Predict success model building in music
The client
Sony Music Benelux is a major music company representing local and international artists across Belgium, the Netherlands, and Luxembourg. The company is known for combining creative vision with data-driven strategies and is a key player in shaping regional and international music trends.
The challenge
Sony Music Benelux aimed to build a predictive model of the success of pop songs, using A.I. in combination with fMRI data. The goal was to test whether brain responses to pop songs could predict the future success of music streaming.
The solution
During fMRI, participants listened to randomly selected songs from a pool of unreleased pop songs. Six months after the release of the songs, we trained an A.I. model by comparing the brain responses
with the actual streaming results of the songs. This allowed us to build a predictive model of musical success.
100
songs used to train the model
10
songs tested for prediction
93,3%
prediction accuracy
The results
Our predictive model achieved 93.3% accuracy in predicting the success of unreleased pop songs.
This result highlights the power of combining neuroscience with A.I. to forecast the success of a song before it hits the market.
See how it works in action
Let's make an impact
together!
Have questions?
Let's explore how neuroscience and A.I. can drive your success