Combine combine
The Combine activity is a type of Targeting activity that enables you to segment your inbound population effectively. It allows you to merge multiple populations, exclude specific segments, or retain only the data shared across several targets.
The following segmentation options are available:
-
Union: merges the results of multiple activities into a single unified target.
-
Intersection: retains only the elements that are common across all inbound populations.
-
Exclusion: removes elements from one population based on specified criteria.
Configure the Combine activity combine-configuration
Follow these common steps to start configuring the Combine activity:
- Add multiple activities such as Build audience activities to form at least two different execution branches.
- Add a Combine activity to any of the previous branches.
- Select the segmentation type: union, intersection or exclusion.
- Click Continue.
- In the Sets to join section, check all the previous activities you wish you join.
Union combine-union
Within the Combine activity, you can configure a Union by selecting a Reconciliation type to determine how duplicate records are managed:
-
Keys only (default): retains a single record when multiple inbound transitions share the same key. This option is only applicable when the inbound populations are homogeneous.
-
A selection of columns: allows you to specify which columns are used for data reconciliation. Select Add attribute.
In the following example, a Combine activity is used with a Union to merge the results of two queries, Loyalty Members and Purchasers, into a single, larger audience that includes all profiles from both segments.
Intersection combine-intersection
In the Combine activity, you can configure an Intersection. For this, you need to follow the extra steps below:
-
Select the Reconciliation type to define how duplicates are handled:
-
Keys only (default): retains a single record when multiple inbound transitions share the same key. This option is only applicable when the inbound populations are homogeneous.
-
A selection of columns: allows you to specify which columns are used for data reconciliation. Select Add attribute.
-
-
Enable Generate completement if you wish to process the remaining population. The complement contains the union of all inbound activity results, excluding the intersection. An additional outbound transition is added to the activity.
The following example illustrates the use of the Intersection between two query activities. It is used to identify profiles who are Loyalty Members and have made a purchase within the last month.
Exclusion combine-exclusion
In the Combine activity, you can configure an Exclusion. For this, you need to follow the extra steps below:
-
In the Sets to join section, choose the Primary set, which represents the main population. Records found in the other sets are excluded from this primary set.
-
When needed, you can adjust inbound tables to align targets from different dimensions. To exclude a target from another dimension, it must first be brought into the same targeting dimension as the main population. To do this, click Add a rule and define the conditions for changing the dimension. Reconciliation is then done using either an attribute or a join.
-
Enable Generate completement if you wish to process the remaining population. The complement contains the union of all inbound activity results, excluding the intersection. An additional outbound transition is added to the activity.
The following Exclusion example shows two queries configured to filter profiles who purchased a product. The profiles who do not have a loyalty membership are then excluded from the first set.
