Documentation Audience Manager Audience Manager User Guide

Data Security in Audience Manager

Last update: April 13, 2021

Topics:
Data Governance & Privacy

Audience Manager takes data security and privacy very seriously. We work to keep our systems secure and protect your valuable data.

Audience Manager security practices include external and internal audits, activity logging, training, and other procedures designed to help protect our systems and your valuable data. We believe a secure product helps build and maintain the trust customers place in us.

In Audience Manager, we think about security in three main categories:

Security Type	Provides Support For
Information security	Enterprise-level authentication, encryption, and data storage practices
Data leakage/transparency	Deep and actionable insight into on-site activities that constitute or contribute to data leakage
Process/policy enhancements	Clients, by working with industry best practices for privacy and data security

Systems, Training, and Access

Processes that help keep our system and your data secure.

External Security Validation: Audience Manager tests security on an annual and quarterly basis.

Yearly: Once a year, Audience Manager undergoes a full penetration test conducted by an independent third-party company. The test is designed to identify security vulnerabilities in the application. The tests include scanning for cross-site scripting, SQL injections, form parameter manipulation, and other application-level vulnerabilities.
Quarterly: Once each quarter, internal teams check for security vulnerabilities. These tests include network scans for open ports and service vulnerabilities.

Systems Security: To help keep data safe and private, Audience Manager:

Blocks requests from unauthorized IP addresses.
Protects data behind firewalls, VPNs, and with Virtual Private Cloud storage.
Tracks changes in the customer and control-information databases with trigger-based audit logging. These logs track all changes at the database level, including the user ID and IP address from which changes are made.

Security Assets: Audience Manager has a dedicated network operations team that monitors firewalls and intrusion-detection devices. Only key personnel have access to our security technology and data.

Security Training: Internally, our commitment to security extends to developers who work on our product. Adobe provides formal training to developers on how to build secure applications and services.

Secure Access: Audience Manager requires strong passwords to log on to the system. See password requirements.

Privacy and Personally Identifiable Information (PII)

Processes that help keep personal information safe. For additional privacy information, see the Adobe Privacy Center.

PII Data: Audience Manager contractually prohibits customers and data partners from sending PII information into our system. Additionally, the Unique User ID (UUID) does not contain or use PII data as part of the ID-generation algorithm.

IP Addresses: Audience Manager does collect IP addresses. IP addresses are used in data-processing and log-aggregation processes. They are also required for geographic/location look-ups and targeting. Additionally, all IP addresses within retained log files are obfuscated within 90 days.

Data Partitioning

Processes that help protect data owned by individual clients.

Trait Data Partitioning: Your data (traits, IDs, etc.) is partitioned by client. This helps prevent accidental information exposure between different clients. For example, trait data in cookies is partitioned by customer and stored in a client-specific sub-domain. It cannot be read or used accidentally by another Audience Manager client. Furthermore, trait data stored in the Profile Cache Servers (PCS) is also partitioned by customer. This prevents other clients from accidentally using your data in an event call or other request.

Data Partitioning in Reports: Client IDs are part of the identifying key in all reporting tables and report queries are filtered by ID. This helps prevent your data from appearing in the reports of another Audience Manager customer.

Inbound Server-to-Server (S2S) Transfers

Adobe Audience Manager supports two main methods of transferring S2S on-boarded data files to our systems:

Both methods are designed with the security of our customer and partner data in mind while data is in flight between their systems and our system.

SFTP: For the SFTP option, most customers choose to deliver files via the Secure FTP (SFTP) protocol, which uses the Secure Shell (SSH) protocol. This method ensures that files are encrypted while in flight between the customer’s systems and Adobe’s system. For each customer, we create a jailed drop-box location on our SFTP servers, which is tied to a user account on that system. Only the customer’s credentialed and privileged internal system users can access this jailed drop-box location. This jail is never accessible to other customers.

Amazon Web Services S3 via HTTPS: For the S3 delivery option, we recommend that all customers configure their S3 clients to use the HTTPS encryption method for file transfers (this is not the default, so it must be explicitly configured). The HTTPS option is supported both by the s3cmd command line tool as well as the S3 libraries available in every major programming language. With this HTTPS option enabled, customer’s data is encrypted while in flight to our systems. For each customer, we create a separate S3 bucket sub-directory that can be accessed only by that customer’s credentials and those of our internal system users.

To add PGP encryption to your data files, see File PGP Encryption for Inbound Data Types.

Protecting Data by Escaping

Note that Audience Manager does not escape outgoing data to secure it against possible cross-site scripting (XSS), etc. It is the responsibility of the client to escape incoming data.

HTTP Strict-Transport-Security

HTTP Strict-Transport-Security (HSTS) is an industry-wide web security mechanism which helps protect against cookie hijacking and protocol downgrade attacks.

The policy instructs the web browser that once a secure HTTPS call was made to a given domain, no subsequent unsecure calls (HTTP) should be allowed to that domain. This protects against man-in-the-middle attacks, where an attacker might try to downgrade HTTPS calls to unsecured HTTP calls.”

This policy improves data security between clients and Adobe Edge servers.

Example

Let’s say the yourcompany.demdex.com domain sends traffic to the DCS via HTTP. HSTS upgrades the calls to use HTTPS instead, and all subsequent DCS calls coming from yourcompany.demdex.com will use HTTPS instead of HTTP.

See HTTP Strict Transport Security - Wikipedia for more information about HSTS.