Creating and activating segments on your unified data is a key feature of Data Cloud. While this feature makes Salesforce Data Cloud more than just any other Customer Data Platform, anyone who has worked with it may know that it can sometimes be tricky – from Data Cloud’s subscriber key selection criteria, to related attributes in activations, and considerations of your segment refreshes.
This article aims to compile all you need to know about Activations on Data Cloud along with its limitations and considerations, as well as some workarounds that could be considered to help you navigate through it. There will also be a special focus on tips and tricks relating to Data Cloud Activations into Salesforce Marketing Cloud.
What Are Activations?
Activation is the process of publishing segments from your data cloud instance into your activation target (the target platform destination of your segment). While your segment refers to the population of people that fits within the segment criteria that you have defined, activations take that population of people to send into your activation target (target destination), with the additional ability to select (supplementary) attributes relating to that segment.
Activations differ from Data Actions – read about the differences between them in the guide below:
Activating on Related Attributes
When setting up your Activation, you may select between direct attributes or related attributes. However, you will find that, unlike segmentation, activating related attributes comes with various exceptions you should be aware of.
I have compiled some important system limitations and considerations outlined by Salesforce for activating related attributes:
- When activating a segment with 10M population with related attributes in the activation; the segment will not activate.
- 20 related attributes per activation.
- 100 Activations with Related Attributes.
- Work around: After reaching this limit, delete related attributes from existing activations, delete unused activations, or create activations with only direct attributes.
Perhaps most importantly of these limitations, you can only activate related attributes that sit on the same, singular data model object path. The UI reminds you of this by graying out other path options once you have selected your first related attributes to include in your activation.
This is a rather significant limitation of the system as it forces and limits your available selections. While a potential work around would be to simply create new relations to that data model object (which in itself could also lead to complications in managing these relations), realistically there will also be data model objects that cannot be connected to one another (i.e. data model objects that have a many to many relationship).
In addition to only being able to select attributes within a singular path, you may only select data model objects that are “hops away” from your segment-on entity.
A potential work around should your use case require a significant amount of related attributes would be to consider using data cloud to build and activate your segments, but enriching your segment population with data on an external platform (such as Salesforce Marketing Cloud).
Segment Refreshes: Considerations
As part of setting up your Activations, you may select the type of refresh: Full refresh, which updates all contacts in a segment, or Incremental refresh, which updates only new, updated, or deleted segments.
In the case that the segment is used in an ongoing journey, there is a risk of removing the contact from the data extension prior to journey completion through these refresh settings as a contact may no longer fit in the segment criteria but have already been injected into an active journey in Marketing Cloud. This will have implications if using “Journey Data” for email personalization in journey, as they will be removed – Marketing Cloud cannot find the associated Journey Data of the contact.
Work around: set automation to import the Shared Data Extension from Data Cloud into a separate data extension that will then be used as the audience for your journey. Data retention policy can be applied to data extension to remove old records.
Activating to Salesforce Marketing Cloud: Subscriber Key Selection
It is important to note that the selection criteria for segmentation and activation are different, and separate from one another. This point is particularly important in the case of segmenting and activating on the Unified Individual. Subscriber keys are imperative to the functionality of Salesforce Marketing Cloud. As such, it is important to understand how Data Cloud makes this important choice for you.
When creating an Activation into Marketing Cloud, you will be prompted to select a contact point (email or mobile), determined by the communication channel of this segment. Once a channel selection is made, you are then allowed to select a source priority order; this allows you to prioritize the selection of subscriber keys on the basis of its data source.
In the case where under a Unified Profile, the Individual IDs under that Unified Profile come from the same source, Data Cloud will decide which subscriber key to write based on 2 criteria in this order:
- SubscriberKey with the best Engagement Score in Marketing Cloud.
- The earliest subscriber key based on alphanumerical order (0- 9, A-Z).
The email address that is activated will follow with the necessary subscriber key. It should also be noted that despite documentation suggestions, it is possible for Data Cloud to activate back contacts with null email addresses.
You may work around this by adding ‘Individuals with null email addresses’ in the Exclude tab of your segment criteria, or by setting email addresses as a non-nullable field in the Shared Data Extension of the segment (post-activation), which would result in the rejection of this contact upon reactivation as it fails to meet the criteria of the data extension.
Conclusion
Activations in Data Cloud is an incredible tool that enables its users to convert its unified insights into actionable data across various systems.
Hopefully, these practical tips and tricks will help you navigate, and get more out of data with Data Cloud.