Google Universal Cart: Walmart & Target

Executive Summary

Google’s Universal Cart — a single persistent shopping layer spanning Search, Gemini, YouTube, and Gmail — is rolling out to Walmart, Target, Nike, Sephora, Ulta Beauty, Wayfair, and Shopify merchants this summer 2026.
For brand managers, the business consequence is unambiguous: products not structured for AI agent discovery will be invisible in Universal Cart, losing purchase consideration without ever appearing in a results page.
The counterintuitive angle: Walmart and Target — both of which have spent hundreds of millions building proprietary AI shopping tools — signed on anyway, suggesting they believe Google will dominate the discovery moment even if it doesn’t own the transaction.

There’s a moment that keeps recurring in conversations with brand managers right now, and it goes something like this: “We optimized for Amazon search, then we optimized for Google Shopping, and now you’re telling me I have to optimize for a cart that lives inside someone’s Gmail?” Yes. That’s exactly what I’m telling you.

At Google I/O on May 19, Google unveiled the Universal Cart — an AI-powered shopping hub that follows a shopper across every Google surface. A user searches for sneakers in Google Search, adds a pair. They ask Gemini to compare laptops, add one to the same cart. They watch a YouTube review of a backpack, add it. A promotional email arrives in Gmail for a jacket they’d been eyeing — one click, into the cart. Price history, stock alerts, and compatibility warnings included. One checkout.

As of this week, that cart is moving from announcement to reality. Modern Retail reports that Walmart and Target — along with Nike, Sephora, Ulta Beauty, Wayfair, and a number of Shopify merchants including Fenty and Steve Madden — are preparing to go live this summer in the United States. YouTube and Gmail integrations will follow, with Canada, Australia, and the UK coming later in 2026.

One Cart, Every Surface — and the Protocol Nobody Mentions

What makes Universal Cart different from Google’s previous commerce attempts (and there have been several: Google Checkout, Buy on Google, Google Express) is the infrastructure underneath it. Google co-developed the Universal Commerce Protocol (UCP) with retail partners in January 2026. UCP is an open standard — a common language for AI-driven commerce that allows checkout to happen either directly through Google or as a handoff to the retailer’s own site.

That last part matters more than it might seem. Retailers were not going to hand Google their checkout pipeline without some guarantee they could keep the transaction relationship. UCP gives them a lane: Google handles discovery and cart management, the retailer can retain the purchase, the loyalty point, the post-transaction data. Whether brands in those retailers’ catalogs retain any meaningful relationship with the end customer is a different question — and one most brand managers haven’t asked yet.

What we’re seeing at Epinium is that the brands who will thrive in agentic commerce aren’t necessarily the ones with the biggest budgets. They’re the ones whose product data is clean, complete, and structured for machine interpretation — not just for human search queries.

Epinium data

In five-plus years of auditing catalog data for brands selling across Amazon and major European marketplaces, Epinium’s operations team has consistently found that over one-third of SKUs in a newly onboarded brand’s catalog arrive with missing or incorrect attribute fields. That figure was a listing quality problem before AI-powered shopping existed. In 2026, it is a discoverability problem at scale — every empty field is a signal that an AI agent will interpret as low confidence.

The Universal Cart doesn’t browse. It decides. An AI agent evaluating two similar products in a shopper’s consideration set will surface the one with richer, more complete structured data — not necessarily the one with the better human-readable description. This is a fundamental shift in what ‘good’ catalog content means.

If you want to understand how AI agents actually rank and choose products, the framework in Agentic Commerce: How Brands Get Chosen by AI Agents is still the clearest map we’ve seen.

FREE TRIAL Explore Platform — audit your catalog data for AI agent readiness → ✓ 7 days free ✓ No card required ✓ Your own data

Why Walmart and Target Said Yes — What Isn’t Being Said

The detail that should give brand managers pause isn’t that Universal Cart exists. It’s that Walmart and Target chose to join it. Both companies have invested aggressively in proprietary AI shopping infrastructure. Walmart has Sparky. Target has its own personalization stack. Neither company is naive about ceding ground to Google.

Their reasoning appears to be distribution arithmetic: Google reaches 90%+ of US internet users through Search alone. If the discovery moment is shifting to AI-powered surfaces — Gemini, AI Overviews, YouTube recommendations — then refusing to participate in Universal Cart doesn’t protect their platforms. It just means their products appear in a Google surface that routes purchase intent somewhere else.

That’s the real signal here. These retailers aren’t endorsing Google’s checkout vision. They’re hedging against being left out of AI-mediated discovery. And that same calculation applies to every brand that sells through them.

There’s an honest counterpoint worth raising: Google has abandoned commerce products before. Buy on Google was shut down in 2023. The Universal Cart faces the same adoption challenge every previous attempt faced — consumer habit change is slow, and buying something through Gmail feels strange to most shoppers today. Google’s own timeline (summer US rollout, international expansion later) is cautious. The technology exists; the behavior change does not yet.

But the structural stakes are different this time. UCP as an open standard means the infrastructure outlives any single Google product. If Google retreats again, the protocol stays. Other AI surfaces — and there will be many — can adopt it. The catalog data work brands do now isn’t spent on a Google gamble. It’s invested in a new standard for machine-readable commerce.

The 1,300-brand case study in Why 1,300 Brands Now Pay to Track Their AI Search Presence shows exactly how fast this gap opens once AI-mediated discovery takes hold.

What Brand Managers Should Do Before Summer

The honest answer is: run a catalog audit. Not a rewrite, not a redesign — a specific check against the attribute completeness requirements of the UCP feed specification and Google’s Merchant Center enrichment guidelines. The brands that will surface in Universal Cart are not the ones that pay Google more. They’re the ones whose data the AI can parse with confidence.

Three things to check immediately: attribute completeness (every required field populated, no placeholder text), image quality for AI vision models (clean backgrounds, multiple angles, no lifestyle-only shots), and category taxonomy alignment (your internal categories mapped to Google’s product taxonomy). These are unglamorous tasks. They’re also the ones that determine whether your product appears in a Walmart shopper’s Gemini cart this summer.

Frequently Asked Questions

What is Google Universal Cart and how is it different from Google Shopping?

Google Shopping is a product listing and ad surface — you appear there by running Shopping campaigns or having a Merchant Center feed. Universal Cart is a persistent shopping layer that travels across Google’s entire ecosystem: Search, Gemini, YouTube, and Gmail. A shopper can add a product from a YouTube video directly to the same cart they’re building in a Gemini conversation. It’s not a new ad product; it’s a new commerce infrastructure layer built on the Universal Commerce Protocol (UCP) open standard.

Does a brand need to do anything specific to appear in Universal Cart?

Yes. Being in a participating retailer’s catalog (Walmart, Target, Nike, Sephora, Ulta, Wayfair, or Shopify-powered stores) is the entry point — Universal Cart surfaces products from those catalogs. But appearing prominently requires that your product data meets UCP feed specifications and Google Merchant Center enrichment standards. Brands with incomplete attribute data, low-quality images, or misaligned category taxonomy will be systematically deprioritized by the AI agent managing cart curation. Submitting a Merchant Center feed is necessary but not sufficient.

Can smaller brands that don’t sell through Walmart or Target participate?

Yes — Shopify merchants are included as launch partners, which opens Universal Cart to a very large base of direct-to-consumer brands. Brands selling through Shopify-powered storefronts that have enabled UCP-compliant checkout will have their products eligible for Universal Cart surfaces. The Fenty and Steve Madden inclusion in the launch list confirms this. Brands outside these channels will need to wait for broader UCP adoption or prioritize getting onto a participating retail platform.

What happens to customer data and purchase attribution when someone buys through Google?

The UCP model supports two flows: checkout directly through Google (Google handles transaction, retailer receives the order), or a handoff where the shopper is routed to the retailer’s own checkout. In the Google-handled case, customer purchase data is managed under Google’s privacy terms — the retailer receives order data but may not receive the full behavioral profile. Attribution becomes complex: Google counts the assist, the retailer records the sale, and the brand may see it as a retail order with no visibility into the discovery journey. Brands should pressure-test this with their retail partners before summer.

Is there a risk that Google becomes a middleman that erodes brand-consumer relationships?

This is the most important question and the least asked. When a shopper buys through Universal Cart, they experienced the product through a Gemini recommendation, not a brand story. They added it to a Google cart, not a brand cart. The loyalty point may go to the retailer; the repurchase prompt may come from Google. Brands that rely on direct discovery — brand search, owned channels, direct-to-consumer — face a structural risk if they let AI-mediated commerce become the primary purchase path. The mitigation isn’t to avoid Google. It’s to ensure that AI agents recommending your product have enough structured signal to reinforce brand identity, not just match a category query.

The Universal Cart is not a threat to well-prepared brands — it’s an amplifier. The brands with clean, complete, AI-readable catalog data will appear in more carts, on more surfaces, with less paid media required to get there. The brands that aren’t ready will discover what invisibility looks like in 2026: their products exist, but no AI agent has enough confidence to surface them.

Google has tried to own commerce before. The difference this time isn’t the cart. It’s the protocol — and the seven major retailers who just agreed to speak it.

Ready to audit your catalog data for AI agent readiness? Epinium Platform gives brand managers and catalog teams a clear view of attribute completeness, AI-readiness scores, and the specific gaps that prevent products from surfacing in AI-powered shopping experiences. Start your free 7-day trial →

No card required · 7 days free · Your own data