covered the timeline of Google’s AI shopping features, from Virtual Try-On to AI Mode, in our guide to Google’s AI shopping updates for 2026. This piece is about what sits underneath those features: the commercial mechanics, the data requirements, and what they mean for your brand’s organic visibility in Google’s Universal Commerce Protocol (UCP).

What Agentic Commerce Actually Means

For most of the past decade, the online shopping journey followed a recognisable pattern. A shopper searched for something. They clicked through to a website. They browsed, compared, and (sometimes) bought.

AI is collapsing that journey.

Shoppers are now using platforms like Google Gemini, ChatGPT, Amazon’s Rufus, and Microsoft Copilot to describe what they need in natural language, receive curated product recommendations, and in some cases complete the purchase without leaving the AI platform at all.

According to eMarketer, AI platforms are projected to account for over $20 billion in US e-commerce sales in 2026, nearly four times the 2025 figure. More than 70% of shoppers have already integrated large language models into some part of their shopping journey. And shoppers arriving at retail sites from AI platforms are, according to Adobe data, thirty times more likely to purchase than average.

The behaviour shift is not gradual. It is structural.

Google’s own data shows that AI-assisted shopping queries are now two to three times longer than traditional keyword searches. A shopper no longer types “blue linen shirt.” They describe a context: “I need something smart casual to wear to an outdoor wedding in June, not too formal, ideally natural fabric.” The AI interprets the occasion, the formality, the season, and the fabric preference, and it returns products that fit.

That is contextual search. And it is where product discovery is heading.

The Platforms Driving This

It is worth understanding the landscape, because agentic commerce is not a Google-only story.

Google has built agentic checkout across both AI Mode in Search and the Gemini app, enabling purchases to be completed directly inside those interfaces via Google Pay. It launched UCP in January 2026, an open standard co-developed with Shopify, Target, Walmart, Wayfair, Etsy, and more than twenty other partners, to create a shared technical language between AI platforms and merchant systems.

Amazon’s Rufus assistant has been expanded significantly through 2025, providing conversational product guidance using catalogue data, customer reviews, and user context. Its Auto Buy functionality can authorise purchases automatically when a product hits a target price set by the shopper.

OpenAI introduced shopping panels inside ChatGPT in April 2025, allowing product discovery and purchase through third-party integrations. It has announced checkout capabilities, though its approach has been more iterative than Google’s.

Microsoft’s Copilot added price tracking, shopping panels, and in-platform checkout through 2025, and introduced a personalised shopping agent that retailers can embed directly into their own websites and apps.

Meta is enabling personalised recommendations and direct purchasing across Facebook, Instagram, and Shopify-powered sites through its Business AI.

Every major platform is building toward the same destination: a shopping experience that begins and potentially ends inside an AI interface, with the retailer’s product data doing the work that a website used to do.

The SEO Shift Nobody Is Talking About Enough

Here is the part that should concern every e-commerce marketing team.

Traditional SEO was about optimising pages to rank. Agentic commerce is about optimising data to be selected. Those are different disciplines, and most brands are still operating in the first mode while the competitive game has moved to the second.

Search Engine Land put it directly in their analysis of Google’s Universal Commerce Protocol: “Feed engineering is the new SEO because AI agents do not read your landing pages. They read your data.”

When an AI agent handles a query like “sustainable waterproof boots for a hiking trip in Scotland in November,” it is not crawling your product pages and interpreting your on-page copy. It is checking structured data fields against a standard. Is the material listed? Is the waterproofing attribute specified? Is stock availability current? Is the price accurate? Is there a return policy attached? Is November 2026 likely to be rainy?

A well-written product title with targeted keywords does very little in this context. A complete, accurate, attribute-rich product feed does everything.

The shift is from keywords to attributes. From page authority to data quality. From ranking to being selected.

URBN, which operates Anthropologie, Free People, and Urban Outfitters, shared their approach at NRF 2026. Rather than trying to optimise their entire catalogue at once, which runs to thousands of SKUs across multiple brands, they focused first on their highest-impact categories: dresses and denim. They standardised language, attributes, and taxonomy in those categories before expanding. That is the practical model for any retailer trying to get agent-ready without it becoming an overwhelming project.

What Google’s Universal Commerce Protocol Changes

UCP is worth understanding specifically, because it is the infrastructure standard that will govern how AI agents interact with UK retailers when European rollout arrives later in 2026.

At its core, UCP creates a shared language between AI platforms and merchant systems. It covers the full purchase cycle: discovery, cart creation, checkout, order confirmation, and post-purchase support. The goal is that any AI agent can transact with any merchant without requiring a bespoke integration for every platform.

For e-commerce brands, this means Google Merchant Centre has become significantly more important than most teams currently treat it.

Merchant Centre is no longer just the backend for Shopping ads for the performance marketing and paid media teams. It is the connection point between your product catalogue and Google’s entire AI commerce layer. Inventory, pricing, promotions, shipping terms, return policies, and product attributes all flow through it. When that data is incomplete, out of date, or inconsistent with what is on your live site, the AI cannot confidently recommend your products.

Google is also rolling out new Merchant Centre data fields specifically designed for conversational commerce: answers to common product questions, compatible accessories, and product substitutes. These are the fields that allow an AI agent to move from simply returning a product to genuinely recommending a solution. Brands that populate them early will have a meaningful advantage when these fields become standard requirements. OpenAI are doing the same.

One practical point worth flagging for Shopify merchants: Stripe’s Agentic Commerce Suite automatically supports all agentic protocols, including UCP, through a single setup. For brands on other platforms, Salesforce Commerce Cloud, Magento, or custom builds, a UCP compatibility assessment is worth doing now rather than when the UK rollout is already underway.

Traditional SEO Still Matters. Here Is Why.

It would be easy to read all of this and conclude that the job is now purely about data hygiene and feed management. That would be a mistake.

UCP and the broader agentic commerce infrastructure handle the transaction layer. They do not handle the trust layer.

Before an AI agent recommends a product, the underlying systems still need to make a judgement about whether to surface that brand at all. That judgement draws on domain authority, backlink profile, third-party brand mentions, and wider credibility signals. A brand with a clean product feed but weak organic authority will find its products are well-structured but not surfaced.

GEO, or Generative Engine Optimisation, AI Search extends SEO. It does not replace it. The brands best positioned for agentic commerce are the ones with strong foundations in both: authoritative domains, quality inbound links, and solid brand reputation, combined with clean, complete, machine-readable product data. Our AI Search and GEO Services cover how we approach this with clients across both layers.

Brands that pivot entirely to feed optimisation while letting their broader SEO foundations slide will face a visibility problem that is harder to diagnose and slower to fix.

What to Do Right Now

If you are an e-commerce brand or marketing team thinking about where to focus, here is a practical starting point.

Treat your Google Merchant Centre feed as a commercial asset, not an admin task. Audit it for missing attributes, pricing discrepancies between the feed and your live site, incomplete return policies, and out-of-date availability data. A feed error in a paid search context wastes some budget. In an agentic commerce context, it removes your product from the AI layer entirely. Our E-commerce SEO services include feed architecture review as part of how we build organic visibility for retail clients.

Start with your highest-impact categories. Do not try to optimise everything at once. Identify the categories where AI-assisted discovery is most likely to drive commercial value and standardise your data there first.

Look at your real-time data infrastructure. Agentic commerce platforms require current pricing and availability to function accurately. If your feed updates overnight in batches, that is a problem in an environment where AI agents are making live purchase decisions throughout the day.

Prepare for new attribute requirements. The conversational commerce fields now rolling out in Google Merchant Centre and OpenAI, common product questions, compatible products, substitutes, are a preview of where the standard is heading. Getting ahead of them now is straightforward. Playing catch-up when they become requirements is not.

And maintain your SEO fundamentals. Domain authority, quality content, earned links, and brand reputation are still the trust signals that determine whether your products get surfaced in the first place.

Our Expert Summary on Agentic Commerce

Agentic commerce is not waiting for brands to be ready. The platforms are live. The protocols are published. Major retailers across fashion, homewares, grocery, and consumer goods are already in production with AI-assisted checkout.

The competitive window is now, specifically because most brands are still treating this as something to monitor rather than something to act on.

The brands that will capture agentic commerce visibility are not necessarily the ones with the biggest budgets. They are the ones whose product data is clean, complete, and structured well enough for an AI to recommend them confidently. That is a solvable problem, and it starts with understanding what your data actually looks like from the AI’s perspective.

If you want to know where you stand, we offer a complimentary consultant-led AI visibility diagnostic audit for e-commerce brands. A comprehensive, practical audit of where you are losing AI search visibility against your competitors and what to do about it.