Audience Composer is a new type of audience targeting technology developed to address shortcomings in traditional DMPs and CDPs. Rather than simple, true/false audience classifiers and targeting guesswork, Helm uses machine learning to assemble and push audiences scored by receptivity to the GlassBox programmatic model builder.
There is an important difference between knowing what a user
Clients retain ownership of all data stored and processed by audience composer and can access it directly using a Presto distributed SQL query engine, enabling users to quickly assemble new audience composites from cross-channel, first-party data and push the data to programmatic RTB targeting systems.
Audience Composer identifies overlapping attributes across 160,000 audiences to develop scalable composite audiences, scored by media receptivity.
The attributes include user behaviors, traits, and interests —Including demographic information, media consumption habits, hobbies, local weather, product affinities, in-market signals, brand affinities, political views, and shopping patterns.