To build rich user and multi-user profiles, we identify key indicators throughout the user life cycle, as well as offering the option to ingest existing datasets and combine this with explicit user signal. These profiles are used in conjunction with our enriched metadata, real-time trending topics, and understanding of the wider user-base, to power our world-class recommendation engine.
There is a huge variation when it comes to individual consumption patterns and behaviours. This is exacerbated by the mixing of linear and on-demand content. Many existing solutions fail in the content landscape of today. Our nuanced approach to the challenge allows us to offer a much more effective solution.
By leveraging the connections within the unified metadata repository, better content filtering can be provided, related content can be highlighted and in turn, a better signal is collected to unlock collaborative filtering.
The next level builds upon user activities and their hidden signals for content desirability. Using the latest ML/AI techniques, prediction models are generated and help drive user engagement and ultimately retention.