How Salesforce Data Cloud Works

Before Data Cloud can harmonize your data, it requires a secure ingestion pipeline to your external environments.

• Data Source Objects (DSOs): These are the native objects residing inside your source systems—such as the Lead, Contact, or Account objects within Salesforce Sales Cloud.
• Data Streams: The actual functional connections established to transition data from your DSOs into Data Cloud. Data Streams seamlessly support both real-time integrations and batch uploads.

To accelerate implementation, Salesforce offers pre-built Starter Bundles (e.g., Sales Cloud and Service Cloud bundles). Deploying these bundles automates the connection setup for standard enterprise objects, removing the need for manual configuration.

Data Categorization: Profile vs. Engagement

When connecting custom objects or external data lakes, Data Cloud requires you to categorize the incoming data schema. This taxonomy determines how data behaves during identity resolution workflows.

The Architecture: DLOs vs. DMOs

The journey from raw data ingestion to standard data modeling relies on two foundational architecture components:

 1. Source System / DSO (Raw Source)
 ↓ Connected via Data Stream
 2. Data Lake Object / DLO (Physical Storage Container)
 ↓ Standardized via Data Mapping
 3. Data Model Object / DMO (Unified Customer 360 Shape)

Creating a Data Stream automatically provisions a corresponding Data Lake Object (DLO). The DLO serves as the exact physical storage container for your raw data inside Data Cloud. However, at this stage, the data is unstructured and unaligned with the broader ecosystem.

To make this information usable, fields from the DLO must be mapped directly to a Data Model Object (DMO). DMOs are the standard semantic templates that comprise the global Customer 360 model.

4. The Foundation of Unification: The Individual Object

The single most critical DMO for true profile unification is the Individual object.

In complex multi-org enterprise environments, a single human being often exists under completely different unique keys across multiple systems. They might be tracked as a Lead ID in Sales Cloud, a converted Contact ID in a secondary business unit, and a Subscriber Key within Marketing Cloud.

By mapping every single profile-category data stream to the central Individual Object, you establish a universal anchor point. This mapping enables Data Cloud's Identity Resolution engines to successfully cross-reference disparate source records and consolidate them into one single, cohesive, and actionable customer profile.