Documentation

This is the early access documentation preview for Custom Views. This documentation might not be in sync with our official documentation.

Define Product Types

To begin defining your product data model, look at your products and identify the basic Product Types and the Attributes needed for each Product Type.

Product Types group similar products together. By similar, we do not mean products having the exact same set of attributes, but those with a significant percentage of attributes that are common (usually 80% or above, ideally higher than that), and a smaller number that will be populated for a subset of the products. To get a better understanding of how different elements of the data model interact, see product information data model.

Always plan your Product Type structure carefully, keeping in mind the different kinds of products you'll be managing and the kind of information you'll need for them. A good way to validate is to check how well a subset of your product conforms to your data model design. The following are some things we suggest you keep in mind:

  • Avoid a large number of Product Types, if possible, as they are more complex to maintain.
  • Pay attention to the number of Attributes per Product Type.
  • Store information in a Product that relates to:
    • The customer finding the product (explore, search and filter)
    • The sales process
    • The processing of product information in downstream systems such as order processing or fulfillment

Think if you really need a certain information as a dedicated attribute or as part of a group of information, for example, provided as one block of plain text or JSON. The goal is to identify the most effective model for your specific needs.

After you've identified the Product Types and Attributes you'd need, creating them in the Merchant Center is a straightforward process.

When defining Attributes, think carefully about the data type and constraints. It can be challenging to change them later, especially if you already have Products associated with the Product Type.

The example Algolia integration of the Store Launchpad for B2C Retail replicates the sample data's product modeling in Algolia. Other integrations are likely to be modeled similarly. Adding Attributes to Product Types will not need any change to the integration code. Changing the product modeling, including deleting or modifying Product Types, may require changes to the integration code or search schema to prevent potential data loss.

Please refer to the documentation of your connected third-party search service, for what you need to do when Product Types change.