Technical
··7 min read

Data Governance Overview in Adobe RTCDP

Diagrams are simplified to focus on these data governance features

Note that this article is not legal or compliance advice. Please seek advice from your legal teams.

Introduction

In today’s digital landscape, data has become one of the most valuable assets for businesses. As companies collect increasing amounts of customer data across various touchpoints, effective data governance is crucial to build trust, ensure compliance, and derive maximum value.

Adobe’s Real-time Customer Data Platform (RTCDP) puts powerful data governance capabilities in the hands of enterprises. With robust access controls, data lineage tracking, and privacy compliance features, RTCDP can help with these operations from consent integration to flexible policy enforcement.

Consent Integration

In jurisdictions that require an explicit opt-in, obtaining consent is the first step. My recommendation is to use a Consent Management Platform (CMP) to control data collection and maintain these records. The CMP consent status can also be used by other client-side applications such as non-Adobe analytics. More recently, advertising tags in Europe also require a certified CMP to be in compliance (TCF) as soon as the start of 2024.

The Adobe Web SDK can be configured in different modes:

  1. Implied opt-in: IN
  2. Implied opt-out: OUT, PENDING

Implied opt-in means the business does not need to obtain the visitor’s consent (or the “opt-in”) before collecting their data, and hence all visitors to the website are treated as opted-in by default. However, the visitor can opt out by rejecting the cookies through the consent banner. This use case is similar to CCPA.

Implied opt-out means that the visitors should be treated as opted-out by default and cookies should not be set.

The Web SDK supports the two options above and for more granular opt-ins, these can be sent into the Profile using this method. By collecting these flags (personalize, “data” share) into the Profile store, you can accept updates from different sources such as affiliated websites or even call centres. The consent policy feature can filter out any opt-outs downstream to the Destinations.

Glossary

Labels

Marketing Actions (M.A)

Attribute-Based Access Control (ABAC)

In simple terms, ABAC is a permissions layer that can restrict access to attributes in a schema, profile viewer, audiences and fields visible while audience building. Tailored permissions are granted based on need-to-know principles. This is also a way to simplify the audience creation process by curating the attributes required for that Role.

Whilst one can go quite granular with this feature, it is important to balance the effort to manage multiple Labels and Roles and the business requirements.

Use cases for this feature are multi-brand businesses sharing a single RTCDP Sandbox, and businesses that create audiences using sensitive fields for customer experience (healthcare or financial data).

In this example, an admin creates two Roles: A & B

Role A only has access to Label A

Role B only has access to Label B

The Labels applied to the Schema flow onto the attributes in the Profile therefore, the Profile Viewer and Audience builder uses this inheritance to either show or hide attributes based on the user’s Role. Users (or logins) can belong to multiple Roles and will the visibility will be adjusted to the more restricted rule.

The fact that ABAC can prevent the use of attributes selectively by certain Roles, this can losely act a filtering method. However for more strict enforcement at Destinations, we should look to:

Data Usage Policies (a.k.a Governance/DULE Policies)

Data Usage Policies prevent Audiences to be mapped to Destinations inadvertently. It does so by check the Governance Policies which in turn are made from rules connecting Labels and Marketing Action.

In the most common scenarios, ABAC would hide the Audiences for unauthorized users so they would not be able to map Audiences to Destinations. Even if they do have access, the policies will present any Audiences using fields that are violating the policy. e.g. the policy can prevent the mapping of an email address field labelled: C2- Data cannot be exported to a third-party. to a Destination with Marketing Action: Export to Third Party.

Example of policy in the UI:

Consent Policies

The typical use case for the Consent Policies is to automatically filter out Profiles that have opted-out. These policies can be setup at channel or data label levels.

They follow an IF … THEN include… logic.

In the IF part, Marketing Actions and/or Labels can be used to define a granular condition so as to either apply the filteration based on the Destination (Marketing Action) or the data (label).

THEN, typically will contain to the standard consent flag however, custom profile attributes can also be selected.

In this example, we only want to let through the Profiles to Destinations with Marketing Action: Export to Third Party if this Profile attribution consentOptOut is N (False).

Data Lifecycle/Hygiene

In RTCDP, there are a robust set of tools to manage large, complicated data operations in order to orchestrate consumer experiences. The two methods to achieve this are:

  1. Automated dataset expirations
    • This can be also used to delete for temporal audiences or campaigns based Datasets so as not to bloat the system.
  2. Delete records
    • Record deletions are meant to be used for data cleansing, removing anonymous data, or data minimisation.

As this feature removes data at the Dataset level, it will also clear the Identity and Profile stores. Audiences could also be updated due to the last Profile fragment removal or data changes affecting the audience qualification.

Privacy Service

The Adobe Experience Platform Privacy Service is the best method for compliance for privacy regulations like the General Data Protection Regulation (GDPR) use cases such as:

  1. Data subject access requests
  2. Data subject delete requests

Note that the privacy request needs to include individual RTCDP stores to fully remove the data:

Profile Management

The use case for profile management is to reduce the number of Profiles and licence compliance for those customer close to their licence limits. The main culprit for overages is online data containing non-durable cookie IDs or low authentication rates.

Since the value of this data decreases over time, there are two methods to maintain a lean and efficient instance. Both methods have a configurable rolling deletion logic or TTL:

  1. Experience Event Expirations
    • Dataset level
    • Deletes purely based on timestamp
  2. Pseudonymous profiles data expiration
    • Sandbox level
    • Deletes based if there aren’t durable ID/Cross device ID on the event

Note that Audiences cannot retrieve historical data further than the TTL set in either method. Since the data is only removed from the Profile Store, with the correct licence and Data Lake storage, you can query the data and push it back into the Profile Store using the Data Distiller.

Conclusion

With consumers and regulators demanding more data transparency and security, the brands that engender trust and compliance will be the ones that succeed. These RTCDP features will help your organisation efficiently navigate current and future requirements.

Views are my own, content is educational only, All Rights Reserved.