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

Which types of datasets are subject to TTL?

TTL applies only to time-series datasets. Record-type datasets (such as entity datasets and classification datasets) are not subject to TTL. They are listed in the table above for reference, with n/a indicated in both the Data Lake TTL and Profile Store TTL columns.

Will this change apply to production sandboxes only or will it apply to dev sandboxes as well?

This change applies 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
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
  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: Not applicable - Experience event retrieval is not supported via data sources.

• 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).

How does the new TTL affect use cases that require longer data retention (e.g., excluding profiles who received an email in the past 120 days, or capping emails over a year)?

The new TTL policy will limit the look-back period for system-generated dataset data in the profile store to 90 days and in the data lake to 13 months. Use cases that require access to data beyond these periods will be impacted. For example, audience segmentation or frequency capping based on events older than 90 days in the profile store will no longer be possible using system datasets.

What alternatives are available for retaining data longer than the TTL?

Customers who require longer retention have two options:

• Export to external storage: export relevant data from AJO datasets before the TTL expiration. Adobe Journey Optimizer supports exporting datasets to various cloud storage destinations (Amazon S3, Azure Blob, Google Cloud Storage, etc.). Learn more
• Data Distiller derived datasets: customers with a Data Distiller entitlement can set up automated queries to copy critical data into a derived dataset in the data lake, which can be stored without a TTL. Learn more

What should customers do to prepare for the TTL change?

• Review your use cases and identify any that require data retention beyond the new TTLs.
• Set up automated queries to copy critical data to derived datasets before data is deleted.
• Work with your Adobe representative to discuss any additional needs or potential TTL extensions (planned for future releases).

How are customers notified before the TTL is enforced on existing sandboxes?

Adobe notifies impacted customers before enforcing the TTL on existing sandboxes. For this rollout, Adobe sent an in-product notification and published this change in the product release notes, giving customers about two months’ notice before the guardrail takes effect on October 1, 2026.

Can I delete Journey Optimizer system-generated datasets?

Journey Optimizer system-generated datasets are protected and cannot be deleted through the standard Adobe Experience Platform UI. These datasets are essential for Journey Optimizer functionality and are managed by the system.

If you need to permanently remove a Journey Optimizer system dataset (e.g., for QA environments, sandbox cleanup, or specific data hygiene requirements), please contact Adobe Engineering or Adobe Customer Care. These datasets require specialized backend procedures to ensure complete and safe removal.

note
NOTE
For routine data cleanup within these system datasets, use the Data Lifecycle operations available through the Privacy Service to delete specific records or identities. Learn more

AI Knowledge Reference

This section contains structured knowledge intended to support interpretation, retrieval, and question answering related to this topic.

For complete understanding, this information should be combined with the documentation on this page. Neither source is intended to stand alone; the page describes the feature, while this section provides additional context that helps disambiguate terminology, intent, applicability, and constraints.

• TL;DR: This page explains the time-to-live (TTL) retention guardrails on Journey Optimizer system-generated datasets in the profile store (90 days) and the data lake (13 months), including rollout timing and options for retaining data longer.

Intents:

• Understand how long system-generated dataset data is retained in the profile store versus the data lake.
• Identify which datasets are subject to TTL.
• Plan for the enforcement on existing customer sandboxes starting October 1, 2026.
• Retain data beyond the TTL using dataset export through Destinations or Data Distiller derived datasets.
• Understand the impact on segmentation, computed attributes, tracking, and Customer Journey Analytics.

Glossary:

• Time-to-live (TTL): retention guardrail after which system-generated dataset data is dropped (product-specific)
• Profile store: store subject to the 90-day TTL for impacted datasets (product-specific)
• Data lake: store subject to the 13-month TTL for impacted datasets (product-specific)
• Time-series dataset: dataset type that is subject to TTL
• Record-type dataset: dataset type not subject to TTL (marked n/a in both TTL columns)
• Data Distiller: entitlement that allows creating derived datasets stored in the data lake without a TTL (product-specific)

Guardrails:

• Profile store TTL: 90 days for Journey Optimizer system-generated dataset data (hard limit; TTL extensions are not currently supported).
• Data lake TTL: 13 months for Journey Optimizer system-generated dataset data (hard limit; TTL extensions are not currently supported).
• AJO Message Export Dataset and AJO Message Event Metadata Dataset: 30 days data lake TTL (hard limit; both require the Message Export add-on).
• TTL is enforced on new sandboxes and new organizations as of February 2025, and on existing customer sandboxes starting October 1, 2026 (hard enforcement date).
• TTL applies only to time-series datasets; record-type datasets are n/a in both the Data Lake TTL and Profile Store TTL columns.
• TTL enforcement uses the event timestamp, not the ingestion date.
• Journey Optimizer system-generated datasets are protected and cannot be deleted through the standard Adobe Experience Platform UI.
• Computed attributes: initial backfill calculation is limited to the last 90 days of data; subsequent look-back is a maximum of 6 months.
• Segmentation and retargeting look-back is limited to 90 days on affected profile store data.

Terminology:

• Canonical name: Time-to-live — Acronym: TTL
• Do not confuse: “profile store TTL” (90 days) ≠ “data lake TTL” (13 months)
• Do not confuse: “time-series dataset” (subject to TTL) ≠ “record-type dataset” (n/a, not subject to TTL)
• Do not confuse: dataset TTL (drops system-generated dataset data in the profile after 90 days) ≠ deletion of the profiles themselves (the profiles are not dropped)

FAQ:

• Q: Which datasets are subject to TTL? — Only time-series datasets; record-type datasets (such as entity and classification datasets) are not subject to TTL and are marked n/a.
• Q: Does the 90-day profile store TTL delete profiles? — No; the system-generated dataset data in the profile is dropped after 90 days, not the profiles themselves.
• Q: Can I increase the TTL? — TTL extensions are not currently supported; you can export data through Destinations, or, with a Data Distiller entitlement, create derived datasets stored without a TTL.
• Q: When does enforcement reach existing sandboxes? — Starting October 1, 2026.
• Q: Which timestamp is used for enforcement? — The event timestamp, not the ingestion date.
• Q: Can I delete Journey Optimizer system-generated datasets? — No; they are protected and cannot be deleted through the standard Adobe Experience Platform UI; contact Adobe Engineering or Adobe Customer Care for permanent removal.