Sub-event components in data feeds

Sub-event components are dimensions and metrics based on fields inside an array or map in your XDM schema. They allow you to analyze data at a level more granular than the event level, such as the individual products in a purchase. For information about using this data in segments, see Sub-events.

Use the following information to understand how sub-event components from array and map fields appear in your Customer Journey Analytics data feeds.

Understand sub-event components

Sub-event components in the XDM schema

In the XDM schema, each element of an array (a string array or an object array) is a sub-event. Each entry in a map field is also a sub-event, as described in Map fields in data feeds. Dimensions and metrics based on the fields within a sub-event are sub-event components.

To view sub-events within the XDM schema in Adobe Experience Platform, select Schemas, then expand an event that contains sub-events.

In the following example, Product list items is an object array containing various sub-event components.

Sub-event differences between Analysis Workspace and data feeds

Sub-event components are represented differently between Analysis Workspace and data feeds in Customer Journey Analytics.

Sub-event differences between Adobe Analytics and Customer Journey Analytics

Sub-event data (such as multiple product details in a single purchase event) appears differently in Customer Journey Analytics data feeds than in Adobe Analytics data feeds. The following table compares how each product represents sub-event data.

Sub-event example: Products in a purchase event

A customer purchases two products in a single order: one cordless drill and two drill battery packs. Your implementation sends a single purchase event that includes both products in the productListItems object array:

{
  "eventType": "commerce.purchases",
  "timestamp": "2026-09-16T14:32:07.512Z",
  "commerce": {
    "purchases": { "value": 1 }
  },
  "productListItems": [
    { "SKU": "CD-2000", "name": "Cordless Drill", "quantity": 1, "priceTotal": 129.99 },
    { "SKU": "BP-2000", "name": "Drill Battery Pack", "quantity": 2, "priceTotal": 39.98 }
  ]
}

This event contains two sub-events, one for each object in the productListItems array. The following table shows which fields belong to the event and which belong to its sub-events.

Add sub-event components to a data feed

When you add a sub-event component to a data feed, a dialog prompts you to add the other components from the same sub-event.

Fields from the same sub-event appear on the canvas as a collapsible nested group rather than a flat item.

This group reflects the underlying data structure.

In the data feed output, all of these components appear as a nested array in a single column.

For information about how to add components, including sub-event components, to a data feed, see Create a data feed.

Query sub-event data in data feed output

Because sub-event data appears differently in Customer Journey Analytics data feeds, the queries you use for it differ from those you use for Adobe Analytics data feeds.

The following examples show how to find events that include a specific product. The examples use Google BigQuery syntax, where UNNEST expands an array into rows. Other data warehouses support the same approach with their own array functions, such as FLATTEN in Snowflake or explode in Databricks.

{
  "row_id": "01K3F2M9-...-4821",
  "timestamp_utc": "2026-09-16T14:32:07.512000Z",
  "product_list_items": [
    { "category": "Power Tools", "product": "Cordless Drill", "quantity": 1, "revenue": 129.99,
      "events": {"event1": 1}, "merchandising": {"eVar10": "DrillBundle"} },
    { "category": "Power Tools", "product": "Drill Battery Pack", "quantity": 2, "revenue": 39.98,
      "events": {"event1": 1}, "merchandising": {"eVar10": "DrillBundle"} }
  ]
}

SELECT row_id, timestamp_utc, product_list_items
FROM `project.dataset.cja_data_feed` AS f
WHERE EXISTS (
  SELECT 1
  FROM UNNEST(f.product_list_items) AS item
  WHERE item.product = 'Cordless Drill'
);

SELECT f.row_id, f.timestamp_utc, item.product, item.quantity, item.revenue
FROM `project.dataset.cja_data_feed` AS f,
     UNNEST(f.product_list_items) AS item
WHERE item.product = 'Cordless Drill';

Power Tools;Cordless Drill;1;129.99;event1=1;eVar10=DrillBundle,Power Tools;Drill Battery Pack;2;39.98;event1=1;eVar10=DrillBundle

SELECT hitid_high, hitid_low, post_evar10
FROM aa_hit_data
WHERE REGEXP_CONTAINS(product_list, r'(^|,)[^;]*;Cordless Drill;')

Use map fields in data feeds

Map fields in your XDM schema store key-value pairs. Data feeds export each map as an array of objects, the same way as other sub-event data. Each object contains the map key and its value as separate fields.

Field names in the output come from the component IDs that you configure for the data feed, not fixed names such as key or value. The examples in this section use sample component IDs.

Simple maps

Simple maps are the map type that you can create in your own schema. Each key is a string, and each value is a string or an integer.

For example, a survey map stores each question as a key and the response as a value:

{
  "_yourtenant": {
    "surveyResponses": {
      "How did you hear about us?": "Search engine",
      "How likely are you to recommend us?": 9
    }
  }
}

In the data feed output, survey_question and survey_answer are the component IDs for the key and the value:

{
  "survey_responses": [
    { "survey_question": "How did you hear about us?", "survey_answer": "Search engine" },
    { "survey_question": "How likely are you to recommend us?", "survey_answer": 9 }
  ]
}

Identity map

Each identity in the identityMap field is exported as one object. The object contains the identity namespace (the key), along with the identifier, authenticated state, and primary flag. The namespace repeats for each identity in that namespace.

Only the identity map attributes that exist as dimensions in your data view and that you add to the data feed are exported.

{
  "identity_map": [
    { "identity_namespace": "ECID", "identity_id": "83290187457380573620940587193016478103", "authenticated_state": "ambiguous", "is_primary": true },
    { "identity_namespace": "CRMID", "identity_id": "C-1048576", "authenticated_state": "authenticated", "is_primary": false }
  ]
}

Nested maps

Some Adobe-defined fields, such as segmentMembership, are maps of maps. Data feeds flatten these into a single array, with the first-level key and second-level key as separate fields in each object. The first-level key repeats in each object it applies to, so no data or relationships are lost.

For example, segment_namespace and segment_id are the component IDs for the first-level key and the second-level key:

{
  "segment_membership": [
    { "segment_namespace": "ups", "segment_id": "04a81716-43d6-4e7a-a49c-f1d8b3129ba9", "status": "realized" },
    { "segment_namespace": "ups", "segment_id": "53cba6b2-a23b-454a-8069-fc41308f1c0f", "status": "exited" }
  ]
}