Understanding Lytics / Product Documentation / Descriptive & Predictive Modeling

Behavioral Audiences

Since using Lytics scores directly requires a level of understanding of the data science at play, Lytics also offers out-of-the-box behavioral audiences.

These audiences are essentially blends of scores that can be used alone or as a rule in a custom audience. For example, Casual Visitors describes users who come and go without showing much activity per session. The definition using Lytics scores is Users with an intensity score less than 25. Now, instead of needing to know exactly what "intensity" means in this context, or what the significance of the number 25 is, the audience Casual Visitors can be used as a building block.

An audience definition using Casual Visitors

In this example audience, the Casual Visitors characteristic is being used to filter the users who have a high affinity for "Computing". Note that the number of people who have a high affinity for "Computing" is 276,046, but the number of people who have a high affinity for "Computing" and qualify as a Casual Visitor is 185,747.

Reducing the size of a audience is a great way to make a campaign more efficient. Additionally, splitting an audience based on behavioral properties is a great way to introduce different modes of communication for different archetypes of users. For instance, Casual Visitors may not stick around long enough to answer their own questions, try engaging them with a slideout on early page visits. The opposite of Casual Visitors, Deeply Engaged Visitors, are probably determined to find that information on their own and would find a popup to be annoying. Splitting the audience keeps both archetypes engaged without accidentally detering anyone.

The Full Set of Out-of-the-box Behavioral Audiences

All behavioral audiences

  • Frequent Users: People who interact with your brand a lot.
  • Infrequent Users: People who interact with your brand occasionally.
  • Deeply Engaged Users: People who show a lot of activity when they do interact with your brand.
  • Casual Visitors: People who show little activity when they do interact with your brand.
  • Likely To Re-engage: People likely to come back based on their past activity patterns.
  • Unlikely To Re-engage: People not likely to come back based on their past activity patterns.
  • At Risk Users: People who's interaction behavior is changing for the worse.
  • Binge Users: People who show a lot of activity when they do interact with your brand.
  • Perusers: People who visit often but rarely interact deeply with your brand.

Taking it further

Remember that each of these out-of-the-box behavioral audiences can be recreated with the audience builder using Lytics scores user fields. Mastering Lytics scores will open up the possibility of new combinations of score thresholds that result in new nehavioral audiences to be used as building blocks in campaigns.