DEV Community

AI True Labs
AI True Labs

Posted on Originally published at geoready.aitruelabs.com

Why ChatGPT Shopping Skips Shopify Product Pages (Fixing Liquid JSON-LD, OAI-SearchBot & llms.txt)

Conversational product discovery (OAI-SearchBot, ChatGPT-User, PerplexityBot) doesn't "browse" a 2.4MB client-rendered storefront the way human shoppers do. Instead, answer engines extract deterministic pricing, stock status, return policies, and brand provenance from three machine-readable layers:

  1. Unblocked Edge Bot Access (robots.txt + zero HTTP 403 JS challenges on OAI-SearchBot)
  2. Complete @graph JSON-LD (Product -> Offer with explicit priceCurrency, numeric price, full https://schema.org/InStock URI, MerchantReturnPolicy, and Organization.sameAs)
  3. Clean Markdown Context (/llms.txt + Accept: text/markdown negotiation so specs aren't truncated inside 18,000+ tokens of theme Liquid/SVG boilerplate)

The 3 Most Common Shopify Theme Failures We See in Live Audits

After running live edge audits across DTC storefronts, three silent bugs account for most invisible SKUs:

1. Fragmented Microdata Instead of Unified JSON-LD @graph

Many legacy or heavily customized Liquid themes emit partial Product JSON-LD from the theme plus conflicting duplicate Offer blocks from review or bundle apps—often omitting priceCurrency on variant arrays or formatting price with currency symbols ("$29.99" instead of "29.99"), which breaks strict schema parsers.

2. Missing FAQPage & Numeric Specification Tables

When a buyer asks ChatGPT "Which stainless steel cat fountain has a <30dB USB pump and ships to the US?", LLMs prioritize pages with explicit <table> spec rows and FAQPage JSON-LD entities over vague lifestyle copy.

3. Missing /llms.txt Storefront Directory

Shopify doesn't natively serve a root /llms.txt summary of your hero collections, shipping SLAs, and return policy—returning a 404 HTML page instead.

Production-Ready Shopify Liquid JSON-LD Snippet (snippets/geoready-product-schema.liquid)

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Organization",
      "@id": "{{ shop.url }}/#organization",
      "name": {{ shop.name | json }},
      "url": {{ shop.url | json }}
    },
    {
      "@type": "Product",
      "@id": "{{ shop.url }}{{ product.url }}#product",
      "name": {{ product.title | json }},
      "description": {{ product.description | strip_html | truncate: 500 | json }},
      "sku": {{ product.selected_or_first_available_variant.sku | json }},
      "brand": { "@id": "{{ shop.url }}/#organization" },
      "offers": {
        "@type": "Offer",
        "url": "{{ shop.url }}{{ product.url }}",
        "priceCurrency": {{ cart.currency.iso_code | json }},
        "price": "{{ product.selected_or_first_available_variant.price | divided_by: 100.0 }}",
        "availability": "https://schema.org/{% if product.available %}InStock{% else %}OutOfStock{% endif %}"
      }
    }
  ]
}
</script>
Enter fullscreen mode Exit fullscreen mode

Run a Free 5-Second Live Edge Check on Your Storefront

You can test whether OAI-SearchBot and PerplexityBot can cleanly parse your store's robots.txt, JSON-LD Offer graph, token-bloat ratio, and /llms.txt using our free edge validator:

Top comments (0)