When the Scoring Framework template configuration is complete, Scoring Framework creates the resources used to generate and store personalized promotion targets.
Required Editions
| Available in: Lightning Experience |
| Available in: Enterprise, Performance, Unlimited, and Developer Editions with Loyalty Management and Data Cloud |
When activation completes, Salesforce creates the following resources.
- A scoring framework app instance.
-
Training and scoring batch data transforms.
- A training batch data transform with the same name as the template configuration is created.
- When a template configuration is activated for the first time, a new scoring batch
data transform with the same name as the template configuration is created. If a scoring
batch data transform already exists, a new transform definition row is added to the
existing definition.
Important A scoring batch data transform supports up to 5 transform definitions. Ensure yours contains 4 or fewer definitions before activating a new template configuration, or activation fails.
To learn how to access and manage batch data transforms, see Batch Data Transforms.
- When a configuration template is activated for the first time, a target Data Model Object (DMO) is created with the same name as the template. Predicted values are written to this DMO, and outputs from subsequent template activations are also stored in the same DMO.
- The scoring batch data transform creates prediction records.
- When no products or categories are selected, the system creates 4 records per member, one for each of the 4 periods immediately before the scoring period. Each record holds the prediction for the period that follows it. For example, to generate targets for December, the data transform creates records for August, September, October, and November. The November record stores the prediction for December.
- When products or categories are selected, the system creates records for each combination of member and selected product or category, plus one aggregate record per member. The system also creates all of these records for 4 periods. For example, if you select 2 products, the system creates 3 records per member (one per product and one for the aggregate), multiplied across 4 periods, creating 12 records per member.