Datatrics has a feature where you can recommend child products to profiles and make sure you won’t recommend multiple child products from the same parent in one recommendation. We also make sure the most potential child product will be recommended (best matching size, color, etc.).
Example: The algorithm decides that it should recommend 4 different versions of the same shirt (different color) to your visitor. These however all have the same parent_id. By setting it to show just one child only the most relevant one will be shown and the other spots will be filled by items that do not share the same parent_id. This prevents 4 almost identical products (barring their color) being shown as recommendations.
To use this feature, your content needs to have the following requirements:
Parent products are optional, but if you can’t add parent data (like images) to the child product, make sure these items are present in Datatrics.
Child products need to contain one of the following attributes, the value should be the same per child and contain the itemid of the parent if you want to use parent data in child product recommendations:
- parent_id
- item_group_id
Now if you want to recommend the most potential child products to profiles (and prevent duplicates), the only thing you need to do is check the following box when filtering/selecting content in a touchpoint:
With the setting above, Datatrics will only show a maximum of one child product per recommendation (if you recommend multiple) and recommend the most potential one. This is based on conversions, shopping carts and product views. Note that product views can only be an input for the algorithm, if the URL is different per child product.
Using data from the parent product
To use parent product data when recommending a child product, you can use the following variable:
{{item.parent.VARIABLE}}VARIABLE can be changed with every product attribute of the parent. A few examples:
{{item.parent.image}}
{{item.parent.name}}
{{item.parent.description}}