Trait qualification, or trait realization, is treated differently in Audience Manager, depending on trait type. See Trait Qualification by Trait Type for details about trait type qualification.
Additionally, see Real-time Segment Population and Total Segment Population for details about segment qualification.
|Trait qualification happens in real-time, as users qualify for a trait in their browser. Your users will start qualifying for a rule-based trait approximately 4 hours after you create the trait in the UI. Rule-based traits allow you to use recency and frequency controls for ad frequency capping and other use cases.
|Trait qualification happens after an inbound file is processed, i.e. the inbound file is imported into Audience Manager and that is when the trait qualification happens. You should wait approximately 4 hours after creating an onboarded trait before uploading an inbound file for processing. For onboarded traits, the maximum number of qualifications for a user profile is 1.
|For algorithmic traits, the maximum number of qualifications for a user profile is 1.
|A folder trait sums up the trait qualifications of the traits it contains. Read Folder Traits: About for more information.
|Active Audience Traits and Data Source Synced Traits
|An Active Audience trait contains all of the devices under management in your Audience Manager account. Data Source Synced Traits track all of the users associated with a data source. Read more about Active Audience Traits and Data Source Synced Traits.
Depending on the type of results that you want the graph to show (filtered by Device ID or Cross-Device ID), the metrics have different meanings:
When filtering the results by Device ID:
When filtering the results by Cross-Device ID:
Think of the numbers this way. In the image above, from the Trait Details view, 90,173 represents the number of active devices, that visited your properties yesterday. The Total Trait Population of 55,757 represents the amount of users currently qualified for this trait. The Total Trait Population figure is meant to show the total amount of users who could be used for segmentation/targeting. Typically, users will remain part of a trait for 120 days.
Because we run two different computational jobs to calculate the two populations, the Total Trait Population always lags behind the Unique Trait Realizations by 24 hours. In the graph above, you can see about 90,400 Unique Trait Realizations and a Total Trait Population of about 90,300 for February 5th. The 90,400 profiles are added to the Total Trait Population on the following day.
To further drive the point home, if you experienced a spike of 10,000 visitors right now, they would show up in tomorrow’s Unique Trait Realizations, but would only show up 24 hours later in the Total Trait Population.
Any change in trait realizations reflects in segment populations.
The Real-time Segment Population counts the number of devices that have qualified for the selected segment and have reached your properties, within the selected time interval.
The Total Segment Population counts the number of devices that have qualified for the selected segment within the selected time range. The 1 Day report represents the most up to date segment population count.
Think of the numbers this way. In the image above, from the Segment Details view, 9,993 represents the number of active devices, that visited your properties yesterday, and qualified for the segment. The Total Segment Population of 699,532 represents the total number of devices currently qualified for this segment. The Total Segment Population figure is meant to show the total number of devices that could be used for segmentation/targeting.
Because we run two different computational jobs to calculate the two populations, the Total Segment Population always lags behind the Real-time Segment Population by 24 hours. In the graph above, you can see a 8,116 Real-time Segment Population and a Total Segment Population of 742,000 for February 2nd. The 8,116 profiles are added to the Total Segment Population on the following day.
To further drive the point home, if you experienced a spike of 10,000 visitors right now, they would show up in tomorrow’s Real-time Segment Population, but would only show up 24 hours later in the Total Segment Population.
We enforce a limit of 150,000 trait qualifications for each user profile, whether it is an authenticated profile (DPUUID) or a device ID (UUID). Note that while the DPUUIDs are unique to a specific instance of Audience Manager, UUIDs are shared across the Audience Manager platform. For UUIDs, we impose a fairness policy when storing trait qualifications. An algorithm ensures that an equal share of the UUID profile is made available for every instance of Audience Manager.