The Filter builder dialog is used to create new or edit existing filters. The dialog is titled New filter or Edit filter for filters that you create or manage from the Filters manager.
Specify the following details ( is required):
Element | Description |
---|---|
Data view | You can select the data view for the filter. The filter you define is available as a filter in the Settings tab of a data view. |
Project-only filter | An info box to explain that the filter is only visible in the project where it is created and that the filter will not be added to your component list. Enable Make this filter available to all your projects and add it to your component list to change that setting. This info box is only visible when you create a quick filter and turn the quick filter info a regular filter using Open builder from the Quick filter interface. |
Title | Name the filter, for example, Last month mobile customers . |
Description | Provide a description for the filter, for example, Filter to define the mobile customers for the last month . |
Tags | Organize the filter by creating or applying one or more tag. Start typing to find existing tags you can select. Or press ENTER to add a new tag. Select to remove a tag. |
Definition | Define your filter using the Definition builder. |
To verify whether your filter definition is correct, use the constantly updated preview of the results of the filter at the top right.
To create an audience from the filter and share the audience with Experience Platform, select Create audience from filter. See Create and publish audiences for more information.
Select:
You use the Definition builder to construct your filter definition. In that construction, you use components, containers, operators and logic.
You can configure the type and scope of your definition:
You can always change these settings later.
A vital part of the construction of your filter definition is using dimensions, metrics, existing filters and date ranges. All these components are available from the component panel in the Filter builder.
To add a component:
To edit a component:
To delete a component:
You can group multiple components in one or more containers and define logic within and between containers. Containers allow you to build complex definitions for your filter.
You can use in a container for the following actions:
Container action | Description |
---|---|
Add container | Add a nested container to the container. |
Exclude | Exclude the result from the container in the filter definition. A thin red left bar identifies an exclude container. |
Include | Include the result from the container in the filter definition. Include is the default. A thin gray left bar identifies an include container. |
Name container | Rename the container from its default description. Type a name in the text field. If you provide no input, the default description is used. |
Delete container | Delete the container from the definition. |
You can build filters that contain rolling date ranges. So, you are able to answer questions about ongoing campaigns or events. For example, you can build a filter that includes everyone who has made an online purchase over the last 60 days.
This video demonstrates the functionality using Adobe Analytics. However, the functionality is similarly available in Customer Journey Analytics. Be aware of the following differences in terminlogy.
Adobe Analytics | Customer Journey Analytics |
---|---|
Segments | Filters |
Visitor | Person |
Visit | Session |
Hit | Event |
You can build a filter using filters. When you use filters in a filter, you can optimize your filter and reduce the complexity.
Imagine you want to filter on the combination of device type (2) and US states (50). You could either build 100 filters, each for the unique combination of device type (mobile phone versus tablet) and US state. To get the Californian tablet users, you would use one of the 100 filters:
Or, you could define 52 filters: 50 filters for the US states, one for mobile phone and one for tablet. And then stack the filters to obtain the same results. To get the Californian tablet users, you would stack two filters:
When you use a dimension in the Filter builder, you have the options to specify the attribution model for that dimension. The attribution model you select determines whether data qualifies for the condition you have specified for the dimension component.
Select within the dimension component and select one of the Attribution models from the popup:
Models | Description |
---|---|
Repeating model (default) | Include instance and persisted values for the dimension to determine qualification. |
Instance | Include only instance values for the dimension to determine qualification. |
Non-repeating instance | Include unique instance (non-repeating) values for the dimension to determin qualification. |
As part of a filter definition you have specified the following condition: Page Name equals Women. Similar as to the example above. You repeat this filter definition using the two other attribution models. So you have three filters each with their own attribution model:
The table below explains, for each attribution model, which incoming events are qualified for that condition.
Women Page - Attribution - attribution model |
Event 1: Page Name equals Women |
Event 2: Page Name equals Men |
Event 3: Page Name equals Women |
Event 4: Page Name equals Women (persisted) |
Event 5: Page Name equals Checkout |
Event 6: Page Name equals Women |
Event 7: Page Name equals Home |
---|---|---|---|---|---|---|---|
Repeating (default) | |||||||
Instance | |||||||
Non-repeating instance |
An example report on events using the three filters looks like: