Organizing Shopify product metafields for reliable Algolia-powered furniture search

What was happening

The Shopify catalog was connected to Algolia for product search. The search layer depended on product data coming from Shopify, including custom product information stored in metafields. Without a deliberate metafield structure, similar attributes could be entered inconsistently, become difficult to index, or require one-off search rules that were hard to maintain as the furniture catalog grew.

Why the obvious solution was not enough

Algolia can only search and filter the data it receives. The solution therefore could not be treated as a front-end search problem alone. The source catalog needed a stable field model in Shopify so products could be indexed predictably and future merchandising changes would not require rebuilding the search implementation each time.

How the solution works

1. Audit which product attributes need to participate in search, filtering, merchandising, or display before creating new metafields.
2. Group related product metadata under a consistent Shopify metafield namespace and use clear keys that describe the business attribute rather than the current visual treatment.
3. Choose metafield types deliberately so comparable values are stored in comparable formats instead of mixing free text, numbers, references, and lists unpredictably.
4. Keep the product record in Shopify as the source of truth and expose only the metafields Algolia actually needs for searchable or filterable product records.
5. Normalize repeated values and naming conventions before indexing. Search quality degrades quickly when the same concept is represented by multiple spellings or inconsistent formats.
6. Map the selected Shopify fields and metafields into the Algolia product record so search and filters use structured attributes rather than parsing long product descriptions.
7. Test new and edited products to confirm metafield changes are reflected in the Algolia index and that missing optional values do not break the product record.
8. Document the metafield organization so future catalog editors know which field controls which part of search and can add products without creating new data conventions.

const searchRecord = {
  objectID: product.id,
  title: product.title,
  handle: product.handle,
  productType: product.productType,
  attributes: {
    ...product.metafields.search,
  },
};

indexProductInAlgolia(searchRecord);

What should be verified before shipping

• Compare several products in the same furniture category and confirm equivalent attributes use the same metafield keys and value formats.
• Edit a Shopify metafield value and verify the corresponding Algolia record updates after the normal synchronization path runs.
• Test products with optional metafields left empty so the index remains valid and the storefront search does not render broken filters.
• Search and filter using the structured attributes and confirm results match the Shopify catalog data rather than stale or manually duplicated values.
• Add a new product using the documented metafield structure and confirm it becomes searchable without adding one-off code for that product.
• Review which fields are actually indexed so unnecessary internal or presentation-only data is not added to the search record.

What changed

The search implementation had a cleaner contract between Shopify and Algolia: Shopify remained the source of product truth, metafields provided structured catalog attributes, and Algolia consumed a predictable record for customer-facing discovery.

Reusable lesson

Search quality starts with catalog architecture. Before tuning ranking or UI, make the source product data consistent enough that the search engine receives dependable attributes to work with.