Shopify Food & Beverage Nutrition Schema for AI Shopping Agents

Why Food Brands Are Invisible to AI Dietary Queries

The food and beverage category is one of the most AI-agent-driven verticals in e-commerce. Users routinely ask AI shopping agents questions that are structurally nutritional: "low-sodium crackers under 140mg per serving," "high-protein vegan snacks with at least 20g protein," "tree nut-free granola bars," "Certified Vegan protein powder." These queries require structured, machine-readable nutritional data — not text descriptions or product imagery.

Shopify's default product JSON-LD outputs only name, description, price, and availability. Nutrition facts — even when displayed beautifully as an FDA-compliant label image or an app-generated HTML table — are completely invisible to AI agent crawlers. The nutrition data exists on the page for humans; it is not in the structured data for machines.

The schema.org NutritionInformation type and the RestrictedDiet enumeration exist precisely to solve this problem. Adding them to your Shopify product JSON-LD lets AI agents accurately filter, rank, and surface your products in response to dietary-intent queries that represent a substantial and fast-growing share of food e-commerce traffic.

Food product query types requiring structured data

Shopify default vs. nutrition-aware JSON-LD

{
  "@type": "Product",
  "name": "Dark Chocolate Protein Bar",
  "description": "20g protein. Gluten free.
    Contains: Milk, Soy.",
  // nutrition in text only
  // allergens in description text
  // invisible to AI dietary filters
}

{
  "@type": "Product",
  "nutrition": {
    "@type": "NutritionInformation",
    "calories": "210 calories",
    "servingSize": "1 bar (60g)",
    "proteinContent": "20 g"
  },
  "additionalProperty": [{
    "propertyID": "containsAllergen",
    "value": "Milk"
  }]
}

Food Product JSON-LD Patterns

Minimum viable NutritionInformation

The baseline pattern: add a nutrition property to your Product with a NutritionInformation type. At minimum, include calories and servingSize — these two fields unlock the broadest set of AI dietary queries:

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Dark Chocolate Almond Protein Bar",
  "description": "20g protein per bar. Made with real almonds and 72% dark chocolate.",
  "brand": { "@type": "Brand", "name": "NutriCraft" },
  "nutrition": {
    "@type": "NutritionInformation",
    "servingSize": "1 bar (60g)",
    "calories": "210 calories",
    "fatContent": "9 g",
    "saturatedFatContent": "3 g",
    "carbohydrateContent": "18 g",
    "sugarContent": "6 g",
    "fiberContent": "4 g",
    "proteinContent": "20 g",
    "sodiumContent": "180 mg"
  },
  "offers": {
    "@type": "Offer",
    "price": "32.00",
    "priceCurrency": "USD",
    "availability": "https://schema.org/InStock"
  }
}

Allergen declaration with containsAllergen

Declare each allergen as a separate additionalProperty entry with propertyID: "containsAllergen". Use standard allergen names matching FDA Big-9 and EU Big-14 allergen lists. Add mayContainAllergen as a distinct propertyID for cross-contact risk so AI agents can distinguish between intentional ingredients and manufacturing cross-contamination:

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Dark Chocolate Almond Protein Bar",
  "nutrition": {
    "@type": "NutritionInformation",
    "servingSize": "1 bar (60g)",
    "calories": "210 calories",
    "proteinContent": "20 g"
  },
  "additionalProperty": [
    {
      "@type": "PropertyValue",
      "propertyID": "containsAllergen",
      "name": "Contains allergen",
      "value": "Milk"
    },
    {
      "@type": "PropertyValue",
      "propertyID": "containsAllergen",
      "name": "Contains allergen",
      "value": "Tree nuts"
    },
    {
      "@type": "PropertyValue",
      "propertyID": "containsAllergen",
      "name": "Contains allergen",
      "value": "Soy"
    },
    {
      "@type": "PropertyValue",
      "propertyID": "mayContainAllergen",
      "name": "May contain allergen (cross-contact)",
      "value": "Peanuts"
    }
  ]
}

Dietary certifications — Organic, Vegan, Kosher, Non-GMO

Use hasCertification for third-party verified certifications with an issuing body. For self-declared dietary designations (no certifying body), use additionalProperty: dietaryRestriction with schema.org RestrictedDiet enumeration values:

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Single-Origin Ethiopian Espresso Beans",
  "hasCertification": [
    {
      "@type": "Certification",
      "name": "USDA Organic",
      "certificationStatus": "https://schema.org/CertificationActive",
      "issuedBy": {
        "@type": "Organization",
        "name": "USDA Agricultural Marketing Service",
        "url": "https://www.ams.usda.gov"
      }
    },
    {
      "@type": "Certification",
      "name": "Fair Trade Certified",
      "certificationStatus": "https://schema.org/CertificationActive",
      "issuedBy": {
        "@type": "Organization",
        "name": "Fair Trade USA",
        "url": "https://www.fairtradecertified.org"
      }
    }
  ],
  "additionalProperty": [
    {
      "@type": "PropertyValue",
      "propertyID": "dietaryRestriction",
      "name": "Dietary restriction",
      "value": "https://schema.org/VeganDiet"
    },
    {
      "@type": "PropertyValue",
      "propertyID": "dietaryRestriction",
      "name": "Dietary restriction",
      "value": "https://schema.org/GlutenFreeDiet"
    }
  ]
}

Full food product pattern — nutrition + allergens + certifications

Production-ready pattern combining all three signal types. This is the complete JSON-LD for a food product surfaceable across calorie, allergen, macro, and certification queries:

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Pea Protein Powder — Vanilla Bean",
  "brand": { "@type": "Brand", "name": "PlantForce" },
  "nutrition": {
    "@type": "NutritionInformation",
    "servingSize": "1 scoop (30g)",
    "calories": "120 calories",
    "fatContent": "2 g",
    "saturatedFatContent": "0 g",
    "carbohydrateContent": "3 g",
    "sugarContent": "1 g",
    "fiberContent": "1 g",
    "proteinContent": "25 g",
    "sodiumContent": "340 mg"
  },
  "additionalProperty": [
    {
      "@type": "PropertyValue",
      "propertyID": "dietaryRestriction",
      "value": "https://schema.org/VeganDiet"
    },
    {
      "@type": "PropertyValue",
      "propertyID": "dietaryRestriction",
      "value": "https://schema.org/GlutenFreeDiet"
    }
  ],
  "hasCertification": {
    "@type": "Certification",
    "name": "Certified Vegan",
    "certificationStatus": "https://schema.org/CertificationActive",
    "issuedBy": {
      "@type": "Organization",
      "name": "Vegan Action",
      "url": "https://vegan.org"
    }
  },
  "offers": {
    "@type": "Offer",
    "price": "54.00",
    "priceCurrency": "USD",
    "availability": "https://schema.org/InStock"
  }
}

Shopify Liquid implementation with metafields

Store nutrition data in a nutrition metafield namespace. Gate on a product tag (food-product) to avoid injecting nutrition schema on non-food products in a mixed catalog:

{% if product.tags contains 'food-product' %}
{% assign n = product.metafields.nutrition %}
{% if n.calories %}

{% endif %}
{% endif %}

Nutrition property reference

Common Mistakes

1. Putting nutrition facts only in the product description or image

An FDA-compliant nutrition label image, a formatted HTML nutrition table, or a description line like "190 calories per serving" — all invisible to AI agent crawlers. Text and images are not machine-readable for structured filtering. AI agents answering "protein bars under 200 calories" read the NutritionInformation.calories property in JSON-LD, not description text. The nutrition panel on your page can remain for human readers; the JSON-LD serves AI agents.

2. Listing allergens only as a sentence in the ingredients text

Writing "Contains: Milk, Soy, Almonds" in product description or ingredients text is not machine-readable for allergen exclusion queries. AI agents filtering "nut-free protein bars" look at the containsAllergen additionalProperty — not description text. Allergen sentence in description plus structured containsAllergen declaration is the correct pattern: keep both for legal compliance and AI readability.

3. Conflating "contains" with "may contain" allergens

Using containsAllergen for both intentional allergen ingredients and cross-contact manufacturing risk overstates the actual allergen content. For users with severe allergies filtering on "completely peanut-free," a product that says containsAllergen: Peanuts but is actually peanut-free with a "may contain" warning will be excluded unnecessarily. Use containsAllergen for intentional ingredients only; use a separate mayContainAllergen additionalProperty for cross-contact risk.

4. Using self-declared organic without a certification entity

Declaring hasCertification: { name: "Organic" } without an issuedBy Organization leaves the certification claim unverifiable. AI agents may discount or ignore uncertified organic claims. For USDA Organic certification, include issuedBy: { name: "USDA Agricultural Marketing Service" } to tie the claim to a recognized authority. For uncertified products made with organic ingredients, use additionalProperty: organicIngredients rather than a certification claim.

5. Injecting nutrition schema on non-food products in a mixed catalog

Shopify stores selling both food and non-food products (e.g., a wellness brand selling supplements, yoga mats, and apparel together) must gate nutrition JSON-LD on a tag. Injecting NutritionInformation on a yoga mat or water bottle creates invalid structured data and can cause schema validation errors that degrade all structured data on that page. Use a product tag like food-product or edible to restrict injection.

Implementation Checklist

• Add nutrition: { "@type": "NutritionInformation" } with at minimum calories and servingSize
• Include all available macros: proteinContent, carbohydrateContent, fatContent, sugarContent, sodiumContent
• Declare each intentional allergen as a separate additionalProperty: containsAllergen entry
• Declare cross-contact allergens separately as additionalProperty: mayContainAllergen
• Use hasCertification with issuedBy Organization for certified Organic, Vegan, Kosher, Fair Trade, Non-GMO claims
• Use additionalProperty: dietaryRestriction with schema.org RestrictedDiet values for self-declared dietary designations
• Store all nutrition data in a nutrition.* metafield namespace for Liquid injection
• Gate nutrition JSON-LD on a product tag (e.g., food-product) in mixed catalogs
• Run CatalogScan to verify NutritionInformation and allergen properties appear in your AI readiness report

Related Resources

• Age Verification and Restricted Product Schema
• Product Safety Certifications Schema
• Material Composition Schema (REACH/TSCA)
• Product Condition Structured Data
• CatalogScan Signal Reference