Google Universal Cart lets shoppers add products and check out inside Google Search, Gemini, YouTube, and even your email without ever visiting your website. Google launched it on May 19 with Nike, Walmart, Sephora, and Wayfair on board, so the checkout page you control is no longer a guaranteed part of the sale.
At Google’s developer conference, a Google engineer typed a question into Gemini, the AI picked a product, and money moved in twenty-three seconds. The shopper never visited a website, a checkout page, or a store.
I have been watching this shift play out inside my own store, and it should worry every seller who is not paying attention yet. Here is how Universal Cart works, what it takes away from your store, and the three steps that get your products picked by AI shopping agents.
Get My Free Mini Course On How To Start A Successful Ecommerce Store
If you are interested in starting an ecommerce business, I put together a comprehensive package of resources that will help you launch your own online store from complete scratch. Be sure to grab it before you leave!
Table of Contents
Key takeaways
- Universal Cart moves checkout onto Google, across Search, Gemini, YouTube, and email.
- AI shopping is getting big fast. Morgan Stanley estimates AI shopping agents could drive up to $385 billion in US ecommerce spending by 2030.
- The Universal Commerce Protocol (UCP) reads your live data through an API, so accurate stores get recommended with confidence.
- You stay merchant of record, and you still lose abandoned-cart flows and on-page upsells.
- AI agents can’t see what they can’t read. Amazon blocks AI crawlers, so even top Amazon listings can be invisible to AI.
- Win with clean product data, fresh reviews, trusted mentions, and named demand that makes shoppers ask for your brand.
How Google Universal Cart works across Search, Gemini, and YouTube
Universal Cart is one shopping cart that lives across everything Google owns. A shopper can add a product while searching, chatting with Gemini, watching YouTube, or reading email, and the cart then works on its own.
It hunts for a lower price, tracks the price history, and pings the shopper when an out-of-stock item comes back.
It can even look at the parts in the cart and warn the buyer when two of them will not work together.
For my entire career, checkout was the one piece I controlled. People found my handkerchiefs through search or an ad, came to my store, and bought on my turf, and Google just pulled that final step into its own backyard.
Which brands have signed on to Google Universal Cart
Nike, Walmart, Sephora, and Wayfair are Universal Cart launch partners, along with Shopify sellers like Fenty and Steve Madden.
When the largest retailers jump in this quickly, it tells you the shift is real.
How Google’s virtual try-on keeps shoppers on Google
Google’s virtual try-on tool keeps shoppers on Google by letting them upload a photo inside search and see how your clothing looks on their own body before they ever land on your site.
It is the same trend as Universal Cart. Research, comparison shopping, and buying are all migrating onto Google’s turf.
What the Universal Commerce Protocol (UCP) does for your store
The Universal Commerce Protocol, or UCP, is a common language that lets any AI shopping agent read your store and complete a sale. Google co-built it with Shopify, Etsy, Wayfair, Target, and Walmart, and Visa, Mastercard, Stripe, and Home Depot back it.
UCP is API-based, which means Google can look up your actual products, real prices, and live inventory the instant a shopper asks. The old way relied on a crawled copy of your site from days or weeks ago, a snapshot that was usually out of date as soon as it was taken.
Under the old system, a price change or a sold-out size could show the wrong information in search for days. With UCP, the agent sees what is true right now.
That makes your products more likely to get recommended. When the AI knows your data is live and accurate, it trusts you, and it surfaces the brands it trusts.
| Crawled product data (old way) | Universal Commerce Protocol | |
|---|---|---|
| Data source | A copy of your site crawled days or weeks ago | Live lookup through an API |
| Prices and stock | Can be wrong for days after a change | Accurate in real time |
| Checkout | Shopper clicks through to your store | The AI agent can complete the sale |
| AI confidence | Lower, because the data may be stale | Higher, so the agent recommends you more readily |
Why AI shopping is growing so fast
AI shopping is growing fast because shoppers already use AI to research and compare, and Google just filled the last gap by letting them finish the purchase in the chat or video. In my own store, sales from AI-driven traffic have grown exponentially over the last year.
I sell handkerchiefs, linens, and embroidered products, which is about as boring and old-school as a category gets. If AI shopping is hitting my category, it is hitting yours.
Wall Street is betting big on the shift. Morgan Stanley estimates that AI shopping agents could account for up to $385 billion in US ecommerce spending by 2030.
From Google’s side, Universal Cart is a smart, aggressive land grab, and I understand why it made the move. The problem is what it does to the seller on the other end of the transaction.
How agentic checkout skips your website
Agentic checkout means a shopper can add your product from a search result or a video and buy it without ever seeing your store. For twenty years the game was simple: get the click, bring them to your site, and let your homepage, product page, and reviews do the selling.
Universal Cart can skip every bit of that, so the page you spent months optimizing might never load.
Tushy, the bidet company, shows how much a store can depend on its own pages. When it started, almost nobody was searching for a bidet, so its job was convincing a skeptical first-time buyer on its own website with funny, disarming copy that made an awkward product feel normal.
Tushy grew from $8 million in revenue to $40 million in about a year, and one better landing page more than doubled its conversion rate, from 2.5 percent to 5.5 percent.
In a Universal Cart world, the customer may buy without seeing that page. The persuasion has to happen before the cart, because the product page may never get its chance.
Why AI shopping agents miss great products
AI shopping agents can only recommend products whose data they can find and read. When the cart compares your product against every competitor Google can see, an algorithm makes the buying decision, and a lot of great sellers are about to get a nasty surprise.
A friend of mine sells the top-rated standing desk on all of Amazon, with thousands of reviews. When you ask an AI which standing desk to buy, his never comes up.
The first reason is that Amazon has been blocking AI crawlers, adding bot after bot to its block list to keep companies like OpenAI and Google from scraping its product data.
All the reviews and ranking that made him number one on Amazon do the AI no good, because the AI cannot see inside Amazon.
The second reason is that his own website was never set up for an AI to read, so the agents had nothing clean to pull. He did everything right by the old rules and still ended up invisible under the new ones.
What sellers keep and lose with Universal Cart
With Universal Cart you keep merchant-of-record status, the customer’s email, and your post-purchase flows, and you lose abandoned-cart triggers plus the on-page upsells that fire at checkout.
As merchant of record, you keep the customer relationship and the transaction on paper. You still get the customer’s email, so you can follow up, win the repeat purchase, and run your post-purchase flows.
Some of what you count on for revenue quietly stops working. If a customer checks out inside Google and never lands on your site, there is no cart on your store to abandon, so your abandoned-cart sequence has nothing to trigger on.
The bigger loss is the moment of the sale. When checkout happens inside Google, the “frequently bought together” offers, order bumps, and “add this for 10% off” upsells on your product page never appear, and for many stores those upsells are a real chunk of average order value.
| What you keep | What you lose |
|---|---|
| Merchant of record status | Abandoned-cart flows for Google checkouts |
| The customer’s email address | On-page cross-sells and order bumps |
| Post-purchase and repeat-purchase flows | Your product page’s chance to persuade |
Merchant of record on paper means little when Google controls discovery, comparison, and the cart all at once. You are renting space inside Google’s store, and we watched this exact movie play out with Amazon, where the terms are great right up until the day they change.
How to prepare your store for agentic checkout
Preparing your store for agentic checkout takes three steps: get your product data and reviews in shape, earn mentions on the sites AI already trusts, and build named demand outside the algorithm.
Step 1: Clean up your product data and reviews
Clean product data means titles, prices, specs, and stock counts that are complete, accurate, and written in your customer’s language, because that data is what AI agents read. This boring work now matters more than your homepage design.
Vague descriptions and pretty marketing copy are invisible to an agent. It will pick your product only when your product data matches the query.
Google just added a tool for this.
Inside Merchant Center, Google rolled out conversational attributes, new feed fields built so the AI can answer the longer, more nuanced questions shoppers ask.
You can add a direct question and answer, a popularity rank, and related products straight into your feed. The fields are optional, which is exactly why most sellers will ignore them and why filling them in is your opportunity.
The store owner who fills these in writes the answers the AI gives a shopper about their product.
Reviews matter just as much, because agents lean hard on fresh customer reviews as a live trust signal. A store with rich, recent, genuine reviews gets mentioned, and a store with three reviews from 2022 gets skipped.
That review request email you keep meaning to automate just became a ranking factor, so set it up this week.
Your store also needs UCP support, because UCP is the pipe that lets an agent finish a sale. Shopify stores are in good shape, because Google named Shopify in the launch and the plumbing is built into the platform.
If you are on another platform, ask whether it supports UCP yet. An agent that cannot complete a checkout on your store will quietly buy from a competitor whose checkout it can finish.
Step 2: Earn brand mentions on the sites AI already trusts
The sites AI already trusts are Reddit threads, YouTube reviews, and best-of roundups, and earning genuine brand mentions there is the biggest agentic shopping opportunity almost nobody is acting on yet.
Reddit is now the single most-cited source across every major AI engine, responsible for roughly 40 percent of all citations.
Google made that official on May 7, when it started pulling Reddit threads into search results as community perspectives. When a shopper asks an AI for the best product in your category, the answer often comes from a Reddit thread, a YouTube review, or a best-of roundup.
If your brand genuinely shows up in those conversations, you get pulled into the recommendation. If it never shows up, you stay invisible no matter how good your product is.
The right way to earn Reddit mentions is to show up in the communities where your customers already hang out and answer real questions like a human being. Get real customers talking about you and work your way into honest roundups and reviews.
Fake posts hyping your own brand will get you banned.
Step 3: Build named demand outside the algorithm
You build named demand outside the algorithm through podcasts, press, newsletter and creator sponsorships, and partnerships with people your buyer trusts, so customers ask for your brand by name and the AI has to deliver your products.
Graza, the squeeze-bottle olive oil brand, reportedly crossed into the tens of millions in revenue within a couple of years.
Graza got there through founder PR, podcast appearances, chef partnerships, and social media. When it was small, that named demand was its entire moat, and it still is.
Step 2 earns your way into the sources the AI reads, and this step is the layer above it. Get on podcasts in your category, pitch press, sponsor newsletters and creators your customer already follows, and partner with people who carry weight with your buyer.
When a customer asks for your brand by name, the AI cannot quietly swap in the cheapest alternative.
Here is a simple test of where you stand. Ask Gemini or ChatGPT to recommend a product in your category today and see whether your name comes up, and if it does not, that is your homework.
Why sellers who move first will win the agentic shopping shift
Sellers who move first win the agentic shopping shift because they adapt while everyone else is still in denial. I have watched that exact pattern play out through every major ecommerce shift in my nineteen years.
Agentic checkout is here to stay, so adapt early.
Frequently asked questions
What is Google Universal Cart?
Google Universal Cart is a single shopping cart that works across Google Search, Gemini, YouTube, and email, letting shoppers add products and check out without visiting the seller’s website. Google launched it on May 19 with partners including Nike, Walmart, Sephora, and Wayfair.
Do sellers still get the customer’s email with Universal Cart?
Sellers stay the merchant of record with Universal Cart, so they still receive the customer’s email and can run post-purchase and repeat-purchase flows. They lose abandoned-cart flows for Google checkouts and the upsells on their own product pages.
Does Shopify support Google’s Universal Commerce Protocol?
Shopify supports the Universal Commerce Protocol, which Google co-built with Shopify, Etsy, Wayfair, Target, and Walmart. Sellers on other platforms should ask their platform whether it supports UCP, because AI agents favor stores where they can complete checkout.
What are conversational attributes in Google Merchant Center?
Conversational attributes are optional Merchant Center feed fields that help Google’s AI answer detailed shopper questions. Sellers can add direct questions and answers, a popularity rank, and related products, so the AI gives shoppers the seller’s own answers.
Why do AI shopping assistants often miss top-rated Amazon products?
Amazon blocks many AI crawlers from reading its site, so AI assistants often cannot see Amazon reviews or rankings. Even a top-rated Amazon product can be invisible to AI unless the brand also has readable product data on its own website.
How much US ecommerce spending could AI shopping agents drive by 2030?
Morgan Stanley estimates that AI shopping agents could account for up to $385 billion in US ecommerce spending by 2030. Shoppers already use AI to research products, and tools like Universal Cart now let them finish the purchase too.
