Pseudonymous profiles data expiration

In Adobe Experience Platform, you can configure expiration times for Pseudonymous profiles, allowing you to automatically remove data from the Profile store that is no longer valid or useful for your use cases.

Pseudonymous profile pseudonymous-profile

A profile is considered for Pseudonymous data expiration if it meets the following conditions:

  • The stitched profile’s identity namespaces match what the customer has specified as a pseudonymous or unknown identity namespace.
    • For example, if the profile’s identity namespace is ECID, GAID, or AAID. The stitched profile has no IDs from any other identity namespace. In this example, a stitched profile does not have either an email or CRM identity.
  • No activity has taken place in a user-defined amount of time. Activity is defined either by any Experience Events being ingested or customer-initiated updates to the profile attributes.
    • For example, a new page view event or age attribute update is considered as an activity. However, a non-user-initiated audience membership update is not considered as an activity. Currently, to compute data expiration, the tracking at a profile level is based on the time of event for Experience Events and time of ingestion for profile attributes.

Access access

Pseudonymous Profile data expiration cannot be configured through the Platform UI or APIs. Instead, you must contact support in order to enable this feature. When contacting support, please include the following information:

  • The identity namespaces to be considered for Pseudonymous profile deletes.
    • For example: ECID only, AAID only, or a combination of ECID and AAID.
  • The amount of time to wait before deleting a pseudonymous profile. The default recommendation for customers is 14 days. However, this value may differ based on your use case.

Frequently asked questions faq

The following section lists frequently asked questions regarding Pseudonymous profiles data expiration:

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

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

Granularity

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

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

Identity types

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.

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

Removed items

Pseudonymous Profile data expiration removes both event and profile records. As a result, the profile class data will also be removed.

Experience Event data expiration only removes events and does not remove profile class data. The profile class data is only removed when all the data is removed across all datasets and there are no profile class records remaining for the profile.

How can Pseudonymous Profile data expiry be used in conjunction with Experience Event data expiry?

Pseudonymous Profile data expiry and Experience Event data expiry can be used to complement each other.

You should always set up Experience Event data expiry in your datasets, based on your needs of retaining data about your known customers. Once Experience Event data expiry is set up, you can use Pseudonymous Profile data expiry to automatically remove Pseudonymous Profiles. Typically, the data expiry period for Pseudonymous Profiles is less than the data expiry period for Experience Events.

For a typical use case, you can set your Experience Event data expiry based on the values of your known user data and you can set your Pseudonymous Profile data expiry to a much shorter duration to limit the impact of Pseudonymous profiles on your Platform license compliance.

What users should be using Pseudonymous profiles data expiration?

  • If you are using Web SDK to directly send data to Platform.
  • If you have a website that serves unauthenticated customers en masse.
  • If you have excessive profile counts in your datasets and have confirmed that this excessive profile count is because of anonymous cookie-based identity namespace.
    • To determine this, you should use the identity namespace overlap report. More information about this report can be found in the identity overlap report section of the preview sample status API guide.

What are some caveats you should be aware of before using Pseudonymous profiles data expiration?

  • Pseudonymous profile data expiration runs at a sandbox level. You can choose to have different configurations for production and development sandboxes.
  • Once you have activated this feature, the deletion of profiles is permanent. There is no way to roll back or restore the deleted profiles.
  • This is not a one-time cleanup job. Pseudonymous profile data expiry will run once per day and delete profiles that match the customer’s input.
  • All profiles that are defined as Pseudonymous profiles will be affected by the Pseudonymous profile data expiration. It does not matter if the profile is Experience Event only or if it only contains profile attributes.
  • This cleanup will only occur in Profile. Identity Service may continue to show the deleted identities within the graph after the cleanup in cases where the profile has two or more associated pseudonymous identities (such as AAID and ECID). This discrepancy will be addressed in the near future.
  • Pseudonymous profile data expiration does not run immediately, and may take up to three days to process.

How does Pseudonymous profiles data expiration interact with guardrails for Identity Service data?

  • The Identity Service “first-in, first-out” deletion system could delete ECIDs from the identity graph, which are stored in Identity Service.
  • If this deletion behavior results in an ECID-only profile being stored in the Real-Time Customer Profile (Profile store), then Pseudonymous profile data expiration will delete this profile from the Profile store.
recommendation-more-help
54550d5b-f1a1-4065-a394-eb0f23a2c38b