Behavior Filtering

更新时间:
复制 MD 格式

Behavior filtering is a method of audience filtering based on a behavioral dataset. This method selects audiences whose behavior records meet specified requirements.

Procedure

  1. Go to Workspace > User Insights > Audience Management > Audience Filter to open the Audience Filter page.

  2. Click the Behavior Filtering tab and select a dataset from the drop-down list on the right.

  3. Set the Statistics Time, Behavior Channel, Behavior Type, Behavior Object, Number of Behaviors, and Behavior Amount as prompted on the interface.

    Note

    • When you update an audience, relative times such as "This Week" and "This Month" are recalculated based on the update date. The audience is then filtered based on the new time range. If you set a specific time range, the results do not change after an update unless the underlying user behavioral data changes.

    • You can make multiple selections for Behavior Channel and Behavior Object. Select All to include all items, or perform a search and then select Select All to include all search results.

    • For behavioral data collected using Quick Tracking, you must select the Behavior Channel, Behavior Type, and Behavior Object in order. Otherwise, you cannot proceed.

  4. Click Add Condition. The condition is added to the Selected Conditions list on the right, and the corresponding number of users is displayed.

  5. Click a condition in the list on the right to view its details on the left.

    • After modifying the configuration, click Update Condition to update the selected condition.

    • Click Add New. The existing conditions on the right are retained, and the editing area on the left is cleared so you can add another condition.

    • Hover your mouse over a condition to display the Remove button. Click Remove to delete the filter condition.

    • You can perform intersection, union, and difference calculations on multiple conditions. The default calculation is union. Click the union icon 并集, intersection icon 交集, or difference icon 并 to switch between methods. You can drag conditions to adjust their calculation level.

      Note

      In multilayer calculations, only the outermost layer (the final layer in the calculation) supports the difference calculation.

      1

  6. Enter a name and remarks for the audience. Select a folder to save the audience in and an associated sub-campaign (for more information, see the Marketing Campaign document). Specify whether the audience is public and select the ID type to save for the audience. Click Save to complete the audience filtering.

    Note
    • Select the ID type for the saved audience from the ID types available in the dataset. If you select a non-user identity ID, the actual number of saved audience members may differ from the displayed count, because the count is based on user identity IDs.

    • When you use an audience for pushes and automated marketing, you can use both the saved ID type of the audience and other ID types from its source dataset.

References