Google AI Overviews for E-commerce: The Structured Data Checklist

Google AI Overviews are the AI-generated summaries that appear at the top of certain search results pages. For informational queries they have been widely discussed, but for shopping queries they represent a meaningful new surface: Google can now surface specific products, compare options, and link directly to sellers, all inside the search results page before a user clicks anything.

This post is a practical checklist of the structured data signals that influence whether your products appear in AI Overviews. Each item includes a brief explanation of why it matters and, where useful, a JSON-LD snippet you can adapt. We also cover one robots.txt trap that affects a significant number of stores.

What AI Overviews Are and How Products Surface

Google AI Overviews use Google's Gemini models to generate summaries at the top of relevant search results. For product queries, the AI can pull together product information from indexed pages, present comparisons, and cite specific products with direct links to the selling page.

Unlike traditional organic results, where a well-optimized page can earn a position through ranking signals, AI Overviews rely more heavily on the machine-readability of your product data. If Google cannot confidently extract structured, accurate information from your pages, your products are less likely to be included, even if your pages rank well in standard search.

The Google-Extended Nuance Most Stores Miss

Google uses a separate crawler named Google-Extended for its AI features, including AI Overviews and the data that feeds Gemini-based experiences. Blocking Google-Extended in your robots.txt opts your content out of these AI features while leaving normal Googlebot crawling (and therefore your standard SEO rankings) completely unaffected.

The trap: many stores have copied a robots.txt template or installed a plugin that blocks Google-Extended as part of a general AI-training opt-out, without realizing it also removes them from AI Overviews. You can check by visiting yourdomain.com/robots.txt and searching for "Google-Extended".

If you see User-agent: Google-Extended followed by Disallow: /, you are opted out of AI Overviews. Removing or adjusting that rule will allow Google-Extended to crawl your pages and consider your content for AI-generated features.

The Structured Data Checklist

The items below correspond to fields Google uses when evaluating product pages for rich results and AI-generated features. Work through them for your most important product pages first.

1. Product JSON-LD: present and server-rendered

The most fundamental requirement. Your product pages need a valid @type: Product JSON-LD block in the server-rendered HTML. If the markup is injected only by JavaScript after page load, Google's normal crawling pipeline handles that with a two-pass rendering system, but it is slower and less reliable than having it in the initial HTML. Serve it in the HTML response from your server.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Your Product Name",
  "description": "A substantive description of the product.",
  "image": ["https://example.com/product-image.jpg"]
}
</script>

2. Name and description quality

The name field should be clear and specific, matching the visible product title on the page. The description should be substantive: at least a few sentences covering what the product is, who it is for, and key specifications. Thin descriptions that repeat the title or consist of a single marketing sentence give the AI little material to work with.

3. GTIN and MPN identifiers

Global Trade Item Numbers (GTIN, covering EAN barcodes and UPCs) and Manufacturer Part Numbers (MPN) uniquely identify a product in a way that the AI can cross-reference against product knowledge graphs. Google has specifically recommended including these for products that have them.

"gtin": "0123456789012",
"mpn": "MFR-PART-1234"

4. Offer with price, priceCurrency, and availability

Every product page should have an Offer block. Price and currency are required. Availability is critical: using the correct schema.org URL makes your listing machine-readable as in-stock or out-of-stock.

"offers": {
  "@type": "Offer",
  "price": "49.99",
  "priceCurrency": "USD",
  "availability": "https://schema.org/InStock",
  "url": "https://example.com/product"
}

5. priceValidUntil

Adding priceValidUntil signals to Google that your price data is actively maintained. A date in the future (refreshed periodically) tells the crawler your pricing information is current. Stale or missing price validity reduces confidence in your listing.

"priceValidUntil": "2026-12-31"

6. shippingDetails with OfferShippingDetails

Google can surface shipping information in rich results and AI Overviews for shopping queries. OfferShippingDetails lets you specify shipping cost, delivery time, and destination. Free shipping or fast delivery are signals that influence product recommendations.

"shippingDetails": {
  "@type": "OfferShippingDetails",
  "shippingRate": {
    "@type": "MonetaryAmount",
    "value": "0",
    "currency": "USD"
  },
  "deliveryTime": {
    "@type": "ShippingDeliveryTime",
    "handlingTime": {
      "@type": "QuantitativeValue",
      "minValue": 0,
      "maxValue": 1,
      "unitCode": "DAY"
    },
    "transitTime": {
      "@type": "QuantitativeValue",
      "minValue": 3,
      "maxValue": 5,
      "unitCode": "DAY"
    }
  },
  "shippingDestination": {
    "@type": "DefinedRegion",
    "addressCountry": "US"
  }
}

7. hasMerchantReturnPolicy

Return policy markup tells shoppers and AI systems what happens if a purchase does not work out. Google has explicitly listed this as a recommended field for product pages targeting rich results.

"hasMerchantReturnPolicy": {
  "@type": "MerchantReturnPolicy",
  "applicableCountry": "US",
  "returnPolicyCategory":
    "https://schema.org/MerchantReturnFiniteReturnWindow",
  "merchantReturnDays": 30,
  "returnMethod": "https://schema.org/ReturnByMail",
  "returnFees": "https://schema.org/FreeReturn"
}

8. aggregateRating and Review

If your product page displays star ratings, mark them up. AggregateRating provides a summarized rating (the average and review count). Individual Review items provide review text. Only mark up reviews that are actually visible on the page. Do not fabricate or aggregate ratings from third-party platforms without confirmation that you have permission and the values match what is displayed.

"aggregateRating": {
  "@type": "AggregateRating",
  "ratingValue": "4.7",
  "reviewCount": "84"
}

9. Images with ImageObject

Provide multiple product images and mark them up with ImageObject. Include at least a main product shot and a few lifestyle or detail images. Google uses image data in AI-generated product summaries. Images must be crawlable (served as static URLs, not CSS background images).

"image": [
  {
    "@type": "ImageObject",
    "url": "https://example.com/product-front.jpg",
    "width": 1200,
    "height": 1200
  },
  {
    "@type": "ImageObject",
    "url": "https://example.com/product-side.jpg",
    "width": 1200,
    "height": 1200
  }
]

10. BreadcrumbList for category context

A BreadcrumbList schema on product pages tells Google where the product sits in your catalog hierarchy (e.g., Home, Running Shoes, Men's). This category context helps the AI understand the product's domain and surface it for category-level shopping queries.

{
  "@context": "https://schema.org",
  "@type": "BreadcrumbList",
  "itemListElement": [
    {
      "@type": "ListItem",
      "position": 1,
      "name": "Home",
      "item": "https://example.com"
    },
    {
      "@type": "ListItem",
      "position": 2,
      "name": "Running Shoes",
      "item": "https://example.com/running-shoes"
    },
    {
      "@type": "ListItem",
      "position": 3,
      "name": "Men's Velocity Runner",
      "item": "https://example.com/running-shoes/velocity-runner-mens"
    }
  ]
}

11. dateModified for freshness

A dateModified field on the page (or in the Product JSON-LD) signals that the content is actively maintained. Product pages that appear stale or neglected are lower-confidence sources for AI-generated answers. Update this field whenever you make meaningful changes to product content or pricing.

12. Canonical URL

Every product page should have a <link rel="canonical"> tag pointing to itself (or to the preferred URL if you have variant URLs). Duplicate or parameter-laden URLs without canonicals confuse crawlers and dilute the signal strength of your structured data.

13. FAQPage markup for product questions

If your product page has a Q&A or FAQ section (common on detail-heavy product pages and DTC brands), mark it up with FAQPage JSON-LD. This can surface directly in search results and provides the AI with explicit question-answer pairs about the product.

14. Brand as an entity with sameAs

Mark up the product's brand as an Organization or Brand entity and include a sameAs URL pointing to the brand's Wikipedia article, Wikidata entry, or official brand page. This anchors the brand to a known entity in Google's knowledge graph, which strengthens the product's associative signals.

"brand": {
  "@type": "Brand",
  "name": "Your Brand Name",
  "sameAs": "https://en.wikipedia.org/wiki/Your_Brand"
}

How to Verify Your Structured Data

After implementing changes, validate with two tools:

• Google Rich Results Test (search.google.com/test/rich-results): paste your product URL and confirm Google can parse your Product markup without errors or warnings.
• Krytho free scan at krytho.com/scan: checks the live page for structured data completeness, crawlability, and the specific signals AI shopping engines look for. Useful for a quick pass across multiple product URLs.

Validation catches syntax errors and missing required fields. It does not guarantee AI Overview inclusion, but it removes the technical barriers that would prevent inclusion.

A Note on Realistic Expectations

Structured data is one input into a complex system. Google AI Overviews also consider page quality, E-E-A-T signals, the competitiveness of your category, and query-specific factors that are not fully documented. The checklist above gives your pages the best possible technical foundation. From there, content quality and domain authority do the rest of the work.

These signals are also not static. Google updates its structured data requirements and recommendations periodically. Check the Google Search Central documentation for Product structured data and re-audit your pages after significant platform updates.