How Google’s Universal Commerce Protocol changes ecommerce SEO

For years, ecommerce ran on a simple model: Google drove traffic, and your site did the selling. Rankings, clicks, and conversion rate determined performance. That model just changed.

With the Universal Commerce Protocol (UCP) and AI Mode, Google can now discover, compare, and complete purchases inside its own AI experiences. Search is shifting from a traffic channel to a transaction layer. Visibility now depends on whether Google’s AI selects your product data.

When AI makes the recommendation and closes the sale, optimization moves upstream. The question isn’t just whether you rank. It’s whether you’re chosen.

Here’s what changed and what SEO and AI optimization teams need to do next.

The shift to agentic commerce

Google launched the Universal Commerce Protocol, or UCP, on Jan. 11. This new open standard is designed to let AI agents discover, evaluate, recommend, and purchase products across the web, all inside Google’s own AI experiences.

What stood out to me wasn’t just the protocol itself, but the ecosystem Google built around it. UCP was developed with platforms like Shopify, Etsy, Wayfair, Target, and Walmart, with payment networks already integrated. That kind of coordination suggests this was planned for the long haul, not just a quick test.

At the same time, Google rolled out three platform-level capabilities that make this real in day-to-day shopping:

• Business Agent gives brands an AI-powered representative inside Search and the Gemini app. Shoppers can ask product questions, compare options, and get brand-level guidance without visiting a website.
• Direct Offers allow merchants to inject exclusive discounts directly into Google’s AI Mode, so promotions now live inside the recommendation engine itself.
• Checkout in AI Mode lets Google complete purchases inside its own interface, turning Google from a traffic broker into a transaction layer.

More importantly, this allows Google to turn everyday conversation into commerce. Instead of waiting for shoppers to type product searches, Gemini can now respond to natural language prompts like “help me plan a camping trip” or “what will get wine out of my couch” by pulling live inventory, pricing, and availability from retailers, and completing the purchase in the same interaction.

Dig deeper: Are we ready for the agentic web?

What this means for ecommerce strategy

When AI intermediates the buying journey, brands compete inside the recommendation layer, not just in search results.

For most of my career, ecommerce worked the same way everywhere — search engines, ads, and marketplaces existed to send people to your site. Your site did the selling. UCP changes that model entirely.

Now AI handles the whole journey. It figures out what someone actually needs, compares the options, and can even complete the purchase. At that point, it doesn’t really matter how good your homepage or category page is if AI never chooses your product in the first place.

I saw this problem years ago, working with a large American candle retailer. People weren’t really shopping for candles. They were trying to get rid of pet smells, calm down after a long day, or make their house feel a certain way. But all we could give Google were scent names and product categories. 

If someone wanted something that killed pet odor without smelling like fake fruit, we probably had the perfect candle, but it almost never got shown because the data couldn’t express that situation.

That’s what changes here. With Gemini and UCP, people can finally describe what they’re dealing with, and the AI can map that to the right products in a brand’s catalog.

And when checkout happens inside Google, everything shifts. You don’t win because someone clicked your site. You win because the AI picked your product. Business Agent pushes that even further by letting brands show up right in the middle of that decision.

In real terms, that can be the difference between moving a few thousand units and moving 10 times that, without changing a single product, just because the right things are finally being matched to the right people.

This creates a very different kind of competition than what we’re used to. In the past, weak data or mediocre pages might push you lower in the results. Now, when product data is incomplete or inconsistent, the AI has little reason to consider you at all.

Brands are competing for inclusion in the system’s recommendation set. That shift changes where the storefront lives. It now exists wherever the AI presents options in that moment.

Dig deeper: Google outlines AI-powered, agent-driven future for shopping and ads in 2026

The new playbook: How SEO and AI optimization help

For a long time, SEO was framed as getting pages to line up with keywords. In reality, search engines have always been trying to understand products well enough to make decisions on a user’s behalf. What’s changing now is how explicit that decision-making has become.

Google is feeding AI Mode, Gemini, and Business Agent with product feeds and structured data, and it keeps adding more fields that describe how products actually get used. Things like common questions, what works with what, and what people buy instead when something is out of stock. That’s how the AI starts to reason, not just match words.

I saw this clearly while working with an outdoor apparel brand. Someone planning a trip to Europe wasn’t really searching for a jacket. They were thinking about rain, cold mornings, long walks, and changing weather. We had the right products, but shoppers had to guess which filters to click or which category to start in to find them.

With agentic commerce, that guesswork goes away. A shopper can just say, “I’m going to Europe in the spring, what jacket should I bring?” and Gemini can look at weather resistance, weight, breathability, and what’s actually in stock, then show the few options that make sense.

That’s what all these new attributes unlock. They let the AI understand products the way a good salesperson would. And when that happens, it’s not a small optimization. It can be the difference between a campaign barely working and one that suddenly takes off.

Dig deeper: How AI-driven shopping discovery changes product page optimization

Competing in the AI selection layer

What matters isn’t page position. It’s whether Google’s AI understands what a product is, who it’s for, and when it should be recommended.

When I worked with a high-end luxury jewelry retailer, one of our biggest challenges was building “user journey” pages. We had to create landing pages for things like anniversary gifts, modern gold, or minimalist style because shoppers weren’t searching for SKUs. They were searching for meaning: 

• “I need something that feels romantic.” 
• “This is for someone who loves simple gold instead of flashy diamonds.” 
• “I want it to look modern, not old-fashioned.”

Those pages worked, but they were slow to build, hard to keep fresh, and almost impossible to personalize.

With Gemini and UCP, that whole layer moves into AI. A shopper can just describe the person, the style, and the budget, and the system puts together the right products in real time. That feels less like search and more like having a personal shopper.

And none of that works without good product content. The descriptions, the specs, the reviews, even how people interact with the site are what give the AI something to reason with.

If your pages are thin or confusing, the AI has nothing solid to work from. For SEOs, this is the moment the fundamentals become decisive again.

The same goes for site experience. When people stick around, buy, and don’t return products, Google learns that your brand is a safe bet. In an AI-driven world, that trust is what keeps you in the recommendations.

Direct Offers then layer paid promotion on top of this organic selection system, creating a blended performance layer where feed quality, content quality, and media strategy all work together inside the AI buying experience.

Dig deeper: How SEO leaders can explain agentic AI to ecommerce executives

Using Google Merchant Center for agentic commerce

Product feed optimization essentials

Merchant Center has evolved beyond a Shopping ad upload tool. It now connects your entire retail operation to Google’s AI. Inventory, pricing, promotions, shipping, and product details all flow through it so Gemini can actually act on them. If that data is wrong or out of sync, the AI can’t confidently sell anything.

That’s why every field suddenly matters. Titles, descriptions, categories, GTINs, brand names, and images aren’t just metadata anymore. They’re how the AI knows what something is and whether it should trust it.

Google is also starting to add more human context into those feeds. Things like common questions, what accessories go with a product, what people buy instead, and how something is used in the real world. That’s how a machine starts to understand products the way a person does.

This is where a lot of brands get blindsided. A small pricing error, a feed that lags behind inventory, or a missing promotion is all it takes for products to quietly fall out of the AI layer. You might get an alert in Merchant Center, but if no one’s watching closely, the impact shows up in lost visibility long before anyone realizes what happened.

If you’re eligible, turning on Business Agent is part of this too. It lets your brand show up inside those AI conversations, not just as a product listing, but as something that can answer questions and close the sale.

And it’s not just the feed. Google is constantly comparing what you tell it in Merchant Center with what it sees on your site. When those two don’t line up, trust drops, and so does visibility.

Product feed optimization essentials checklist

• Complete all available attributes
  • Title, description, product type, Google product category.
  • GTINs, MPNs, brand identifiers.
  • Images – multiple angles, lifestyle shots.
• New conversational commerce attributes (coming soon)
  • Answers to common product questions.
  • Compatible accessories.
  • Product substitutes.
  • Use cases and scenarios.
• Feed quality signals
  • Price accuracy and competitiveness.
  • Availability and inventory status.
  • Shipping and return information.
  • Promotion data for Direct Offers eligibility.
• Business Agent activation
  • Eligible U.S. retailers can activate in Merchant Center.
  • Customize the AI agent’s voice to match brand.
  • Train the agent on product data – coming feature.
  • Enable direct purchases within the chat experience.
• Structured data alignment
  • Ensure website schema markup matches Merchant Center data.
  • Product schema, offer schema, and review schema all contribute to AI understanding.

Connecting Google Search Console to Merchant Center

This connection matters more than most people realize. Search Console used to tell you how pages were doing. Merchant Center tells you how products are doing. In an AI-driven world, those two things are finally tied together.

Linking them turns this from guesswork into something you can actually manage. You can see which products are getting picked up by Google, which ones are getting ignored, and where bad data is quietly killing your visibility. Disapproved items, missing attributes, price mismatches – all of that shows up right where you can act on it.

It also lets you watch how demand is shifting. You can see when impressions move from traditional search into Shopping and AI results, and which products are benefiting. That’s how you know whether your catalog is really working inside the AI layer or just sitting there hoping someone clicks a link.

What to monitor

Once everything is connected, this becomes your early warning system. It’s the dashboard you end up living in. It tells you which products are broken, which ones are being ignored, and which ones are quietly driving everything.

Performance reports show which items are actually getting seen and clicked, whether that’s in Shopping results or inside AI experiences. And the alerts are what save you from surprises. Price mismatches, crawl issues, or policy problems can quietly pull products out of the AI layer without you ever noticing unless you’re watching.

In an AI-driven commerce world, those small data issues don’t just hurt performance. They decide whether your products show up at all.

• Product feed diagnostics
  • Disapproved products and reasons.
  • Missing required attributes.
  • Data quality warnings.
• Performance insights
  • Click-through rates on product listings.
  • Impressions in Shopping results vs. organic.
  • Conversion tracking across channels.
• Issue alerts
  • Broken feeds or crawl errors.
  • Price and availability mismatches between site and feed.
  • Policy violations that limit visibility.
• Action items
  • Set up automatic alerts for feed issues.
  • Regular audits of product data completeness.
  • Monitor new conversational commerce metrics as Google rolls them out.

The future of ecommerce visibility

This shift to agentic commerce is already happening. AI is now deciding what to show and what to recommend.

I’ve seen brands struggle not because their products were wrong, but because the right products never reached the right people. That’s always been the gap in search. People know what they need. The system just doesn’t always connect the dots.

Agentic commerce starts to close that gap. When AI understands both the shopper and the catalog, it can finally match real needs to real products instead of forcing people to guess the right keywords.

That’s what Google has built here. Search has become a system for turning intent into answers.

So the work is clear. Keep your product data clean. Connect Search Console and Merchant Center. And start thinking about how people actually describe their problems, not just how they type queries.