E-Commerce Product Taxonomy & Intent Graph

**A fully resolved intent-to-product graph for e-commerce**

There is a widening gap between how retailers describe products and how consumers search for them. A catalog lists "Carbon-Fiber Road Frame." A shopper asks, "Hey Siri, what's the best bike for steep hills?" This dataset bridges that gap at scale.

We analyze the titles, descriptions, and naming structures of 372 million product listings from 714,000 data-integrity-screened retailers.. These listings are semantically decomposed into 29.6 million canonical product neighborhoods, each enriched across five dimensions: categories, demographics, occasions, use cases, and seasons. 22.3 million buyer intent keywords are matched against these neighborhoods, producing 223 million keyword-to-product connections, updated weekly.

**What's included: 15 tables, 2 views, 1.07B rows**

- PRODUCT_BRIDGE (361M): Product-to-cluster mapping
- INTENT_PRODUCT_MATCHES (223M): Keyword-to-cluster relevance scores
- WEEKLY_DIMENSIONS (148M): Weekly breakdowns across 12 dimensions
- PRICE_HISTORY_EVENTS (95M): Price changes tracked since Feb 2024
- KEYWORD_FINANCIALS (53.6M): Per-keyword CPC, competition, search volume
- TITLE_CANONICALS (29.6M): Cluster metadata with enrichment tags
- FINANCIAL_METRICS (29.6M): 18 financial signals per cluster
- INTENT_PRODUCTS (22.3M): Master keyword table

<p><br/></p>

Views: CLUSTER_FULL, DIM_FINANCIAL_DIMENSIONS

**Financial metrics per cluster:** median price, min/max range, seller count, CPC, competition index, revenue potential, total demand, keyword count, demand per seller, opportunity score (0-100), price trend, 13-week rolling median, price momentum, demand velocity.

**Use cases:** market sizing, gap analysis, competitor research, price intelligence, keyword and ad research, and seasonal demand planning.