The references document data model

In this model, the relationship is maintained using links between documents. References from one document is stored in another document. This process is also called normalization; it establishes the relationship between different collections and defines collections for a more specific purpose:

We use the normalized data approach in the following scenarios:

• When embedding data, it will create duplicate data but not provide significant read performance
• To represent more complex many-to-many relationships
• To model large hierarchical data