DocumentationExperience PlatformData Governance Guide

Automatic policy enforcement

Last update: Wed Jul 12 2023 00:00:00 GMT+0000 (Coordinated Universal Time)

CREATED FOR:

  • Developer
  • User
  • Admin
  • Leader
IMPORTANT
Automatic policy enforcement is only available for organizations that have purchased Adobe Healthcare Shield or Adobe Privacy & Security Shield.

Once data is labeled and data usage policies are defined, you can enforce data usage compliance with policies. When activating audiences to destinations, Adobe Experience Platform automatically enforces usage policies should any violations occur.

NOTE
This document focuses on the enforcement of data governance and consent policies. For information on access control policies, refer to the documentation on attribute-based access control.

Prerequisites

This guide requires a working understanding of the Platform services involved in automatic enforcement. Please refer to the following documentation to learn more before continuing with this guide:

  • Adobe Experience Platform Data Governance: The framework by which Platform enforces data usage compliance through the use of labels and policies.
  • Real-Time Customer Profile: Provides a unified, real-time consumer profile based on aggregated data from multiple sources.
  • Adobe Experience Platform Segmentation Service: The segmentation engine within Platform used to create audiences from your customer profiles based on customer behaviors and attributes.
  • Destinations: Destinations are pre-built integrations with commonly used applications that allow for the seamless activation of data from Platform for cross-channel marketing campaigns, email campaigns, targeted advertising, and more.

Enforcement flow flow

The following diagram illustrates how policy enforcement is integrated into the data flow of audience activation:

An illustration of how policy enforcement is integrated into the data flow of audience activation.

When an audience is first activated, Policy Service checks for applicable policies based on the following factors:

  • The data usage labels applied to fields and datasets within the audience to be activated.
  • The marketing purpose of the destination.
  • The profiles that have consented to be included in the audience activation, based on your configured consent policies.
NOTE
If there are data usage labels that have only been applied to certain fields within a dataset (rather than the entire dataset), enforcement of those field-level labels on activation only occurs under the following conditions:
  • The fields are used in the audience.
  • The fields are configured as projected attributes for the target destination.

Data lineage lineage

Data lineage plays a key role in how policies are enforced in Platform. In general terms, data lineage refers to the origin of a set of data, and what happens to it (or where it moves) over time.

In the context of Data Governance, lineage enables data usage labels to propagate from schemas to downstream services that consume their data, such as Real-Time Customer Profile and Destinations. This allows policies to be evaluated and enforced at several key points in the data’s journey through Platform, and provides context to data consumers as to why a policy violation occurred.

In Experience Platform, policy enforcement is concerned with the following lineage:

  1. Data is ingested into Platform and stored in datasets.
  2. Customer profiles are identified and constructed from those datasets by merging data fragments according to the merge policy.
  3. Groups of profiles are divided into audiences based on common attributes.
  4. Audiences are activated to downstream destinations.

Each stage in the above timeline represents an entity that may contribute to policy enforcement, as outlined in the table below:

Data lineage stage
Role in policy enforcement
Dataset
Datasets contain data usage labels (applied at the schema field level or entire dataset level) that define which use cases the entire dataset or specific fields can be used for. Policy violations will occur if a dataset or field containing certain labels is used for a purpose that a policy restricts.

Any consent attributes collected from your customers are also stored in datasets. If you have access to consent policies, any profiles that do not meet the consent attribute requirements of your policies will be excluded from audiences that are activated to a destination.
Merge policy
Merge policies are the rules that Platform uses to determine how data will be prioritized when merging together fragments from multiple datasets. Policy violations will occur if your merge policies are configured so that datasets with restricted labels are activated to a destination. See the merge policies overview for more information.
Audience
Segmentation rules define which attributes should be included from customer profiles. Depending on which fields a segment definition includes, the audience will inherit any applied usage labels for those fields. Policy violations will occur if you activate an audience whose inherited labels are restricted by the target destination’s applicable policies, based on its marketing use case.
Destination
When setting up a destination, a marketing action (sometimes called a marketing use case) can be defined. This use case correlates to a marketing action as defined in a policy. In other words, the marketing action you define for a destination determines which data usage policies and consent policies are applicable to that destination.

Data usage policy violations occur if you activate an audience whose usage labels are restricted for the target destination’s marketing action.

(Beta) When an audience is activated, any profiles that do not contain the required consent attributes for the marketing action (as defined by your consent policies) are excluded from the activated audience.
IMPORTANT
Some data usage policies may specify two or more labels with an AND relationship. For example, a policy could restrict a marketing action if labels C1 AND C2 are both present, but does not restrict the same action if only one of those labels are present.
When it comes to automatic enforcement, the Data Governance framework does not consider the activation of separate audiences to a destination as a combination of data. Therefore, the example C1 AND C2 policy is NOT enforced if these labels are included in separate audiences. Instead, this policy is only enforced when both labels are present in the same audience upon activation.

When policy violations occur, the resulting messages that appear in the UI provide useful tools for exploring the violation’s contributing data lineage to help resolve the issue. More details are provided in the next section.

Policy enforcement messages enforcement

The sections below outline the different policy enforcement messages that appear in the Platform UI:

Data usage policy violation data-usage-violation

If a policy violation occurs from attempting to activate an audience (or making edits to an already activated audience) the action is prevented and a popover appears indicating that one or more policies have been violated. Once a violation has triggered, the Save button is disabled for the entity you are modifying until the appropriate components are updated to comply with data usage policies.

Select a policy violation in the popover’s left column to display details for that violation.

The violation message provides a summary of the policy that was violated, including the conditions the policy is configured to check for, the specific action that triggered the violation, and a list of possible resolutions for the issue.

A data lineage graph is displayed below the violation summary, allowing you to visualize which datasets, merge policies, audiences, and destinations were involved in the policy violation. The entity that you are currently changing is highlighted in the graph, indicating which point in the flow is causing the violation to occur. You can select an entity name within the graph to open the details page for the entity in question.

You can also use the Filter icon ( ) to filter the displayed entities by category. At least two categories must be selected in order for data to be displayed.

Select List view to display the data lineage as a list. To switch back to the visual graph, select Path view.

If you have created consent policies and are activating an audience to a destination, you can see how your consent policies affect the percentage of profiles that are included in the activation.

An enhancement to consent policy enforcement on batch and streaming destinations including paid media activations has been made. This enhancement is available to customers of Privacy and Security Shield or Healthcare Shield, and proactively removes profiles from batch and streaming destinations as consent status changes. It also ensures that consent changes are propagated immediately so that the right audience is always targeted.

These improvements allow for greater trust in your marketing strategy as it removes the need for marketers to manually add consent attributes to their segment expression. This ensures that no profiles are inadvertently targeted for any marketing experiences once consent has been withdrawn or no longer qualified for a consent policy. The marketing consent policies that set rules for how consent or preference data should be managed across various marketing workflows are now automatically enforced in activation workflows in downstream solutions.

NOTE
There are no UI changes as a result of this enhancement.

Pre-activation evaluation

Once you reach at the Review step when activating a destination, select View applied policies.

View applied policies button in the activate destination workflow

A policy check dialog appears, showing you a preview of how your consent policies affect the consented audience of the activated audiences.

Consent policy check dialog in the Platform UI

The dialog shows the consented audience for one audience at a time. To view the policy evaluation for a different audience, use the dropdown menu above the diagram to select one from the list.

The audience switcher in the policy check dialog.

Use the left rail to switch between the applicable consent policies for the selected audience. Policies that are not selected are represented in the “Other policies” section of the diagram.

Policy switcher in the policy check dialog

The diagram displays the overlap between three groups of profiles:

  1. Profiles that qualify for the selected audience
  2. Profiles that qualify for the selected consent policy
  3. Profiles that qualify for the other applicable consent policies for the audience (referred to as “Other policies” in the diagram)

The profiles that qualify for all three of the above groups represent the consented audience, summarized in the right rail.

Summary section in the policy check dialog

Hover over one of the audiences in the diagram to show the number of profiles it contains.

Highlighting a diagram section in the policy check dialog

The consented audience is represented by the central overlap of the diagram, and can be highlighted like the other sections.

Highlighting the consented audience in the diagram

Flow run enforcement

When data is activated to a destination, the flow run details show the number of identities that were excluded due to active consent policies.

Excluded identities metrics for a dataflow run

Policy enforcement for activated audiences policy-enforcement-for-activated-audiences

Policy enforcement still applies to audiences after they have been activated, restricting any changes to an audience or its destination that would result in a policy violation. Due to how data lineage works in policy enforcement, any of the following actions can potentially trigger a violation:

  • Updating data usage labels
  • Changing datasets for an audience
  • Changing audience predicates
  • Changing destination configurations

If any of the above actions triggers a violation, that action is prevented from being saved and a policy violation message is displayed, ensuring that your activated audiences continue to comply with data usage policies when being modified.

Next steps

This document covered how automatic policy enforcement works in Experience Platform. For steps on how to programmatically integrate policy enforcement into your applications using API calls, see the guide on API-based enforcement.

recommendation-more-help
834e0cae-2761-454a-be4d-62f0fd4b4456