51黑料不打烊

Datasets Time-to-live (TTL) guardrails ttl-guardrail

As of February 2025, a time-to-live (TTL) guardrail is rolled out to Journey Optimizer system-generated datasets in new sandboxes and new organizations as follows:

  • 90 days for data in the profile store,
  • 13 months for data in the data lake.

This change will be rolled out to existing customer sandboxes in a subsequent phase.

Impacted datasets datasets

The table below lists all impacted datasets and their respective Time-To-Live in the data lake and the profile store.

Dataset
Data Lake TTL
Profile Store TTL
AJO Message Feedback Event Dataset
13 months
90 days
AJO Email Tracking Experience Event Dataset
13 months
90 days
AJO Push Tracking Experience Event Dataset
13 months
90 days
AJO Entity Dataset
13 months
90 days
AJO Surfaces Dataset
13 months
n/a
AJO Inbound Activity Event Dataset
13 months
90 days
AJO Classification Dataset
13 months
n/a
AJO Email BCC Feedback Event Dataset
13 months
n/a
acpEntity Event Dataset
13 months
n/a
Journeys
13 months
n/a
Journey Step Events
13 months
n/a
Decision Object Repository - Personalized Offers
13 months
n/a
Decision Object Repository - Fallback Offers
13 months
n/a
Decision Object Repository - Placements
13 months
n/a
Decision Object Repository - Activities
13 months
n/a
ODE DecisionEvents - prod decisioning
13 months
n/a

Frequently Asked Questions faq

The following is a list of answers to frequently asked questions about datasets TLL.

Will this change apply to production sandboxes only or will it apply to dev sandboxes as well?
This change will apply to all sandbox types.
For the 90 day TTL in profile store, are profiles themselves impacted?
The system-generated dataset data in the profile is dropped after 90 days, not the profiles themselves.
If a system-generated dataset data is pushed to Customer Journey Analytics (CJA), will the data in CJA also be impacted by the TTL?
Data in Customer Journey Analytics is kept in sync with Experience Platform. Therefore, a removal of data due to a TTL on system-generated dataset data will also impact the data in Customer Journey Analytics.
Can customers increase the TTL for Journey Optimizer system dataset data in profile store?

TTLs extensions are not currently supported. However, work is planned to optimize the TTL process to allow for these extension requests sometime starting the latter-half of 2025.

note note
NOTE
Data stored in the profile is subject to the Total Data Volume entitlement. Therefore, any data storage increase on the profile as a result of a TTL extension would count against the Total Data Volume entitlement. Learn more
Can customers increase the TTL for Journey Optimizer system dataset data in data lake?
TTLs extensions are not currently supported. Customers can export data through Destinations to retain data longer. Learn more. Additionally, customers with a Data Distiller entitlement can create derived datasets to store the data in data lake without a TTL. Learn more
Will the following capabilities be impacted by the TTLs?
  • Look-up store: No

  • Journey capping: No

  • Offer capping: No

  • Send Time Optimization (STO): No

  • Message frequency capping (i.e., Business rules): No

  • Reporting: No

    note note
    NOTE
    A TTL is already implemented on the Customer Journey Analytics (CJA) connection, which reduces effective max look-back period of impacted dataset data to 13 months.
  • Experience Platform data source: Yes - Experience event retrieval is subject to the 90 day TTL.

  • Computed attributes: Yes - Initial backfill calculation will be limited to last 90 days of data; computed attribute will be updated based on incremental events for subsequent updates. As soon as the subsequent updates reach the look-back period (max 6 months), the TTL essentially no longer affects the computed attribute. Learn more.

  • Segmentation and retargeting: Yes - Segmentation is dependent on data in the profile store; therefore, look-back is limited to 90 days on affected dataset data.

  • Tracking: Yes - Reduces effective max look-back period of impacted dataset data to 90 days. Data from impacted datasets resides for 13 months in data lake.

What timestamp is used for TTL enforcement (e.g., for backfill use cases)?
The event timestamp is used (i.e., not the ingestion date).
recommendation-more-help
b22c9c5d-9208-48f4-b874-1cefb8df4d76