To use product recommendations, you must sync the correct insight data from your commerce platform. Your products must be added as an account-scoped Insight data collection and the collection name must be prefixed with catalog_.
Learn more in The non-developer's guide to Insight data.
Dependencies
The table below details the types of product recommendation, the Insight data collections required to power them, and the fields used to match products between collections.
The unique product identifier is set at the catalogue level. By default, Name and SKU are used, but you can change this.
Learn more in Change the unique product identifier.
Product recommendation type | Required Insight data collections | Product match |
Custom category | Products | N/A |
Best sellers | Products, Orders | Unique product identifier |
Most viewed | Products, WebInsight | Product URL |
Last browsed | Products, WebInsight | Product URL |
Trending | Products, Orders, WebInsight | Unique product identifier and product URL |
Also bought* | Products, Orders | Unique product identifier |
Lookalikes (predictive) | Products, Orders | Unique product identifier |
Best next (predictive) | Products, Orders | Unique product identifier |
New in store | Products | N/A |
Bought together* | Products, Orders, Cart (optional) | Unique product identifier |
Hybrid prediction | Products, Orders | Unique product identifier |
*For Also bought and Bought together recommendations, orders or cart product SKU, or order or cart product URL fields are used to access recommendations for a specific contact.
Data schemas
For product recommendations to work correctly, your product and order data must follow a specific structure. Refer to the data schemas for the required structure:
Read more in the Web behavior tracking section of the Help Centre.
Sync your data to Dotdigital
Sync Insight data through an integration or directly through the API.