Define the event schema and map it to Data Cloud

Updated Nov 8, 2026 · Based on web research, 2026-11-08

Once your backend can authenticate, define the schema for the objects it sends and map each one to a Data Cloud data model object (DMO) — the step that makes your events usable by identity resolution and Personalization.

Define the schema

1. In Data Cloud Setup, open your Ingestion API connector and define each object you’ll send, including its fields and data types.
2. Mark one field per object as its Primary Key — later updates to a record must reference this field.
3. Choose a Record Modified (event-time) field — Data 360 uses it to decide whether an incoming partial update is newer than what’s stored.
4. Download the connector’s object endpoints; each object gets its own REST path under /api/v1/ingest/sources/{source}/{object}.
5. Create a data stream from the connector and map the object’s fields to the target DMO’s fields.
6. Set the stream’s refresh mode to Partial if you want partial updates (only the fields sent) rather than full-record overwrites.

Send an event

Each request posts a batch of records as a JSON body with a top-level data array — one REST call can carry several records for the same object.

{
"data": [
  {
    "customer_id": 1,
    "first_name": "Rachel",
    "last_name": "Rodriguez",
    "email": "rachel@example.com",
    "event_time": "2026-11-08T02:48:33Z"
  }
]
}

A successful call returns HTTP 202 with {"accepted": true} — the data is queued and processed asynchronously, not written synchronously.