The Q2 SMART system categorizes users based on their financial behavior or shared attributes. These categories are called traits and assist in building target audiences for a campaign.
View the list of available traits in the Trait store.
Table 7. Traits expected values
|
Data type |
Expected value |
Example |
|---|---|---|
|
boolean |
true or false |
The user has a checking account or user does not have a checking account |
|
date |
date up to 365 days |
The user completed a transaction for their checking account in the last 120 days |
|
float |
fraction-based numbers |
The user has two checking accounts with the FI; the user's total checking account balance is $5k |
|
enum |
predefined value |
The user's preferred language in UUX is Spanish |
|
string |
alphanumeric results |
SMART identifies the user in the ZIP Code trait value 90210 |
Q2 SMART uses data model machine learning to create its insight traits like Recommended Audiences. This means SMART looks at how different elements of data relate to each other to make highly probable assumptions about users. Sometimes, additional logic is placed around these assumptions to advance accuracy. For example, traits based on an RFM score are complex traits.
Detailed information on insight trait data makeup can be found within the Q2 SMART Product Documentation in the Q2 Customer Portal.