Shopify brand entity in JSON-LD

The brand property on your Product JSON-LD is how AI shopping agents decide which store is the "official" seller of a branded product versus a marketplace reseller. The shape matters: a nested {"@type": "Brand", "name": "..."} entity gets full credit; a plain string gets half; a missing brand gets zero. Most Shopify themes emit the string form by default, leaving 4 of 8 possible points on the table for every PDP — and a missing or weak brand entity is the difference between "the official Allbirds.com store" winning a "where to buy Allbirds" query versus a discount reseller showing up first.

Last updated 2026-04-30 · Ranking-spread signal · 8 pts

What it is

The schema.org Product type allows two shapes for the brand property:

{
  "@type": "Product",
  "name": "Wool Runner",
  "offers": {...}
  // no "brand"
}

{
  "@type": "Product",
  "name": "Wool Runner",
  "brand": "Allbirds",
  "offers": {...}
}

{
  "@type": "Product",
  "name": "Wool Runner",
  "brand": {
    "@type": "Brand",
    "name": "Allbirds"
  },
  "offers": {...}
}

The nested form is what schema.org documents recommend and what every major AI agent's parser is built to expect. The string form was historically permitted as shorthand and is being phased out — it still parses, but agents are increasingly weighting it as a "weaker signal" relative to the nested form.

Why AI shopping agents care

• Official-seller resolution. When ChatGPT is asked "where to buy Allbirds Wool Runners," it has to decide which store is the canonical/manufacturer source vs. a reseller. The store whose JSON-LD identifies itself as the Allbirds brand wins over third-party listings that mention Allbirds in passing.
• Reseller disambiguation. If you're a multi-brand retailer (REI, Backcountry, Zappos), the brand-entity-per-PDP tells agents "this product is by Patagonia, sold by us." Without it, agents have to infer from page text — slow, error-prone, and they often default to the marketplace not surfacing you.
• Knowledge-graph linkage. The nested Brand entity can carry additional fields (url, logo, sameAs linking to Wikipedia/Wikidata) that let agents resolve to a known brand entity in their training data. A known-entity match lifts trust and ranking.
• Cross-product brand consistency. If brand is encoded the same way across all your PDPs, agents learn your full catalog as one brand. Inconsistent brand strings (sometimes "Allbirds," sometimes "AllBirds," sometimes empty) fragment the brand authority across your own site.

How to test it on your store

1. Open any PDP. View source.
2. Find the Product JSON-LD block (search for "@type": "Product").
3. Look for the "brand" field. If it's missing — 0 credit. If it's a plain string ("brand": "Allbirds") — half credit. If it's a nested object with "@type": "Brand" — full credit.
4. Spot-check 3 PDPs across different collections to confirm the shape is consistent — themes sometimes emit one shape on default products and a different shape on app-managed products.

The free CatalogScan scan runs this check automatically and reports the exact shape it found, so you don't have to view-source ten different PDPs by hand.

How to fix it

"brand": {
  "@type": "Brand",
  "name": {{ product.vendor | json }}
}

"brand": {
  "@type": "Brand",
  "name": "Allbirds",
  "url": "https://www.allbirds.com/",
  "logo": "https://www.allbirds.com/assets/logo.png",
  "sameAs": [
    "https://en.wikipedia.org/wiki/Allbirds",
    "https://www.wikidata.org/wiki/Q47073143"
  ]
}

4 mistakes we keep finding