March 2026 LiveRamp Clean Room Release Notes
This release introduces new controls for collaboration and governance, improvements to activation and identity handling, and several quality‑of‑life updates.
Product Updates – General Availability
Org-Specific Default Permissions for Questions
Clean Room owners can already define default permissions for any question created or provisioned into their clean room. This release extends similar control to partner organizations:
Per-organization defaults: Each participating organization can now define default permissions for questions they author in a clean room, rather than inheriting the owner’s defaults.
Stronger protection for partner logic: Partners can more easily avoid unintentionally sharing code or reports with the owner when that isn’t desired.
Applies to new questions: Updated defaults apply to newly created or newly provisioned questions; existing questions retain their existing permission settings.
This improves governance and reduces the risk of unintentionally over‑sharing assets across organizations.
License-Aware Segmentation and Activation
Clean Room segmentation and activation now honor the customer’s Marketplace license attributes directly in the UI:
License-driven behavior: The UI now reflects key license attributes (for example, whether segmentation and activation are allowed, which destinations are allowed, and license expiration).
Clear visibility: Users can see which organizations can segment, which can activate, and which destinations are available.
Prevented misconfigurations: When a Marketplace license no longer permits segmentation or activation, the UI clearly indicates that these capabilities are unavailable and blocks users from enabling or updating them for that clean room.
This reduces confusion and avoids failed workflows caused by license configurations that no longer support a given use case.
Transcoding Enhancements for RampID in Questions and Exports
Transcoding has been enhanced to better support multi‑party workflows while preserving protections against using clean rooms to generate RampID mapping files:
Retain valid RampIDs across joins: List questions can now safely project both owner and partner RampIDs in supported clean rooms, enabling more flexible multi‑party audience workflows.
Destination‑specific encodings: For exports, RampID fields are transcoded into the destination organization’s RampID domain, aligning with how customers expect to receive audiences.
Guardrails preserved: Restrictions that prevent clean rooms from being used to produce cross‑partner mapping files remain in place.
This restores flexibility for complex, multi‑party segmentation and activation while keeping identity protections intact.
Snowflake Tags for Hybrid Clean Room Question Runs
For Hybrid clean room jobs that execute questions in Snowflake, Clean Room now attaches descriptive tags to those questions:
Per‑run identifiers: Each Snowflake question generated by a Hybrid clean room includes tags that reference the originating clean room, question, and run.
Operational transparency: Customers can distinguish LiveRamp‑originated workloads from other Snowflake activity and attribute costs to specific clean rooms and questions.
Improved troubleshooting: The tags make it easier to trace individual runs and understand their impact in Snowflake query history and usage views.
This is currently applied to Snowflake questions produced through the Hybrid clean room integration.
For more information, see “Configure a Snowflake Data Connection for a Hybrid Clean Room”.
Version Selection for Clean Compute Questions
Clean Compute questions that rely on .whl artifacts now support explicit version selection at run time:
Artifact-aware runs: A dedicated artifact folder connection automatically enumerates available .whl files.
UI dropdown: When launching a new Clean Compute run, users can select which artifact version to execute from a dropdown in the UI.
API parameterization: External API callers can specify the desired .whl artifact for a run.
This makes each run explicitly tied to a specific artifact version, improving governance, auditability, and developer ergonomics for Clean Compute‑based workflows.
Improved Version History for Clean Room Questions
The Version History view for clean room questions has been enhanced:
Input/output schema comparison: The comparison view now includes both input and output schema details for each version, making it easier to understand changes over time.
Safer interactions: Usability issues that could previously lead to unintended modifications of active versions have been addressed.
Additional UI refinements and metadata improvements will continue in future releases.
Notifications for Question and Flow Runs
Clean Room is now integrated with Connect’s notification framework for run‑completion events:
Run completion notifications: Users can opt in to email and in‑app notifications when question runs and flow runs complete.
Self‑service preferences: Notification settings can be configured from Connect’s notification preferences page.
Permission-aware delivery: Notifications are available to clean room members with appropriate permissions (for example, the ability to view reports in that clean room).
This reduces the need to manually check run statuses and lays the groundwork for broader event coverage in the future.
Bulk Settings Edit for Flows
Flow builders can now update certain settings across multiple nodes in a flow at once:
Bulk updates: Apply Compute Capacity, Report Access, and Tags to all or a subset of nodes in a single action.
Reduced manual work: Eliminates repetitive per‑node configuration for common settings after a flow is provisioned.
Per-node overrides remain possible: Individual nodes can still be adjusted independently when necessary.
Note
Export‑specific settings are not yet included in this bulk‑edit experience.
Product Updates – Limited Release
Delegated Dataset Assignment with Analysis Rules
For customers participating in the limited release, Clean Room extends how “Partners can assign” interacts with dataset Analysis Rules:
Assignment behavior respects rules: When delegation is allowed and a dataset has Analysis Rules, assignments made by partners are validated against those rules during dataset assignement.
Governed delegation: If Analysis Rule validation fails, runs that rely on those assignments cannot proceed until issues are resolved.
Owner controls: When delegation is disabled, only dataset owners can assign the dataset to questions, regardless of Analysis Rules.
This improves collaboration while maintaining strong governance over how datasets with Analysis Rules are used.
For more information, see "Assign Datasets to a Question".
Historical Exports of Completed Runs
A new historical exports capability is available in limited release:
Export past runs without recomputing: Once a question run has completed successfully, users can later select that run and trigger exports of its results to already‑provisioned destinations.
Recovery for missed exports: Supports cases where exports were not configured prior to setting up an export or when downstream ingestion requires re‑sending a job.
Retention window: Run results are retained for a fixed period (currently 90 days) for export.
This reduces rework and makes it easier to operationalize clean room outputs into external systems and BI tools.
For more information, see “Export Older Runs of Analytical Results”.
Quick Hits and Fixes
Data Connection Views and RampID: Data Connection Views do not currently support data connections that contain RampID fields; work is in progress to close this gap.
Data connection retry: A new API capability allows customers to programmatically retry a data connection, rescan its fields, and return it to a “mapping required” state after configuration changes. Refer to the Clean Room API documentation for details.