Each type of product recommendation has different data dependencies, and the Insight data checker can identify issues in your data. This tool appears in the sidebar when you create a new product recommendation.
For products to be available for selection in a product block, they need to be added to an account's AccountInsight in a collection name that's prefixed with catalog_.
The checks identify two categories of problem:
Warnings
You can still create the recommendation, but the results won't be optimal.Errors
The state of the data prevents us from creating a recommendation.
The Insight data checks on two levels: catalog and product. The catalog level lets you know that your schema is wrong, whereas the product level zooms in on the specific product that's causing the problem.
Validation types
The below table details the Insight data checks that we perform.
Type | Description | Status |
Schema validation | Checks whether the dependent Insight data collections use a valid schema; do they have the correct fields?
| Error |
SKU validation | Checks whether each product has a unique SKU. This field must be unique, otherwise, your best sellers may wrongly attribute sales. | Warning |
URL validation | Checks whether your products have valid links. Without a valid link, contacts aren't able to click through to view the product. Links must be full URLs pointing to your live website. | Warning |