Map Shopify condition metafields to schema so AI recommends your used inventory

Pendium helps e-commerce brands take control over how AI shopping assistants perceive their products, and for resale inventory, that means mapping Shopify metafields directly to schema markup. When a buyer asks Gemini or Claude for refurbished electronics or vintage apparel, the AI relies on the itemCondition structured data property rather than your unstructured product description. Routing your custom Shopify condition metafields into your store's JSON-LD allows AI shopping agents to identify and recommend your open-box, used, or restored inventory to price-conscious shoppers in 2026. Without this code-level translation, AI search engines frequently misclassify second-hand items or exclude them from comparative shopping shortlists entirely.

The default Shopify setup leaves AI guessing

Shopify themes are built with an implicit assumption: every product in your catalog is brand new. When a theme renders standard schema markup, it either omits the itemCondition property entirely or hardcodes it to assume factory-fresh stock. As developer Ilana Davis points out in her analysis of Shopify used product markup, Shopify lacks a native condition setting out of the box, forcing themes to either output incomplete data or guess.

When an AI shopping crawler visits your storefront, it prioritizes structured data over page styling. If that code has no condition field, engines like ChatGPT, Claude, and Google AI Overviews face ambiguity. In many cases, the crawler assumes the product is new and attempts to evaluate your pricing against retail benchmarks. A refurbished laptop listed at $600 looks like an incredible discount on a $1,200 device, but because the condition tag is absent, the model struggles to justify the discrepancy.

Alternatively, the model may cross-reference the SKU or barcode against competing marketplaces like eBay, Back Market, or Amazon Renewed. When third-party platforms provide explicit condition data while your direct store leaves the field blank, AI engines cite those competitor marketplaces instead of your website. You lose the direct sale because the crawler could not verify your product's true status from your own markup.

The problem compounds when you consider buyer intent. As documented in Pendium's guide to AI visibility for DTC brands, AI models give different answers depending on the customer asking the question. A price-conscious shopper explicitly asking for "the best refurbished mirrorless cameras under $800" expects verified second-hand items. If your structured data does not declare the product as refurbished, the AI bypasses your listing in favor of a merchant whose code explicitly states it.

Build a dedicated condition field in Shopify

Writing "Refurbished - Grade A" in your product description text is not enough for machine discovery. Text descriptions are meant for human shoppers reading a page. AI scrapers ingest that text, but they cross-check it against the structured data graph in your HTML header. When you separate product specifications into native custom fields, you give both search scrapers and internal systems a clean data point.

To establish this in your store, use Shopify's custom data architecture. You can review our walkthrough on how to map Shopify size data to schema for AI shopping recommendations to see how structured sizing follows the exact same operational logic.

Follow these steps in your Shopify admin to build the condition field:

• Go to Settings > Custom data > Products.
• Click Add definition to create a structured field.
• Set the namespace and key to custom.condition or use the standard google.condition namespace.
• Select Single line text as the data type.
• Choose List of values or set predefined choices restricted to three values: new, used, and refurbished.

Using a predefined list prevents team members from entering arbitrary phrases like "gently loved" or "open box - excellent." As explained in ShieldKit's guide to fixing condition not declared errors, Google Merchant Center and major retrieval engines demand precise categorical values. Restricting the field options at the admin level prevents syntax errors before they ever reach your theme files.

Once the definition is saved, the condition selector appears at the bottom of every product edit screen in your Shopify admin. You now have a standardized database record for every item in your catalog.

Map the metafield to machine-readable JSON-LD enumerations

Creating the metafield stores the data, but it does not automatically expose that information to search bots. You must bridge the gap between Shopify's database and the public JSON-LD payload in your theme.

The e-commerce data standard defined by schema.org uses specific URIs for product conditions. An AI engine does not look for arbitrary text like "used condition"; it matches against exact schema enumerations.

The four recognized schema conditions

Under schema.org guidelines, the itemCondition property on an Offer type accepts four specific enumeration values. As outlined in Nivk's technical breakdown of resale product schema, these URIs give AI scrapers an absolute reference point:

Shopify Metafield Value	Schema.org Enumeration URI	Meaning
new	https://schema.org/NewCondition	Brand new, unopened original packaging
used	https://schema.org/UsedCondition	Previously owned or opened item showing wear
refurbished	https://schema.org/RefurbishedCondition	Restored to working order by merchant or manufacturer
damaged	https://schema.org/DamagedCondition	Functional or non-functional item with noticeable defects

If you output non-standard values like https://schema.org/PreOwned or simply type "condition": "refurbished", the parser rejects the field. Sticking strictly to these four URIs keeps your store compliant with the standards parsed by ChatGPT, Perplexity, and Google.

Injecting the value into your theme

To output the correct enumeration, open your theme code editor and locate the snippet responsible for structured data. In Dawn and most modern Online Store 2.0 themes, this is usually found in snippets/product-json-ld.liquid or embedded inside sections/main-product.liquid.

Find the section of code that defines the offers array. Inside the Offer object, write a conditional statement that inspects your metafield value and assigns the correct schema URL:

{%- assign schema_condition = 'https://schema.org/NewCondition' -%}

{%- case product.metafields.custom.condition.value -%}
  {%- when 'refurbished' -%}
    {%- assign schema_condition = 'https://schema.org/RefurbishedCondition' -%}
  {%- when 'used' -%}
    {%- assign schema_condition = 'https://schema.org/UsedCondition' -%}
  {%- when 'damaged' -%}
    {%- assign schema_condition = 'https://schema.org/DamagedCondition' -%}
  {%- when 'new' -%}
    {%- assign schema_condition = 'https://schema.org/NewCondition' -%}
{%- endcase -%}

Next, place the variable directly into the JSON object:

{
  "@context": "https://schema.org/",
  "@type": "Product",
  "name": {{ product.title | json }},
  "description": {{ product.description | strip_html | json }},
  "offers": {
    "@type": "Offer",
    "price": "{{ product.selected_or_first_available_variant.price | money_without_currency | remove: ',' }}",
    "priceCurrency": "{{ cart.currency.iso_code }}",
    "availability": "https://schema.org/{% if product.available %}InStock{% else %}OutOfStock{% endif %}",
    "itemCondition": "{{ schema_condition }}"
  }
}

This mapping translates your simple administrative input (refurbished) into the machine-readable web standard (https://schema.org/RefurbishedCondition). When an AI agent reviews the page, the item condition is unambiguous. Tools like the Pendium Agent Experience Engine can then confirm that external bots recognize this status when answering product queries.

Handle mixed-condition catalogs without manual data entry

Updating metadata across dozens or hundreds of products presents an operational challenge. If you run a store with a blend of new arrivals, clearance inventory, and certified pre-owned units, manual data entry invites mistakes.

A single missed product leaves the schema empty, falling back to whatever default the theme carries. You need a reliable method to manage defaults and populate existing items at scale.

Setting a global default

If 90% of your store consists of new items, do not spend hours marking hundreds of items as "new" one by one. Handle this directly inside your Liquid logic by treating empty metafields as new stock.

The Liquid snippet shown above already does this:

{%- assign schema_condition = 'https://schema.org/NewCondition' -%}

By initializing schema_condition with NewCondition, any product that lacks an explicit value in product.metafields.custom.condition outputs as new. You only need to touch the products that deviate from that baseline—your used, open-box, or refurbished units.

Bulk-populating existing inventory

For stores with large catalogs of second-hand goods, updating records one by one through the Shopify admin interface takes too long. As outlined in Branva's analysis of bulk Shopify condition workflows, you can streamline catalog updates using automated batch operations.

Use this operational sequence to populate your condition metafield across large product lists:

• Export your catalog data to CSV using a data tool like Matrixify.
• Include the columns for Handle, Title, Body HTML, and your new Metafield: custom.condition [string].
• Filter your spreadsheet by keywords in the title or description, such as "Refurbished", "Open-Box", "B-Stock", or "Pre-Owned".
• Set the condition column to the matching standardized value (used or refurbished) for those rows.
• Re-import the CSV file to apply the metafield changes across your entire inventory in minutes.

Watch out for direct contradictions between your written copy and your metadata. If your product description opens with "Refurbished Grade B" but your structured data still points to NewCondition because someone forgot to set the metafield, AI models encounter conflicting signals.

When faced with conflicting data on the same page, AI models often drop the listing entirely. Recommending a product with conflicting specs introduces risk of error, and recommendation engines are tuned to avoid surfacing questionable data. Keeping your text description and schema markup aligned protects your listings from being filtered out.

Verify your condition markup across AI shopping engines

Once your theme code is deployed and your products are populated, you must verify that external crawlers receive the new payload. Do not rely solely on how the page looks in a standard web browser.

Start by testing a modified product URL through the official schema.org validator or Google's Rich Results Test tool. Inspect the rendered output under the Offer object to verify that itemCondition appears with the full URI rather than a blank string or local text snippet.

Next, observe how AI assistants index the update. Discovery across platforms like ChatGPT, Perplexity, and DeepSeek does not happen instantaneously; engines must crawl and refresh their context window for your domain.

Pendium monitors how your store appears across seven primary AI search systems: ChatGPT, Claude, Gemini, Grok, Perplexity, DeepSeek, and Google AI Overviews. The platform runs queries simulating varied customer personas, tracking whether your refurbished listings appear when buyers search for discounted, pre-owned, or restored options.

If your catalog contains second-hand, vintage, or reconditioned inventory, you cannot afford to have AI platforms guess what you are selling. Visit Pendium.ai to scan your AI visibility and see how recommendation models classify your product catalog today.