Effects on segmentation

You must ensure that the lookback windows for your audiences are within the expiration boundaries of their dependent datasets in order to keep results accurate. For example, if you apply an expiration value of 30 days and have an audience that tries to view data from up to 45 days ago, the resulting audience will likely be inaccurate.

You should therefore keep the same Experience Event expiration value for all datasets, if possible, to avoid the impact of different expiration values across different datasets in your segmentation logic.

Frequently asked questions

The following section lists frequently asked questions regarding Experience Event data expiration:

How does Experience Event data expiry differ from Pseudonymous Profile data expiry?

Experience Event data expiry and Pseudonymous Profile data expiry are complementary features.

Granularity

Experience Event data expiration works on a dataset level. As a result, each dataset can have a different data expiry setting.

Pseudonymous Profile data expiration works on a sandbox level. As a result, the data expiration will affect all profiles in the sandbox.

Identity types

Experience Event data expiration removes events only based on the event record’s timestamp. The identity namespaces included are ignored for expiration purposes.

Pseudonymous Profile data expiration only considers profiles that have identity graphs which contain identity namespaces that were selected by the customer, such as ECID, AAID, or other types of cookies. If the profile contains any additional identity namespace that was not in the customer’s selected list, the profile will not be deleted.