string categorisation

product_type

Your own internal product taxonomy — free-form, hierarchical, used by you to segment your catalog for product set rules, reporting, and campaign structure. Separate from `google_product_category`, which uses Google's published taxonomy.

Also known as: product_category , internal_category

Channel support

Channel Status Field name Notes
Google Merchant Center Recommended product_type Strongly recommended even though optional — it's the field PMax product groups filter on. Docs →
Meta Commerce Manager Recommended product_type Used as a product set rule filter in Advantage+ Catalog Ads.
TikTok Shop Optional category_path
Pinterest Catalog Recommended product_type
Amazon Not supported Use Amazon's browse-node taxonomy instead.
Bing Merchant Center Recommended product_type

Why it matters

`product_type` is the field channels use when you build product set rules. Where `google_product_category` controls policy and broad ad eligibility, `product_type` controls how *you* segment within those bounds — what ends up in your high-margin product set, your seasonal set, your new-arrivals set.

`product_type` is your taxonomy, complementary to Google's. Where `google_product_category` controls policy and broad ad eligibility, `product_type` controls how you segment within those bounds — what ends up in your "high-margin" product set, your "seasonal" set, your "new-arrivals" set. The two work together: Google's taxonomy is the rules engine, your taxonomy is the strategy layer.

Most starter feeds either skip `product_type` entirely or set it equal to `google_product_category`. Both are wasted opportunities. The field exists to model your catalog the way *you* think about it. A retailer might use sub-brand hierarchies; a marketplace might use fulfillment-method hierarchies; a niche store might use customer-segment hierarchies. The depth and structure are up to you.

The one rule worth following: be consistent. `Apparel > Mens > Outerwear > Jackets > Lightweight` and `Apparel > Men > Outerwear > Jacket > Light` create two product sets that should be one, and the algorithm can't tell that's a typo. Pick a vocabulary, document it, enforce it at feed-generation time.

Format rules

  • Hierarchical, separated by ' > ' (space-arrow-space)

    Applies to: Google Merchant Center, Meta Commerce Manager, Pinterest Catalog, Bing Merchant Center

  • Maximum 750 characters

    Applies to: Google Merchant Center

  • Free-form — no canonical taxonomy to match against

    Applies to: Google Merchant Center, Meta Commerce Manager, Pinterest Catalog, Bing Merchant Center

  • Consistency across SKUs matters — 'Apparel > Mens' and 'Apparel > Men' create two product sets that should be one

    Applies to: Google Merchant Center, Meta Commerce Manager, Pinterest Catalog, Bing Merchant Center

Valid examples

Apparel > Mens > Outerwear > Jackets > Lightweight

Deep hierarchy useful for granular product sets

Home & Garden > Kitchen > Cookware > Frying Pans > Non-stick

Combines vertical and material classification

Common mistakes

Mens jackets, lightweight, outerwear

Comma-separated tags — not hierarchical, won't power product sets

Misc

Catch-all bucket defeats the purpose

Common issues involving this field

FAQ

Can product_type be the same as google_product_category?

It can, and many starter feeds do this. But you're throwing away the field's value. `product_type` lets you model your catalog the way *you* think about it — by margin, by season, by sub-brand — separately from how Google thinks about it.

How deep should the product_type hierarchy go?

Deep enough that any product set rule you want to write can target a single product_type prefix. If you find yourself writing 'product_type contains X AND custom_label_0 = Y' rules, you probably need to push X deeper into product_type.

Does product_type affect SEO or ad delivery?

Not directly — it's not a public-facing field. But because Advantage+ and PMax use it for product set rules, getting it right indirectly drives delivery quality through better segmentation.

Last reviewed: 26 May 2026

Free plan · No card

Why wait? Try it free today.

Stop managing feeds manually. Start optimising with AI in 30 seconds.

  • Free plan, no credit card required
  • 1 brand, 1 feed, 100,000 products per feed
  • Full AI Product Optimisation, Rule Engine, and 200+ channel exports
  • Pay only for AI credits when you need them