Your store is invisible to AI shopping agents

Shopping Agents Are Already Buying

Perplexity processes over 15 million product-related queries every month. ChatGPT completes purchases through Agentic Commerce Protocol, a system built with Stripe that went live in September 2025. Google's Universal Commerce Protocol, backed by Walmart, Target, Shopify, and Etsy, routes purchase intent through a similar pipeline.

These are not search engines. They are shopping agents. A person says "find me a glucosamine supplement for large breed dogs under $40 with third-party testing" and the agent browses catalogs, compares options, and places an order without the person ever visiting a website.

AI-referred traffic to US retail sites grew 393% year-over-year in the first quarter of 2026, according to Adobe Analytics. That traffic converts 42% better than non-AI traffic. Shoppers arriving from Perplexity spend 57% more per order than those coming from traditional search.

If your store can't talk to these agents, you're losing buyers who already decided to purchase.

How Agentic Commerce Works

ACP and UCP follow the same basic pattern. An AI agent receives a shopping request from a user. The agent queries product catalogs through structured feeds and APIs. It compares options across multiple stores using price, availability, reviews, shipping speed, and ingredient specificity. It presents a recommendation. If the user agrees, the agent completes the purchase through an integrated payment flow.

The entire transaction happens inside the chat interface. The user never opens your website, never sees your homepage, never browses your category pages. The agent does all of that programmatically, reading your product data the way a developer reads an API response.

This means the agent needs machine-readable data. Marketing copy doesn't help. A product description that says "premium quality, trusted by thousands" gives the agent nothing to work with. A description that says "500mg glucosamine HCl per chew, 60 chews per bag, chicken flavor, manufactured in a GMP-certified facility, third-party tested by NSF International" gives the agent everything it needs to make a recommendation.

What Makes a Store Visible

Three things determine whether AI shopping agents can find and recommend your products.

The first is structured product data. Every product needs JSON-LD schema markup with specific fields: name, brand, price, currency, availability, SKU, ingredient list, dosage, weight, dimensions, aggregate ratings, and review count. This is the same schema Google has recommended for years, but AI agents depend on it completely. Without it, your products don't exist in their index.

The second is a real-time catalog feed. AI crawlers need current inventory and pricing. If your feed updates once a day and a product goes out of stock at 10 AM, the agent will recommend it at 3 PM, the customer will try to buy it, and the agent will learn not to trust your catalog. Shopify and WooCommerce both support real-time feeds through their APIs, but most stores haven't configured them for AI consumption.

The third is answers to the questions agents ask on behalf of shoppers. FAQ schema on your product pages gives agents the specific data points they need for comparison. "Is this safe for puppies under 6 months?" "Does this contain shellfish-derived ingredients?" "What's the return policy for opened supplements?" If those answers are structured and on your page, the agent cites you. If they're buried in a PDF or missing entirely, the agent uses a competitor's answers instead.

What Most Stores Get Wrong

Most stores treat structured data as an SEO checkbox. They install a schema plugin, let it auto-generate markup, and never look at the output. The result is technically valid schema that's semantically empty.

A product schema that lists the price and name but omits ingredients, certifications, sourcing, and usage instructions gives an AI agent the same information it could get from any competitor. The brands that win agentic recommendations are the ones whose schema answers the questions shoppers actually ask.

The other common mistake is treating this as a future problem. ACP has been live for ten months. UCP launched with over 20 retail partners. The infrastructure is already running. Stores that implement agentic-ready structured data now are building the trust signals that AI agents will use to rank recommendations for the next several years.

The Concrete Steps

Start with your product catalog feed. Confirm your Shopify or WooCommerce store is syncing inventory and pricing to the major AI crawlers in real time, not on a daily batch.

Add complete JSON-LD product schema to every product page. Include ingredients, certifications, manufacturing details, and usage instructions. Test each page with Google's Rich Results validator.

Write FAQ content for your top 20 products, answering the specific questions customers ask in support tickets and reviews. Add FAQ schema markup so AI agents can parse those answers programmatically.

Then check. Search for your products on Perplexity and ChatGPT. If they don't appear, your structured data has gaps. If they appear but a competitor is recommended instead, compare your schema detail to theirs.