The structured data your AI competitors already have

The Gap Between Cited and Invisible

When Perplexity recommends a joint supplement for dogs, it pulls product details from sites that have clean, specific structured data. It cites the product name, price, rating, ingredient list, and dosage from the page's JSON-LD markup. When it skips your product, it's usually because your structured data is either missing, incomplete, or too vague to be useful.

The gap between being cited by AI and being invisible to AI is not about content quality or domain authority. It's about whether your product pages contain machine-readable data that answers the specific questions AI agents need to resolve before making a recommendation.

Most commerce sites have some structured data. Very few have structured data that's detailed enough to compete for AI citations.

The Four Schema Types That Matter

Product schema is the foundation. Every product page needs JSON-LD markup with the product name, brand, price, currency, availability, SKU, description, aggregate rating, and review count. That much is standard. What separates cited products from ignored ones is the additional detail: ingredient lists, dosage instructions, certifications, manufacturing location, weight, dimensions, and material composition. AI agents compare products on these specific attributes. If your schema doesn't include them, the agent uses a competitor's product that does.

FAQ schema gives AI agents the answers to questions shoppers ask. "Is this safe for puppies under 6 months?" "Does this interact with blood thinners?" "What's the recommended serving for a 50-pound dog?" These are the exact queries that drive product recommendations in conversational AI. When your FAQ schema contains specific answers, the AI can cite your page directly. When it doesn't, the AI finds those answers elsewhere.

Review schema provides social proof in a format AI agents can parse and present. Aggregate ratings, review count, and individual review snippets help agents rank products. A product with 847 reviews averaging 4.6 stars gets recommended over an unreviewed product with identical specs, because the agent treats reviews as a trust signal.

How-to schema matters for products with usage complexity. Supplements with specific dosing protocols, skincare products with multi-step routines, and educational products with enrollment processes all benefit from structured how-to content. AI agents surface this information when shoppers ask "how do I use this" or "what's the process for getting started."

How to Audit What You Have

Go to Google's Rich Results Test and enter your product page URLs one at a time. The tool shows exactly which schema types are present and flags missing required fields. Run your top 20 product pages through it.

Most Shopify stores have basic product schema from their theme or an SEO plugin. The output usually looks complete at a glance. Look closer. Check whether the ingredient list is actually populated or just contains the same generic description as the product page body. Check whether FAQ schema exists at all. Check whether review schema includes individual review content or just an aggregate number.

The most common finding is technically valid schema with semantically empty fields. A product description that reads "premium quality supplement for dogs" in the schema is technically correct markup, but it gives an AI agent nothing to differentiate your product from any other "premium quality supplement for dogs" in its index.

Fixing It on Shopify

If you're running a standard Shopify theme, your product schema comes from the theme's JSON-LD template in the product.liquid file. You can edit this directly or use a structured data app. Either way, the fix is the same: add the specific fields that are missing.

For ingredients, create a metafield for each product that contains the full ingredient list. Reference that metafield in your JSON-LD template. Do the same for certifications, manufacturing details, and dosage instructions. This is manual work for the first batch of products, but once the template references the metafields, every new product gets complete schema as long as the metafields are filled in.

For FAQ schema, add a dedicated FAQ section to each product page with 5-8 questions that customers actually ask. Pull these from your support inbox, reviews, and the "Questions" section on Amazon listings for similar products. Structure each Q&A pair as FAQPage schema in a separate JSON-LD block on the page.

If you're running a headless frontend, you have more control. Your product data comes through the Storefront API or a custom backend. Build your JSON-LD generation into your frontend rendering layer so every product page automatically includes complete schema from your product data model.

What to Do This Week

Pick your ten highest-revenue products. Run each through Google's Rich Results Test. Note which schema types are present and which fields are populated with specific data versus generic copy.

For each product, write 5 FAQ entries using real customer questions from support tickets and reviews. Add those to the product pages with proper FAQPage schema.

Fill in the missing product schema fields: ingredients, certifications, manufacturing details, and usage instructions. Every field you add is a data point AI agents can use to recommend your product over a competitor's.

Test again after making changes. Then search for your products on Perplexity and ChatGPT to see whether they appear in recommendations. If they do, check whether the cited details are accurate and specific. If they don't, your competitors' structured data is still more complete than yours.