Best practices for Privacy Service

Last update: March 7, 2024

CREATED FOR:

  • User
  • Admin

Use Privacy Service to automate compliance with data privacy regulations when customers wish to access or delete their personal data from your data stores. To address these evolving business needs, Privacy Service offers a RESTful API and UI to submit access and deletion requests for customer data across Adobe Experience Cloud applications.

This guide outlines the best practices to efficiently process privacy requests and optimize completion response times when managing customer data requests.

Getting started

This guide requires a working understanding of Privacy Service and how it allows you to manage access and delete requests from your data subjects (customers) across Adobe Experience Cloud applications. You are also recommended to have read the guide on creating a privacy job request in the UI or the API, and understand how to perform these operations programmatically.

Prerequisites

Access to Adobe Experience Platform Privacy Service is controlled through granular role-based permissions in Adobe Admin Console. You need the relevant permissions in a product profile to use specific features in the Privacy Service UI and API. Contact your system administrator if you require additional permissions.

Administrators can refer to the guide on managing permissions for Privacy Service for more information.

Privacy job creation guidelines

To streamline your request processing and improve response times, consider the following guidelines when creating privacy jobs. This applies to both the API and UI methods.

  1. Maximize data subjects per request: Include as many data subjects as possible, up to 1000, per request.
  2. Group IDs for efficiency: Group multiple IDs for a single data subject (up to nine) in each request. The IDs can come from different Adobe services in the same request.
  3. Combine access and delete jobs: Include both “access” and “delete” job types in a single request if required by the data subject.
  4. Include only necessary products: Only include products that are required or licensed. Additional products can elongate processing time and increase costs.

Monitor privacy jobs status

To effectively monitor privacy jobs and check their status, Privacy Service provides three methods. The available methods are listed below in order of monitoring efficiency and productivity. Each method includes best practice guidelines to improve your experience, followed by an ideal scenario example that combines all the approaches.

Receive real-time notifications

I/O Events offer near real-time status monitoring through status events. This is the most efficient method as it avoids the need to implement polling mechanisms and incur additional API traffic.

Recommendations:

  • Webhook setup: Set up webhooks to receive push notifications when status changes occur for submitted jobs. This aids in real-time monitoring.
  • Notifications: Use notifications at both the job and product level to help monitor the progress of requests.

See the documentation on subscribing to Privacy Service events for instructions on setting up an event registration for Privacy Service notifications and how to interpret notification payloads.

Retrieve all jobs based on filters

To retrieve all your privacy job data based on any specified filters, perform a GET request to the /jobs endpoint. This API call is useful to provide a high-level view of the current job status for large sets of job IDs with only a single request. It does lack detailed product responses, but they can be found using the /jobs/{jobID} endpoint.

A GET request to the /jobs endpoint is best used to gather or compare the status data of a large set of job IDs but is not meant for regular polling type activities.

Recommendations:

  • Query parameters: Use specific filters to narrow your results, for example: data ranges, regulation types, and status (processing, complete, and so on).

You can view a list of all the current privacy jobs in your organization through the Privacy Service UI. See the managing privacy jobs in the UI documentation for information on how to filter the job request list. Alternatively, see the documentation on the use of the /job endpoint in the Privacy Service API.

The Privacy Service API documentation contains details on the available query parameter filters.

Retrieve detailed responses for a single job

To retrieve detailed responses for a single job, perform a GET request to the /jobs/{jobID} endpoint. This method is intended for deeper information gathering, like product-specific responses and success messages. A call to this endpoint is the best way to see which products have responded and which are still pending, although it is not meant for regular polling activity.

See the /jobs/{JOB_ID} endpoint documentation for details on how to check the status of a specific job.

Ideal scenario example

Use a webhook so that the system can automatically update records and provide reporting or alerts when groups of the IDs from a request are complete. If jobs are still outstanding, the system retrieves these job statuses with a GET request to the Privacy Service API /jobs endpoint and provides a high-level update of the list.

If a particular job is still pending, or has returned an error, you can retrieve the detailed response with a GET request to the /job/{jobId} endpoint.

Access request data

When data-subject information is requested, each service returns data in a format that is consistent with the way they store and use that data. Once all services have completed the request, a .ZIP archive file URL is provided in the job details to allow for this data to be downloaded. See the troubleshooting guide for information on how to download privacy job results.

The following are key items of note relating to the management of the data archive:

  • All archive files are deleted from Experience Platform servers after 30 days. You cannot query customer data that is older than 30 days.
  • The structure of the archive file includes folders for each product included in the request and the data files contained therein. Archive files or folders may be empty if no data was found for the specified ID.
  • The data for previously created jobs is only accessible for 30 days after the completion date. After that time, the data is removed from the system, and a new request must be made.

Recommendations:

  • Protect Data Archives: Both the URL and .ZIP file should be protected, as they may contain personally identifiable information (PII) for the data subject.