ARM for Salesforce Data Cloud

Last Updated: Oct 31st 2025 Applies to: AutoRABIT ARM 25.4.5 and later

Overview

Salesforce Data Cloud (formerly Customer Data Platform) is not just another Salesforce cloud — it’s a real-time customer data platform (CDP) that unifies, harmonizes, and activates data across systems. ARM now supports full Data Cloud metadata deployment through Data Kits, enabling seamless movement of Data Cloud configurations between orgs.

This guide walks you through end-to-end deployment of Data Cloud metadata even if you’ve never used Data Cloud before.

What is Data Cloud?

Data Cloud helps organizations achieve a 360° customer view by combining and activating data from multiple internal and external systems.

Capability

Description

Data Unification

Merge CRM, transactional, web, and third-party data into one profile.

AI & Insights

Built-in Einstein AI generates predictive insights and segmentation.

Governance

Secure and compliant with GDPR, HIPAA, and encryption policies.

Activation

Push unified data to Sales, Marketing, and Service Clouds for personalization.

Ecosystem Support

Works with Stripe, Google Drive, Snowflake, and other external systems.

Key Components of Data Cloud Metadata

Below are the main metadata types that define your Data Cloud configuration:

Before You Begin

Make sure you meet the following prerequisites before attempting a deployment.

Permissions

System Permissions:

API Enabled — so the user can call the APIs needed for ingestion, metadata, streams, etc.
Modify Metadata Through Metadata API Functions — to allow changes to metadata (objects, fields, permission sets) via Metadata API.
View Setup and Configuration — required to access Setup, enabling features like Data Cloud Setup, Data Spaces, etc.

Object and Field Access: Ensure the deployment user has full access to all Data Cloud objects and fields referenced in Data Streams and Data Model Objects.

Step 1: Create a DevOps Data Kit

Go to Setup → Data Kit.
Click New DevOps Data Kit (ensure DevOps, not Standard, is selected).
Provide a descriptive name, e.g., Customer360_DevOpsKit_v1.
Add components such as:
- Connections (external data sources – credentials not migrated)
- Data Streams
- Data Lake / Data Model Objects
- Calculated Insights
- Identity Resolution Rules
- Segmentation Rules
- Data Source Metadata
Click Download Manifest (package.xml).

Note: Standard Data Kits (used in customer solutions) lock certain metadata post-installation.

Step 2: Commit in AutoRABIT ARM

Navigate to Create New → New EZ-Commit.
Select the Data Cloud Source Org and DX repository (recommended).
Choose Fetch Changes from Package Manifest and upload your package.xml.
Do not select “Apply Auto Draft.”
After retrieval, select all components and review for missing/deleted files.
Click Commit.
Once commit is complete, note the revision number. Note: When committing a Permission Set that includes Data Cloud Objects, you don’t need to select “Commit Options for PermissionSet.” ARM automatically identifies and includes the related Object and Field permissions during the commit. This is because Data Cloud metadata is already handled as part of the core Permission Set logic, so manual selection isn’t required for successful deployment.

Step 3: Deploy in ARM

Go to Create New → New Deployment.
Choose Single Revision and pick the revision created in Step 2.
Select the Target Org.
Retrieve and select all components.
Click Deploy.

After the deployment succeeds:

In your target org, navigate to Data Cloud Setup → Data Kits → [Your Kit] → Deploy, to activate the components in the default Data Space.

Step 4: Post-Deployment Tasks

Reauthorize connectors (for ingestion/streaming) — credentials don’t migrate.
Verify deployment under Deployment History.
Check Data Spaces and object/field permissions.
Activate Data Cloud components if required.Best Practices

Metadata Support Matrix

API Name

Support in ARM

Dependent via Package.xml

Independent

DataPackageKitDefinition

Yes

DataPackageKitObject

Yes

DataSource

Yes

DataSourceBundleDefinition

Yes

DataSourceObject

Yes

DataSrcDataModelFieldMap

Yes

DataStreamTemplate

Yes

FieldSrcTrgtRelationship

Yes

DataKitObjectDependency

Yes

DataKitObjectTemplate

Yes

ActivationPlatform

Yes

ActvPlatformAdncIdentifier

Yes

CustomerDataPlatformSettings

Yes

DataConnectorIngestApi

Yes

DataSourceTenant

Yes

DataStreamDefinition

Yes

ExternalDataConnector

Yes

ExternalDataSource

Yes

ExtDataTranFieldTemplate

Yes

ExtDataTranObjectTemplate

Yes

InternalDataConnector

Yes

MarketSegmentDefinition

Yes

MktCalcInsightObjectDef

Yes

MktDataTranObject

Yes

ObjectSourceTargetMap

Yes

ActivationPlatformActvAttr

Yes

StreamingAppDataConnector

Yes

ActivationPlatformField

Yes

ActvPfrmDataConnectorS3

Yes (Excluded for SFDX)

ActvPlatformFieldValue

Yes (Excluded for SFDX)

AIPlugInUtteranceDef

Yes

DataConnector

Yes

DataConnectorS3

Yes

DataSourceField

Yes

ExternalDataTranObject

Yes

Testing Conclusions -

The metadata types which have Package.xml - Yes should follow the above mentioned steps for Commit, Deployments & CI Jobs.
The metadata types which have Package.xml - No can follow the regular steps [any type] for Commit, Deployments & CI Jobs. (Deployments might fail for the components [Details are mentioned in the table]).
Merge, Release Label, Branching Baseline is working for all the metadata types with both DX & NON DX repositories.

** All the above implies for both Constructive and Destructive changes.

Quick Summary for New Users

Create a DevOps Data Kit with all your Data Cloud metadata.
Use EZ-Commit with package.xml (DX repo only).
Deploy via Single Revision and activate from Data Cloud Setup.
Always reauthorize connectors and verify Data Space configurations.
Do not mix standard Salesforce metadata with Data Cloud types.

PreviousARM Supported Metadata Types NextOn-Premises / Dedicated Instances

Last updated 5 days ago

Was this helpful?