Catalog Structure: Unified vs Channel Catalogs
The Catalog page is built on a multi-layer structure that separates raw product data from unified product entities.
This structure allows EVA to both preserve channel-specific data and create a connected product view across systems.
Understanding this separation is essential, because every action in the Catalog page depends on which layer you are working in.
The Big Picture
EVA does not store products as a single flat list. It separates products by source and then creates a unified layer when products are matched.This allows accurate cross-channel analysis while keeping original data intact.
The Structure
The Catalog page consists of three main layers:
Amazon Catalog
The Amazon Catalog contains products sourced directly from Amazon.
Each product is structured using:
- ASIN
- Parent ASIN
- SKU
- Price and status information
This layer reflects how products exist in Amazon and is not modified by matching.
Shopify Catalog
The Shopify Catalog contains products from Shopify at the variant level.
Each row represents a variant and includes:
- Variant ID
- SKU
- Price
- Inventory-related fields
Unlike Amazon, Shopify data is variant-based, meaning a single product may appear as multiple rows.
Unified Catalog
The Unified Catalog is created when products from different channels are matched.
Each unified product represents a connection between:
- One or more Amazon ASINs
- One or more Shopify variants
This layer is where EVA builds a cross-channel product identity.
The Logic
EVA follows a three-step structure:
- Ingest raw product data from each channel
- Keep channel data unchanged in separate catalogs
- Create unified products only when a match is applied
This means:
- Amazon and Shopify data remain independent
- Matching does not overwrite original data
- Unified products act as a linking layer, not a replacement
Why This Matters
Without this structure, cross-channel analysis would be unreliable.
By separating and then connecting data:
- Channel-specific details remain accurate
- Product identity becomes consistent across systems
- Actions like pricing, analytics, and advertising can operate on a unified product
Key Value Pillars
Data integrity is preserved
Raw channel data is never altered by matching.
Cross-channel identity is created
Unified products allow one product to be understood across multiple systems.
Flexibility in workflows
Users can work in raw catalogs or unified view depending on the task.
Pro Insight
- Think of the Unified Catalog as a relationship layer, not a data source.
- Amazon and Shopify catalogs contain the actual product records.
- The Unified Catalog defines how those records are connected.
This distinction is what allows EVA to combine accuracy with flexibility.