Data Cloud brings together data from different parts of your organization, making it accessible across various clouds and applications. This can sometimes be challenging to keep track of, but Data Cloud is designed to be flexible and adaptable. Like any product built on the Salesforce Platform, Data Cloud is also inherently composable.
The power of packaging, specifically Second Generation Managed Packages (2GP), makes any solution you build not just composable within one environment but across hundreds if not thousands. This is especially critical if you utilize features like Data Cloud One.
We won’t go into tremendous detail around composability because there is so much great content on this topic within the Well-Architected framework, but our focus will be to create the equivalent kind of composable system when it comes to Data Cloud. Lets start with talking about how Data Cloud supports packageability.
What is a Data Kit
To understand a Data Kit, it helps to first know what a package is—a collection of metadata elements, including objects, fields, permission sets, flows, apex, layouts, and more.
A Data Kit is a metadata bundle of elements from Data Cloud, designed to fit within Salesforce’s shared metadata framework. It can be packaged in a second-generation managed package (2GP) and installed in an environment, where it unpacks after the managed package installation.
Data Kits are essential for deploying Data Cloud across multiple environments. They handle the complex dependencies within Data Cloud, ensuring that each element installs in the correct order.
Data Kits support both first-generation managed packages and 2GP, though we recommend 2GP for its flexibility. By shifting the source of truth to source control, 2GP enables easier versioning and updates than first-generation packaging.
Data Kits and 2GP Packaging
Lets take a look at the following diagram on how a Data Kit could be packaged up within a 2GP package:
In the diagram above we identified which metadata elements of Data Cloud can be packaged up in a Data Kit. Data Cloud metadata will often have dependencies so the arrows are to indicate dependencies between entities. The groupings represent the following:
- Data Streams: a set of standard and custom connectors to ingest data into Data Cloud.
- Note: Its critical to know that while these data streams are included in the Data Kit, Connector information should already be configured or re-authenticated in target environment . Connector credentials cannot be included within a Data Kit.
- Data Sources: the data processing engines that transform data sources into data objects
- Data Objects: a container and harmonized grouping for the data brought from the connectors and data sources
- Insights: lets you define and calculate multidimensional metrics from your data stored in data objects
- Identity Resolution: uses matching and reconciliation rules to link data about people or accounts into unified profiles
- Data Actions: enables different types of event-driven integrations and orchestrations across Salesforce CRM workflows, Mulesoft and other third party systems
- Segmentation: they let you break down your data into useful segments to understand, target, and analyze your customers
The reason behind showing the dependencies between metadata items within the Data Kit is Data Cloud is driven by its dependencies. Before packaging up your Data Kits it will be critical to map out those metadata items to ensure you’ve adequately captured everything. Those dependencies then drive the sequence in which those metadata items are installed and are able to be configured within the target environment.
For a full list of the kind of metadata you can package into a Data Kit you can find that resource here in the Data Cloud Extensibility Readiness Matrix.
Now lets see what the 2GP package with Data Kit in it will look like within a Production Environment:
Deploying a Data Kit within the production environment first involves installing the 2GP package within your production environment. This makes the Data Kit metadata components available to use within Data Cloud, Salesforce CRM and Agentforce by:
- Bringing in data streams from the Data Kit and authenticate them with connectors into your Data Cloud instance
- Placing the relevant data sources, data objects and other metadata elements into your data spaces that are already within your production environment
- Create Data Cloud Triggered-Flows on DMOs and CIOs that were included in your data kit
- Create Agents in Agentforce with enriched data from Data Cloud.
When deciding what to include in a Data Kit, it’s important to consider the following guidance from the Well-Architected framework around packaging:
- Avoid duplicating metadata across packages to prevent multiple destinations from a source integration.
- Consider how packaging will impact your environment strategies, especially in terms of dependencies on Data Cloud data.
- Think about what your ideal end-state should support based on your Data Cloud and 2GP packaging strategies. Group business units and capabilities to create a modular end state that aligns with the Well-Architected framework’s Separation of Concerns.
Additional Considerations
- Data Cloud today can only be packaged in 1GP and 2GP within partner developer environments. These environments can only be made on request basis via Partner Support.
- Unlocked packages may not be supported as Data Cloud is not available for customer scratch orgs
- When packaging, a single package cannot contain Data Cloud and non-Data Cloud metadata. If you include a Data Kit within a package, you will not be able to add any other kinds of supported core platform metadata within that package.
- Data Kits are critical when it comes to using Data Cloud One to deploy Data Cloud metadata between the home and companion orgs.
Additional Resources
- Packages and Data Kits
- Second Generation Managed Packages
- Data Cloud Data Kits
- Packaging and Data Kits in Data Cloud | Salesforce Trailhead
Written by
