TL;DR
Google is rolling out AI-powered shopping built on Gemini models and a 50-billion-item shopping graph, turning product discovery into a conversation instead of a filter hunt.
Consumers will be able to shop by describing what they want, get follow-up questions from the AI, and check out without leaving Google.
Merchants can get started today with nothing more than a spreadsheet upload to Google Merchant Center — no technical feed required.
84% of small businesses plan to increase their use of technology platforms (U.S. Chamber of Commerce), but the real bottleneck isn't merchant readiness... it's consumer adoption.
Get your products into Merchant Center now. The wave is coming, and the businesses already in the graph will have the head start.
If you've ever Googled a product and seen the shopping carousel across the top of the results, you already know Google's been in the e-commerce game for a while. Merchants send Google a product feed, customers browse, and Google charges for the click.
What's changing is the experience. A lot.
From filters to conversation: what Google actually built
Google is weaving Gemini models into its shopping surface, and the backend powering it is almost hard to believe: a 50-billion-item shopping graph designed to transform online shopping into a "conversational, visual, and agentic experience."
50 billion with a B. That's not a product catalog. That's a searchable model of almost everything available to buy.
In practice, what this looks like is closer to vibe shopping than traditional search. Think of it like vibe coding, where you describe what you want in plain language and the results appear alongside your prompt. A customer searches for a travel bag. The AI asks follow-up questions: Where are you taking it? What style? What size fits in an overhead compartment? It filters intelligently based on the conversation, not just keywords.
The old version of that experience was: pick a category, set your filters, sort by price, scroll. The new version is: just tell it what you want.
And for merchants? If your products are in Google's graph, you're in that conversation.
B2AI: the channel shift that's already happening
Last week we talked about Visa's moves into AI purchasing. The same thread runs through this story.
B2B and B2C aren't the only channels anymore. There's now B2AI: selling to an AI agent that's making (or heavily influencing) the purchase decision on behalf of a human. Whether it's ChatGPT shopping, Google's agentic checkout, or something that doesn't exist yet by the end of this year... the pattern is the same. Your products need to be findable by AI, not just by humans typing keywords.
Google's version goes a step further by closing the loop. Because Google already has your payment information (and has for years if you've used Google Pay), the checkout path compresses from "discover, click, visit merchant site, enter card, confirm" to something closer to: discover, say yes, done. Google's own reporting describes AI agents that can offer discounts and deals to shoppers at exactly the moment they might tip from browsing to buying.
That's a meaningful shift in how purchase decisions happen.
How merchants get in (and it's easier than you think)
Here's the good news for small business owners: you don't need a developer or a custom API to start.
Google already runs a free digital dashboard called Google Merchant Center. It lets you upload your products and inventory so they can be discovered across Google Search, the Shopping tab, Maps, and YouTube. You can start with a spreadsheet. Upload your product names, descriptions, prices, and images, and you're in.
When the agentic shopping experience rolls out more broadly, the businesses already in Merchant Center will have a head start. The feed you build now is the same infrastructure that feeds into the AI layer.
The U.S. Chamber of Commerce's Empowering Small Business report found that 84% of small businesses plan to increase their use of technology platforms. That tracks. Merchants will rush to get their products in front of this new surface. The question is whether consumers follow at the same pace.
The trust gap: big purchases aren't going fully agentic yet
Not every purchase is going to run through an AI agent, at least not right away.
Think about mattress shopping. You've got Purple, Casper, Eight Sleep, and a dozen other brands with slick marketing and strong reviews. The problem is it's hard to know if the product is legitimately good or if the marketing is just really, really good. Most people buying a mattress for the first time still want to watch YouTube reviews, read third-party comparisons, and maybe visit a store.
Same with cameras, TVs, cars... anything where the stakes are high enough that you want proof before you commit.
The AI shopping experience will earn trust fastest on commodity purchases and replenishment orders. Order more toothpaste. Grab a five-pack of candles under $20. Reorder the same Swiffer heads you bought last time. For those, most people are happy to let the AI handle it, especially if you can set a price ceiling and just confirm before it goes through.
For first-time, high-consideration purchases? Consumers will still want the video, the review, the third-party opinion. That's not a flaw in the system... it's just where trust needs more time to build.
What the business model probably looks like
Based on how Google Shopping works today, here's the likely path:
You get organic discovery for free once your products are in Merchant Center. You can then bid to appear higher in results, and Google charges per click (same as search ads). That part isn't new.
What could be different, and this would be a huge win for small businesses, is a model closer to Google's Local Services Ads: you don't pay for the click, you pay for the verified lead or the completed purchase. If the checkout happens inside Google, they have all the data to make that work. No more paying for traffic that bounces. You pay when someone actually buys.
That changes the math for small businesses significantly. If you know your customer lifetime value, you can work backwards to figure out exactly what you're willing to spend per acquisition. The underlying strategy doesn't change... Google is just making it easier to run that math in a place where buyers are already spending time.
If you want to stress-test your own numbers before leaning into a new paid channel, the free strategy diagnosis quiz can help you figure out where marketing actually fits in your growth picture.
What to do this week
If you sell physical products (or even digital ones with clear listings), the move is simple: get into Google Merchant Center now.
Start with a spreadsheet. Add your products, descriptions, prices, and images. You don't need a developer to start, and the platform will walk you through the setup. The AI shopping layer is being built on top of that existing infrastructure. Merchants already in the graph will have the head start when the conversational and agentic experience rolls out more broadly.
Consumer adoption will be the slow variable. Google has to earn that trust the same way Amazon did with one-click checkout and Apple did with Apple Pay: make it frictionless enough that people stop thinking about it and just buy. That takes time. But merchants who wait for consumer adoption before getting their products listed will already be behind.
The conversation is already happening. Make sure your products are in it.