How to structure Shopify Markets Pro duty schemas for AI recommendations

When an international buyer asks ChatGPT or Claude for a product recommendation, the AI optimizes for a frictionless user experience—meaning it won't suggest your Shopify store if it can't confidently parse the landed cost, including duties and import taxes. To fix this visibility gap, merchants using Shopify Markets Pro must explicitly map their HTS codes and Delivered Duty Paid (DDP) pricing rules into their site's JSON-LD. By structuring this data exactly how large language models expect to read it, Pendium users can ensure AI assistants display accurate, localized prices and recommend their products across borders without hesitation.

Where cross-border AI recommendations break down

AI visibility platform crawlers and LLMs operate as advanced decision engines rather than traditional text search indexes. They do not just scrape keyword occurrences. Instead, they calculate whether a product is a viable, low-friction purchase for a user's specific location. When a shopper in the UK queries Claude for a highly specific outdoor gear recommendation, the model acts as an automated customs gatekeeper. If your store hides its international delivery capability behind a dynamic checkout script, the LLM assumes your product will trigger surprise fees at the border or face long transit delays.

To maintain organic reach in modern conversational search, you must expose your international shipping, tax, and return frameworks directly in your raw HTML. LLMs prioritize stores that provide explicit guarantees on import costs and regional returns. If your store lacks clear international return structures, the AI will bypass you for a localized competitor. For details on how to format these policies, see our guide on fixing missing Shopify return policy schema for AI search.

The fundamental technical difference between how search bots read basic international configurations versus comprehensive setups becomes apparent during automated crawls. The table below details what AI crawlers encounter across different setups:

Without structured data, AI agents cannot verify these compliance structures. The crawler defaults to the safest option, which is to omit your brand from its final response.

Mapping tariff codes directly into Shopify schema

The foundation of any cross-border recommendation is the Harmonized Tariff Schedule (HTS) code. When a search engine like Perplexity or Google AI Overviews parses a product page, it reads the HTS code to classify the physical item. This classification tells the model which customs rules, import restrictions, and duty rates apply to the shipment. If this data is missing, the LLM must guess the product category, which increases the margin of error and causes the agent to exclude your product from localized recommendations.

The requirement for 6-digit and 10-digit precision

Historically, search crawlers ignored HS codes because traditional search engines only indexed consumer-facing text. However, modern LLMs are trained on vast datasets of international trade records. A study on ATLAS: Benchmarking and Adapting LLMs for Global Trade shows that fine-tuned language models can process 6-digit and 10-digit customs classifications with high accuracy. This means AI models understand the difference between a synthetic leather boot and a real leather shoe based entirely on the HTS string.

The European Union's Combined Nomenclature contains roughly 29,000 nodes, as documented in open-source classification projects like jfuglas/TullTaxa. To navigate this hierarchy, AI engines require exact schema markup rather than generic product copy. Providing only a 6-digit code is the bare minimum. Injecting a full 10-digit code into your regional schema variant allows the AI to determine precise duty structures and confidently recommend your item to EU buyers.

Updating product metafields for schema injection

Shopify stores the default Harmonized System code under the standard physical_customs_information object, but this field is rarely exposed to the storefront theme or the default JSON-LD payload. To make this data readable to AI crawlers, you must map the code to a custom product metafield and print it to your theme's liquid files.

You can create a single-line text metafield named custom.hts_code and populate it for your international SKUs. Once populated, use a liquid snippet to inject this code directly into your product schema under the mpn (Manufacturer Part Number) or custom extension fields. This simple change allows the Pendium AI visibility platform to verify that search agents are indexing your exact product classification.

{%- if product.metafields.custom.hts_code -%}
  "mpn": "{{ product.metafields.custom.hts_code }}",
{%- endif -%}

Using liquid to surface the code ensures that the data is baked into the initial server-side render. This is critical because many AI search crawlers do not execute heavy client-side JavaScript when gathering structured data.

Configuring JSON-LD for Delivered Duty Paid (DDP)

International buyers hate surprise fees. AI agents, acting as user advocates, are programmed to avoid recommending stores that ship Delivered Duty Unpaid (DDU). When configuring Shopify Markets Pro, the platform calculates duties and taxes at checkout. However, if these calculations are not written directly into your schema, the AI crawler has no way of knowing that your storefront offers a Delivered Duty Paid (DDP) experience.

Isolating base price from import fees

To feed AI agents the exact numbers they need, your structured data must split the product's base price from any localized taxes and import duties. When a crawler hits your localized subfolder, it should find a clean breakdown of these components. This prevents the AI from misinterpreting a duty-inclusive price as a high base price, which would otherwise penalize your store in budget-focused search queries.

You can resolve this mispricing risk by using the priceSpecification property inside your schema.org/Offer markup. This property lets you define the core price, the value-added tax (VAT), and any duty surcharges as distinct nodes. For a step-by-step walkthrough of this split, read our article on how to fix Shopify tax-inclusive JSON-LD so AI displays actual prices.

Validating schema.org/Offer markup for regional variants

Each market in your Shopify Markets Pro setup must render its own unique Offer node. If you sell to the UK, Germany, and the US from a single store, your product page must output three distinct offer blocks when parsed by an AI bot. Below is an example of how to structure this JSON-LD to declare DDP terms and HTS codes for a German buyer:

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Arcturus Waterproof Trail Backpack",
  "description": "Professional 40L hiking backpack with roll-top closure.",
  "category": "Sporting Goods > Outdoor Recreation > Hiking",
  "mpn": "ARC-40-BK",
  "offers": {
    "@type": "Offer",
    "priceCurrency": "EUR",
    "price": "145.00",
    "itemCondition": "https://schema.org/NewCondition",
    "availability": "https://schema.org/InStock",
    "eligibleRegion": {
      "@type": "GeoShape",
      "box": "DE"
    },
    "priceSpecification": [
      {
        "@type": "UnitPriceSpecification",
        "priceType": "https://schema.org/BasePrice",
        "price": "121.85",
        "priceCurrency": "EUR"
      },
      {
        "@type": "UnitPriceSpecification",
        "priceType": "https://schema.org/VAT",
        "price": "23.15",
        "priceCurrency": "EUR",
        "valueAddedTaxIncluded": true
      }
    ],
    "shippingDetails": {
      "@type": "OfferShippingDetails",
      "shippingDestination": {
        "@type": "DefinedRegion",
        "addressCountry": "DE"
      },
      "deliveryTime": {
        "@type": "ShippingDeliveryTime",
        "handlingTime": {
          "@type": "QuantitativeValue",
          "minValue": 1,
          "maxValue": 2,
          "unitCode": "DAY"
        },
        "transitTime": {
          "@type": "QuantitativeValue",
          "minValue": 3,
          "maxValue": 5,
          "unitCode": "DAY"
        }
      }
    }
  }
}

This explicit structure tells the AI crawler exactly how the final price is calculated. It eliminates the ambiguity that often causes search engines to exclude international listings.

Auditing international pages for AI readability

Implementing the schema is only the first step. You must verify that search engine crawlers can access and interpret your newly formatted data. Many Shopify merchants configure Shopify Markets Pro correctly but block international user agents in their security settings or serve incomplete JSON-LD on localized subfolders.

To ensure your international catalogs are indexable, you must analyze your site's structure from the perspective of an AI bot.

• Check subfolder rendering: Ensure that language-specific paths (like /en-ca or /de-de) output the correct localized schema. If a crawler requests a German page but receives US dollar schema, the AI will index the wrong pricing.
• Verify hreflang tags: AI search agents use hreflang tags to map relationships between regional page variants. If these are broken, the AI cannot resolve which page to show the user.
• Audit crawler accessibility: Some Shopify apps and firewall rules block or throttle non-standard search bots. Ensure your robots.txt allows access to agents from OpenAI, Anthropic, and Google.

Pendium provides a specialized tool to run these technical checks. Using the Pendium AI Site Audit tool, you can scan your Shopify domain to identify broken structured data, incorrect heading hierarchies, and missing regional variants that block AI discovery.

By running these audits, you can find and fix gaps before they impact your international conversion rates. Ensuring your Shopify Markets Pro setup is fully optimized for AI visibility means your products will stay on the digital shortlists that define modern e-commerce. Run a free Pendium AI Visibility Scan on your Shopify domain to see exactly how ChatGPT, Claude, and Gemini perceive your cross-border pricing and product data.