IAB Ad Product Taxonomy Mapper: Non-normative reference for product classification

[esteininger]

We've published a non-normative reference implementation for mapping product metadata to the IAB Ad Product Taxonomy 2.0.

What it does

Converts arbitrary product metadata (title, description, category, brand) into standardized IAB Ad Product Taxonomy categories.

Example output:

{
  "iab_product": {
    "primary": "IAB-AP-1121",
    "label": "Consumer Electronics > Wearables > Smartwatches",
    "confidence": 0.95,
    "version": "2.0"
  }
}

Key characteristics

• Standards-aligned: Maps to IAB Tech Lab Ad Product Taxonomy 2.0
• Deterministic first: Keyword matching with optional semantic fallback
• Explainable: Every classification includes matched keywords and confidence
• High performance: Sub-10ms latency for deterministic mapping

Who this is for

• SSP/DSP platform teams
• Publishers needing ad product classification
• Ad operations standardizing product feeds
• Standards engineers validating IAB compliance

Links

• npm: https://www.npmjs.com/package/@mixpeek/iab-ad-product-taxonomy
• GitHub: https://github.com/mixpeek/connectors/tree/main/iab-ad-product-taxonomy

This is intended as a reference implementation for validation and QA, not a production targeting SDK.