About AudienceDB and EventDB

How it works

AudienceDB and EventDB services store structured audience and event data in Amazon Redshift, which is a data warehouse similar to Postgres. New data is stored in EventDB and AudienceDB within one hour after it reaches EventStore or AudienceStore.

You can then query and analyze the data in Amazon Redshift using your preferred SQL client or Business Intelligence (BI) tool.

Prerequisites

AudienceDB and EventDB must be activated for your account. For additional information, contact your account manager. When EventDB or AudienceDB is activated, a database is created in Amazon Redshift to store your data. New data is stored in EventDB and AudienceDB within one hour after it reaches EventStore or AudienceStore.

Each profile has its own database schema within the same Redshift instance, which means that each user has a specific Username and Password for each database schema (profile). The database schemas for AudienceDB and EventDB are created automatically when EventDB or AudienceDB are enabled for a user’s profile.

After AudienceDB and EventDB have been activated, you must enable both and select the attributes to be stored. For more information, see Enable AudienceDB and EventDB.

Tables, views, and normalized views

The columns in the Redshift database tables are named according to the attribute type and the internal attribute ID. Views and normalized views of data contain the same data as tables, but have user friendly names that make it easier to write queries. The normalized view name is similar to the view name but with the attribute ID omitted from the table name. Table names and view names are created as follows:

• Table names
  Column names are a combination of the attribute type and the attribute ID.
  For example: badge_30
• View names
  Column names are a combination of a user-friendly name and the attribute ID.
  For example: visitor - badge - fan (30)
• Normalized view names
  Column names are user-friendly names without attribute IDs.
  For example: visitor - badge - fan

Views and normalized views simplify the process of running queries with aggregations such as SUM() , MIN(), and MAX().

Data retention

The structured data stored in EventDB and AudienceDB remains available in Amazon RedShift for the length of time stated in your contract, usually 90 days. Contact your Account Manager to review the terms of your contract that determine the length of time your data is stored.

AudienceDB tables

Visit and visitor attributes are stored in database table columns according to their attribute type and name. Audiences are stored as columns in the visitors table. The keys for the tables are visit_id or visitor_id.

The following tables are available for visit and visitor data:

• Visit/Session Data: visits
• Visitor Data: visitors

In addition, the following tables exist for special attribute types:

• Arrays: visit_arrays, visitor_arrays
• Set of Strings: visit_lists, visitor_lists
• Tally: visit_tallies, visitor_tallies

For detailed information, see the AudienceDB data guide.

EventDB tables

EventDB table data includes event attributes for all events in the event feed. Table columns are named according to the attribute type and name, with only some attributes referencing internal ID’s. Standard Universal Data Object (UDO) variables are named with a udo_ prefix and most column names match their corresponding attribute names, for example: udo_event_name. For additional information, see Live Events and Feeds.

Event data coming from the Tealium Collect tag also includes information about which tags executed on the page and page performance metrics. For more information, see Tealium Collect.

The following tables are available for event data:

• Event Feed Data: events_{FEED}