The music industry has faced major changes in the past years. Streaming has become one of the most common ways of consuming music. Discover how OpenWT used Artificial Intelligence to help Sony Music to transform its advertising activities, in order to stay ahead in an evolving environment.

Situation

Sony Music is facing the following questions:

  • How to react on declined significance of traditional advertising activities?
  • How to exploit massive amount of data giving access to individual listening behavior?
  • How to create sustainable advertising on a personalized level?
  • How to automate advertising in order to tackle the increased diversity of personalized campaigns?
  • How to measure impact of advertising activities?

OpentWT was asked to leverage Artificial intelligence technology in order to help Sony Music to address these questions and to transform its advertising activities.

Approach

We followed a 3-step approach:

  1. Creation of large scale data warehouse optimized for AI: In a powerful AI environment in the cloud, we designed a data base allowing an efficient access for machine learning algorithms. The data base was fed with several years of historical streaming data and at the same time a pipeline of real-time data was connected to it.
  2. Train AI system: With supervised and unsupervised learning algorithms we exploited the vast amount of data. This enabled us to create powerful hypothesis to detect micro trends and related target audiences which are particularly susceptible for advertising activities.
  3. Tuning and automation: The training of the AI system was accompanied by continuously launching advertising campaigns. The insights of these campaigns allowed to iteratively improve and tune the AI system.

Results

Thanks to the AI solution provided by OpenWT, Sony Music gained highly valuable insights in its advertising activities. The result of the collaboration between the music label and OpenWT is threefold:

  1. Leveraging micro trend: the AI engine scans recent data and detects in an efficient way new promising micro trends which can be leveraged in a sustainable way.
  2. Personalization of campaigns: the AI engine opens the door to automatically create campaigns on a personalized level and to react effectively on the detected micro trends.
  3. Measuring impact: the impact of a campaign can be measured with the introduction of KPIs thanks to the analysis and combination of different data sets.

The results of this project are impressive, and the concept can easily be applied to other industries. Especially if they could leverage a large amount of usage-data, microtrends can be identified to optimize marketing activities.

 

Results before and after