Metafields vs. tags - a Shopify taxonomy guide

It’s a random Monday. You’ve got 300 new products to upload, a supplier spreadsheet that doesn't match your store, and a looming sense of dread. You start tagging.

“Leather”... or was it “leather”? Maybe “material_leather”?

You scroll through the tag list and realise all three already exist -and no one knows which one is actually driving your site filters. Multiply that by color, season, and fit, and suddenly product uploads feel slow, error-prone, and stressful.

This is the product data problem we see with many growing stores. For years, tags were the only tool available, so they ended up doing everything. But it doesn't have to be this way. Today, we recommend a clear separation: metafields for the permanent foundation, and tags for agile merchandising.

1. Metafields as a permanent foundation

Think of metafields as the fixed attributes of a product. Because this data is structured, it is the only reliable way to manage a site as it grows.

• What they are: Material, shape, dimensions, flavor, or core use-cases.

• Why they matter: Metafields power your automated collections, SEO schema, and AI discovery feeds.

• The audit factor: It is much easier to see if a value is missing in a metafield column than it is to hunt for a missing tag in a flat list of 50 others.

2. Category metafields

As of 2026, Shopify has standardised this with the Standard Product Taxonomy. When you assign a product to a category (e.g., Apparel > Shirts), Shopify "unlocks" specific Category Metafields like neckline or sleeve length.

We recommend using these native fields whenever possible. They are "machine-readable" by default, meaning they feed directly into Google Shopping, Meta, and AI search agents without you having to build custom maps.

3. Reporting & analytics

As of February 2026, Shopify has updated its analytics features to use metafields as dimensions and filters in reports.

Previously, you couldn't easily see how specific attributes performed without external tools. Now, you can filter your data by the facts that actually matter to your business:

•  Beauty & food: Filter sales reports by "Key Ingredients" (e.g., Vitamin C vs. Retinol) or dietary tags (e.g., Vegan vs. Keto).

•  Apparel: Analyse returns or conversion rates based on "Fit Type" (Slim, Relaxed, Athletic) rather than just broad categories.

•  Home goods: Compare the performance of different "Material" types across your entire catalog.

4. Tags as a merchandising layer

Tags are for temporary states. They give your team the agility to move fast without over-complicating your underlying data structure.

• What they are: "New in," "Sale," "Summer Collection," or "Staff Picks."

• Why they matter: They are perfect for internal warehouse flags or temporary badges that you might only need for a few weeks.

How to migrate from tags to metafields

Most stores that have this issue are a bit of a mess, and that’s fine. Don’t attempt a "big bang" migration. Instead:

1. Standardise your tags: Start using snake_case (e.g., material_leather instead of just leather). This makes it easy to spot which tags are "data" and which are "merchandising."

2. Move one category at a time: Pick your best-selling collection and move its core attributes into metafields first.

3 Fill the gaps: As we saw in the reporting section, AI agents love detail. If you don’t provide the facts in your metafields, the AI will fill those gaps with its own (often incorrect) narrative.