Export arrays, maps, and objects from Real-Time CDP

Learn how to export arrays, maps, and objects from Real-Time CDP to cloud storage destinations. Additionally, you can export map-type fields to enterprise destinations and limited edge personalization destinations. Read this document to understand the export workflow, the use cases enabled by this functionality, and known limitations. View the table below to understand the functionality available per destination type.

Consider this page your go-to place for anything that you want to know about exporting arrays, maps, and other object types from Experience Platform.

Bottom line up front

Get the most important information about the functionality in this section, and continue below to the other sections in the document for detailed information.

• For cloud storage destinations, the ability to export arrays, maps, and objects depends on your selection of the Export arrays, maps, objects toggle. Read more about it further down on the page.
• You can export arrays, maps, and objects to cloud storage destinations in JSON and Parquet files. For enterprise and edge personalization destinations, the exported data type is JSON. People and prospect audiences are supported, account audiences are not.
• For file-based cloud storage destinations, you can export arrays, maps, and objects to CSV files, but only by using the calculated fields functionality and concatenating them into a string by using the array_to_string function.

Arrays and other object types in Experience Platform

In Experience Platform, you can use XDM schemas to manage different field types. Before support for array exports was added, you were able to export simple key-value pair type fields such as strings out of Experience Platform to your desired destinations. An example of such a field that was supported for export previously is personalEmail.address:johndoe@acme.org.

Other field types in Experience Platform include array fields. Read more about managing array fields in the Experience Platform UI. You can now export array objects such as the example below.

organizations = [{
  id: 123,
  orgName: "Acme Inc",
  founded: 1990,
  latestInteraction: "2024-02-16"
}, {
  id: 456,
  orgName: "Superstar Inc",
  founded: 2004,
  latestInteraction: "2023-08-25"
}, {
  id: 789,
  orgName: 'Energy Corp',
  founded: 2021,
  latestInteraction: "2024-09-08"
}]

In addition to arrays, you can also export maps and objects from Experience Platform to your desired cloud storage destination. Read more about maps and objects in Experience Platform.

Prerequisites

Connect to a desired cloud storage destination, progress through the activation steps for cloud storage destinations and get to the mapping step. When connecting to the desired cloud destination, you must select the Export arrays, maps, objects toggle on. Get more information in the section below.

Export arrays, maps, objects toggle

When connecting to a file-based cloud storage destination, you can set the Export arrays, maps, objects toggle on or off.

Toggle this setting on to enable the export of arrays, maps, and objects to JSON or Parquet files. You can select these object types in the source field view of the mapping step when activating audiences to cloud storage destinations. However, with this setting on, you cannot use the calculated fields option to transform data on activation.

With this toggle off, you can use the calculated fields option and apply various data transformation functions when activating audiences. However, you can not export arrays, maps, and objects to JSON or Parquet files and must configure a separate destination for that purpose.

Export arrays, maps, objects toggle on

With this setting on, you can export entire objects (for example person.name) and arrays by selecting them via the source field selector in the mapping step of the activation workflow.

With this option selected, the user interface blocks users from using calculated fields, and the Add calculated fields control is disabled, as shown below. To use calculated fields for data transformations, set up a destination connection with the toggle off.

Export arrays, maps, objects toggle off

With this option set to off, you can use the calculated fields option and apply various data transformation functions when activating audiences. However, you can not export arrays, maps, and objects to JSON or Parquet files and must configure a separate destination for that purpose.

You can export arrays, maps, and objects to CSV files by using the calculated fields functionality and concatenate them into a string by using the array_to_string function. Read more about using that function.

Read more about working with calculated fields to perform transformations on data exported to cloud storage destinations.

Sample exported files

By using this functionality, you can export Parquet and JSON files where the data preserves the structure from Experience Platform. View below an example of an exported JSON file.

{
  "person_name_firstName": "John",
  "person_name_lastName": "Smith",
  "_acmeinc_customer_hs_main_address_scalar": "Oak Avenue No 12",
  "_acmeinc_customer_hs_locations_array": [
    "home address 12",
    "office address 12"
  ],
  "_acmeinc_customer_hs_date_array": [
    "2024-11-14",
    "2024-11-15"
  ],
  "_acmeinc_customer_hs_customer_obj_emails_array0": "john.smith@example.com",
  "_acmeinc_customer_hs_customer_obj": {
    "emails_array": [
      "john.smith@example.com",
      "j.smith@example.com"
    ],
    "name_scalar": "John Smith"
  },
  "_acmeinc_customer_hs_addresses_array_obj": [
    {
      "is_primary": true,
      "streetName_scalar": "Maple Street",
      "streetNo_int": 12
    },
    {
      "is_primary": false,
      "streetName_scalar": "Pine Road",
      "streetNo_int": 45
    }
  ],
  "_acmeinc_customer_hs_addresses_array_obj0": {
    "is_primary": true,
    "streetName_scalar": "Maple Street",
    "streetNo_int": 12
  }
}