Google has reworked its shopping experience around generative AI and the Gemini family of models. For consumers, the shift promises conversational product discovery, AI-generated comparison briefs, and — where available — automated “agentic” checkout that can buy on your behalf when preconditions are met. For merchants and developers, the new surface combines two sets of APIs (shopping / merchant APIs and Google’s GenAI / Gemini APIs) and requires updated feed practices, privacy controls, and technical integration.
Google AI Shopping is built on the Gemini API—its most intelligent AI models are currently the Gemini 3 Pro and Gemini 3 Flash—and the CometAPI also provides them.
What exactly is “Google’s AI shopping” ?
Google’s AI shopping umbrella includes conversational and agentic shopping inside Search and the Gemini app, AI-assisted product discovery (Shopping Graph + LLMs), and agent-driven checkout flows that can track prices and complete purchases on behalf of users. The goal: let users explain what they want conversationally (text, images, preferences), surface matching products, and — in some cases — let an AI “buy for me” when the price, size and other constraints are met. For merchants, this means discovery can happen without a shopper ever visiting your storefront; visibility now depends on how well Google understands your product data and how prepared your systems are to accept agent-driven requests.
Why this matters right now: At the National Retail Federation (NRF) conference earlier this month, Google unveiled the Universal Commerce Protocol (UCP) and a suite of "Agentic Commerce" tools that fundamentally shift how consumers discover and buy products.
Gone are the days of keyword stuffing and static product feeds. We have entered the era of the AI Shopping Agent—autonomous digital concierges capable of researching, negotiating, and executing purchases on behalf of users. For merchants, this isn't just a feature update; it is a rewriting of the rules of engagement.
What is the Universal Commerce Protocol and Why Does It Matter?
The most headline-grabbing announcement of 2026 is undoubtedly the Universal Commerce Protocol (UCP). To understand its magnitude, we must look at the friction plaguing modern e-commerce. Historically, if an AI assistant (like Gemini or ChatGPT) found a product for a user, the "buying" part was a clumsy hand-off: a link click that dumped the user onto a product page to start the checkout process from scratch. This friction caused cart abandonment rates to hover near 70%.
The "HTTP" of Shopping
UCP solves this by establishing a standardized language for commerce. Just as HTTP allows any browser to read any website, UCP allows any AI agent to communicate with any merchant system.
Co-developed with retail giants like Shopify, Walmart, and Target, UCP transforms the entire transaction stack—discovery, authentication, payment, and fulfillment—into a protocol layer. This means an AI agent can now:
- Query real-time stock without scraping.
- Negotiate pricing based on user loyalty status.
- Execute the payment using stored credentials (via Google Wallet or Apple Pay).
- Handle post-purchase support (returns/tracking) autonomously.
For merchants, the implication is stark: If your store cannot "speak" UCP, you are effectively invisible to the next generation of AI shoppers.
How Does Google’s AI Mode Change the Customer Journey?
The consumer-facing side of this revolution is Google Shopping AI Mode. Integrated directly into Search and the Gemini app, this interface replaces the traditional list of blue links with a dynamic, conversational workspace.
The Death of Keywords
In AI Mode, users no longer search for "men's running shoes size 10." They prompt: "I need marathon training shoes for flat feet, under $150, that I can pick up today in Austin."
Query Fan-Out and Deep Reasoning
Gemini uses a technique called Query Fan-Out. It breaks this single complex prompt into dozens of sub-queries:
- Check local inventory in Austin stores.
- Filter for "stability" shoes (for flat feet).
- Cross-reference pricing < $150.
- Analyze reviews for "marathon" durability.
It then synthesizes this information into a "Dynamic Product Panel"—a temporary, personalized micro-store built just for that user.
Agentic Checkout
Perhaps the most disruptive feature is Agentic Checkout. Once the user selects a product in the AI panel, they don't need to visit your site. They simply say, "Buy it." Google’s agent utilizes the UCP to securely transmit payment and shipping data to your backend, creating the order in your system as if the user had checked out natively. You remain the merchant of record, but the experience happens entirely off-site.
How do merchants get their products into Google’s AI surfaces?
The three entry points
- Merchant Center (feeds & APIs): Upload product data, images, pricing, availability, shipping and returns into Merchant Center. Google uses this canonical store of product truth for Shopping surfaces.
- Free listings + Paid ads: Products can appear in free listings (organic surfaces) and ads (Performance Max, AI Max). Google has been expanding how free listings feed into AI Mode search, so having clean, complete product data increases the chance of appearing in AI results.
- Structured data on your site & APIs: Schema/Product JSON-LD on product pages and programmatic APIs (Content API / Merchant API) help Google match and verify product information and—where supported—enable agentic checkout interactions. Recently Google signaled a transition to a newer Merchant API to streamline these integrations.
What should merchants and retailers do to prepare?
Goals for merchants
- Ensure product feeds are complete, accurate, and continuously synchronized (title, description, GTIN, availability, price, images, shipping, returns).
- Support structured data and the Merchant API / Merchant Center integration so Google’s Shopping Graph ingests canonical product data.
- Audit policies and checkout / returns workflows so agentic purchases can be fulfilled reliably.
How do merchants get their products into Google’s AI surfaces?
- Merchant Center (feeds & APIs): Upload product data, images, pricing, availability, shipping and returns into Merchant Center. Google uses this canonical store of product truth for Shopping surfaces.
- Free listings + Paid ads: Products can appear in free listings (organic surfaces) and ads (Performance Max, AI Max). Google has been expanding how free listings feed into AI Mode search, so having clean, complete product data increases the chance of appearing in AI results.
- Structured data on your site & APIs: Schema/Product JSON-LD on product pages and programmatic APIs (Content API / Merchant API) help Google match and verify product information and—where supported—enable agentic checkout interactions. Recently Google signaled a transition to a newer Merchant API to streamline these integrations.
Technical checklist (practical)
- Migrate or confirm API usage: Google announced the Merchant API (the successor to the Content API for Shopping); merchants should review the migration path and the API samples for product insertion and inventory updates. The Content API will be sunset (merchant teams should confirm timelines and adopt the Merchant API).
- Improve structured metadata: Google now offers generative marketing features that can create product copy, image variants and ad creatives from product data — a productivity multiplier for catalog-heavy sellers. Combine those with A/B testing and human review to avoid hallucinated or inaccurate product claims. Ensure product attributes (size, color variants, GTIN/ISBN, material, dimensions) follow Google’s Product Data Specification. Accurate metadata reduces misclassification by AI.
- Support real-time inventory signals: If a merchant expects to be selected for agentic checkout flows, low-latency inventory and accurate availability are mandatory to avoid failed orders. Use the Merchant API or supported local inventory APIs.
- Add 3D/AR assets and virtual try-on support: If you sell apparel, eyewear, jewelry or cosmetics, invest in 3D models and AR-enabled assets. Google surfaces these in virtual try-on experiences and they can materially improve conversion when surfaced in AI recommendations. Merchant help documents detail formats and publishing steps.
Schema: Implementing Agent-Readable Structured Data
You must go beyond basic Schema.org markup. AI agents require specific attributes to understand "soft" constraints like usage scenarios and sustainability.
Below is an example of advanced JSON-LD markup that includes returnPolicy, hasEnergyConsumptionDetails, and material, which are heavily weighted by the Gemini ranking algorithm.
<!-- Example: Advanced JSON-LD for Agentic Discovery -->
<script type="application/ld+json">
{
"@context": "https://schema.org/",
"@type": "Product",
"name": "Ultra-Grip Trail Runner 5000",
"image": [
"https://example.com/photos/1x1/photo.jpg",
"https://example.com/photos/4x3/photo.jpg"
],
"description": "Professional grade trail running shoe designed for wet terrain and flat feet.",
"sku": "0446310786",
"mpn": "925872",
"brand": {
"@type": "Brand",
"name": "ApexRun"
},
"review": {
"@type": "Review",
"reviewRating": {
"@type": "Rating",
"ratingValue": "4.5",
"bestRating": "5"
},
"author": {
"@type": "Person",
"name": "Jane Doe"
}
},
"aggregateRating": {
"@type": "AggregateRating",
"ratingValue": "4.4",
"reviewCount": "89"
},
"offers": {
"@type": "Offer",
"url": "https://example.com/anvil",
"priceCurrency": "USD",
"price": "149.99",
"priceValidUntil": "2026-11-20",
"itemCondition": "https://schema.org/NewCondition",
"availability": "https://schema.org/InStock",
"merchant": {
"@type": "Organization",
"name": "ApexRun Official Store"
},
"hasMerchantReturnPolicy": {
"@type": "MerchantReturnPolicy",
"applicableCountry": "US",
"returnPolicyCategory": "https://schema.org/MerchantReturnFiniteReturnWindow",
"merchantReturnDays": 30,
"returnMethod": "https://schema.org/ReturnByMail",
"returnFees": "https://schema.org/FreeReturn"
}
},
"material": "Recycled Polymer",
"audience": {
"@type": "Audience",
"audienceType": "Marathon Runners"
}
}
</script>
Integrate the Merchant/Content API (example)
To be surfaced reliably in AI Mode you should provide live product data via Google’s APIs rather than static file uploads. Google now provides the Merchant API as the recommended modern interface (examples and client libraries are available). Below is a brief example (Node.js) that demonstrates inserting/updating a product using the Content API / Merchant API client libraries.
samples and the canonical docs; follow the official samples and auth flows in production.
// Node.js example: insert a product into Merchant Center using googleapis
// Note: this example assumes you have set up OAuth2 or a service account and have
// merchantId. See Google Merchant API docs for full auth flow.
import {google} from 'googleapis';
async function insertProduct(merchantId) {
const auth = new google.auth.GoogleAuth({
scopes: ['https://www.googleapis.com/auth/content']
});
const authClient = await auth.getClient();
const content = google.content({version: 'v2.1', auth: authClient}); // or Merchant API version
const product = {
offerId: 'SKU-12345',
title: 'Waterproof Hiking Jacket — Pacific Series',
description: 'Lightweight insulated waterproof hiking jacket, breathable membrane.',
link: 'https://example.com/product/SKU-12345',
imageLink: 'https://example.com/images/SKU-12345-main.jpg',
contentLanguage: 'en',
targetCountry: 'US',
channel: 'online',
brand: 'TrailCo',
availability: 'in stock',
condition: 'new',
price: {value: '199.00', currency: 'USD'}
// add GTIN, shipping, tax, and other required fields as needed
};
try {
const res = await content.products.insert({
merchantId: merchantId,
resource: product // note: param is 'resource'
});
console.log('Inserted product:', res.data);
} catch (err) {
console.error('Error inserting product:', err);
}
}
// Usage
insertProduct('YOUR_MERCHANT_ID_HERE');
Why this matters: programmatic feeds mean AI surfaces see the freshest prices, stock, and variant information — and Google explicitly recommends APIs for automation and advanced use.
Exact method names, resource paths, and client library packages differ across evolutions of Google’s APIs — follow the Merchant API samples and sample GitHub repositories. Also, Google has announced the Content API will be replaced by the Merchant API (a migration window applies)
How does “agentic checkout” work and is it safe?
How it works
Agentic checkout is a higher-trust flow in which a user explicitly delegates a limited set of shopping actions to Google’s agent (Gemini). Typical mechanics: (1) user opts in and provides or confirms payment and shipping details (Google Pay), (2) user defines triggers/constraints (price, seller reputation, return policy), and (3) Google monitors and executes purchases when conditions are met; the transaction is routed via participating merchants and confirmed to the user. Agentic checkout in targeted rollouts (initially U.S.-only for many features).
Safety, privacy, and industry responses
Agentic commerce raises payment, identity, and liability questions. Payments networks and industry stakeholders (e.g., Mastercard) are actively discussing standards for “agentic commerce” — focusing on authentication, never-expose-sensitive-data patterns, and dispute/resolution pathways — because agents acting autonomously change how refunds, fraud detection, and merchant acceptance are handled. This is an active regulatory and standards conversation. From a consumer perspective: only opt in when you trust the account protections (2FA, payment tokens) and read the delegated-permission scope.
If an AI agent offers to buy for you, what should you check?
When Agentic Checkout or “Buy with Google / Gemini buy” appears, read the permissions carefully — Google will require explicit opt-in for automated purchases and stored payment methods (Google Pay). If you want the assistant to monitor price/stock and act automatically, define precise triggers (e.g., “Buy if price drops below $120 and free shipping”).
- Consent & Scope: Confirm exactly what the agent will do (one-time purchase vs. ongoing purchases).
- Payment & Identity: Validate which payment provider is used (AP2 / payment partners) and whether you must re-authenticate.
- Return policy & receipt: Ensure the merchant’s return and tax responsibilities are clear in the interaction.
How do you get your site indexed for Google Shopping and AI surfaces?
Search engines and Google’s Shopping Graph still rely on standard signals. Follow these steps:
Required checklist for discoverability
- Merchant Center & Verification: Create a Merchant Center account, verify your site, and claim website ownership.
- Accurate product markup: Add structured data (schema.org/Product and Offer) to each product page so Google can parse key attributes. Here’s a JSON-LD template:
<script type="application/ld+json">
{
"@context": "https://schema.org/",
"@type": "Product",
"name": "Waterproof Hiking Jacket — Pacific Series",
"image": ["https://example.com/images/SKU-12345-main.jpg"],
"description": "Lightweight insulated waterproof hiking jacket with breathable membrane.",
"sku": "SKU-12345",
"brand": {"@type": "Brand", "name": "TrailCo"},
"offers": {
"@type": "Offer",
"url": "https://example.com/product/SKU-12345",
"priceCurrency": "USD",
"price": "199.00",
"availability": "https://schema.org/InStock"
}
}
</script>
Structured data doesn’t guarantee display in AI Mode, but it increases the likelihood that Google extracts reliable product attributes for the Shopping Graph and generative surfaces.
- High-quality images and 3D assets: Where possible, supply 3D models and AR files (glTF, USDZ) for virtual try-on. Google’s merchant docs explain formats and hosting options.
- Fast, mobile-friendly pages: AI surfaces and visual features favor fast, responsive pages; poor performance undermines visibility.
What Strategic Shifts Must Retailers Make in 2026?
The technical implementation is only half the battle. The shift to agentic commerce requires a fundamental rethink of your marketing strategy.
1. From SEO to AIO (Artificial Intelligence Optimization)
Traditional SEO was about keywords and backlinks. AIO is about Entity Authority. You need to ensure your brand is an entity in Google's Knowledge Graph.
- Action: Audit your "About Us" and "Policy" pages. AI agents read these to verify legitimacy. If your return policy is vague, the agent will perceive your store as "high risk" and avoid purchasing for the user.
2. The "Zero-Click" Reality
Merchants must accept that traffic to their actual websites will decrease. This sounds alarming, but it is offset by higher conversion rates via off-site channels (Agentic Checkout).
- Action: Shift your KPIs. Stop obsessing over "Sessions" and "Pageviews." Start measuring "Share of Recommendation" (how often an AI suggests your product) and "Agent-Originated Revenue."
3. Data Integrity is King
In the past, you could get away with a missing image or a slightly inaccurate description. In 2026, an AI agent comparing attributes will ruthlessly filter out products with incomplete data.
Action: Use the Click Potential score in Merchant Center Next to identify products with data gaps. Ensure every attribute—color, material, pattern, size_system, age_group—is populated.
Conclusion
The launch of the Universal Commerce Protocol and Google's AI shopping features marks the end of the "Search" era and the beginning of the "Agent" era. For consumers, it promises a future where shopping is as easy as conversation. For merchants, it presents a binary choice: adapt your data infrastructure to speak the language of agents, or fade into obscurity.
If you'd like to learn more about AI in Google Shopping, you can check out CometAPI and learn about the Gemini API to avoid being “cheat” by AI. Developers can access Gemini 3 Pro and Gemini 3 Flash through CometAPI, the latest models listed are as of the article’s publication date. To begin, explore the model’s capabilities in the Playground and consult the API guide for detailed instructions. Before accessing, please make sure you have logged in to CometAPI and obtained the API key. CometAPI offer a price far lower than the official price to help you integrate.
Use CometAPI to access chatgpt models, start shopping!
Ready to Go?→ Sign up for gemini API today !
If you want to know more tips, guides and news on AI follow us on VK, X and Discord!