When you're deep into building magic with Salesforce Data Cloud whether that's crafting powerful identity resolution rules, tuning calculated insights, or slicing your data into sharp segments eventually, you'll hit a wall: "How do I move all of this safely to another environment?
Think of a Data Kit as a metadata suitcase for Salesforce Data Cloud. It doesn’t carry customer data — it carries your configuration: the brainwork, the logic, and the structure of your data intelligence. It lets you version, migrate, and deploy Data Cloud configurations across sandboxes and production environments in a structured, reliable way.
Salesforce offers two flavors of Data Kits depending on your use case:
A single DevOps Data Kit can be re-used across multiple stages (Dev → QA → UAT → Prod), making it the preferred option for teams focused on compliance, velocity, and repeatability.
Data Cloud configurations have dependencies, so the publishing sequence is critical. You can't publish a calculated insight before its referenced DMO exists. Salesforce handles this behind the scenes for most deployment methods — but if you’re crafting your own manifest, you'll want to respect the order:
A place to view the deployment status and deployment errors. A few sample errors are shown in the screenshot below.
Salesforce offers flexibility depending on your comfort level and workflow. Here's how you can get your Data Kit from dev to destination:
For the admin who loves the UI, this is the easiest path:
PS: Don’t forget to click ‘View/Add Dependencies’. Adding just the data kit doesn’t add the components inside it.
Great for smaller teams or one-off deployments.
For DevOps teams or advanced users who love Git and CI/CD:
sf project deploy start --manifest manifest/package.xml --target-org your-sandbox
This method is version-controllable, repeatable, and ideal for release pipelines.
Want pinpoint control? You can manually write your own package.xml with exactly what you want to deploy:
<?xml version="1.0" encoding="UTF-8"?>
<Package xmlns="http://soap.sforce.com/2006/04/metadata">
<types>
<members>MessagingSession_Home</members>
<members>SalesforceDotCom_Home</members>
<name>DataKitObjectDependency</name>
</types>
<types>
<members>JoinMStoConversation</members>
<members>MessagingSession_Home</members>
<members>SalesforceDotCom_Home</members>
<name>DataKitObjectTemplate</name>
</types>
<types>
<members>MIAW_Chat_Transcript</members>
<name>DataPackageKitDefinition</name>
</types>
<types>
<members>MIAW_Chat_Transcript_1743688369456</members>
<members>MIAW_Chat_Transcript_1743688371158</members>
<members>MIAW_Chat_Transcript_1743688371409</members>
<members>MIAW_Chat_Transcript_1743688397273</members>
<members>MIAW_Chat_Transcript_1743688410331</members>
<members>MIAW_Chat_Transcript_1743688410951</members>
<name>DataPackageKitObject</name>
</types>
<types>
<members>MIAW_Chat_Transcript</members>
<members>Salesforce_Home</members>
<name>DataSource</name>
</types>
<types>
<members>MIAW_DataStreams</members>
<name>DataSourceBundleDefinition</name>
</types>
<types>
<members>MIAW_Chat_Transcript</members>
<members>MessagingSession_Home1</members>
<name>DataSourceObject</name>
</types>
<types>
<members>MessagingSession_Home_1743688371181</members>
<name>DataStreamTemplate</name>
</types>
<version>63.0</version>
</Package>
Deploy it with CLI or plug it into your automated pipeline. This is DevOps gold.
|
Method |
Best For |
Tools Needed |
|
Change Set |
Admins, quick tests |
Salesforce UI |
|
Manifest + CLI |
DevOps, versioning |
Salesforce CLI, Git |
|
Manual package.xml |
Fine-tuned deployments |
VSCode, CLI |
Data Kits are a game-changer in managing Data Cloud deployments. Whether you're working solo or across a multi-org enterprise, they give you the structure, control, and flexibility you need to move fast without breaking things.
And best of all? You choose how to use them — clicks, code, or full CI/CD.