This article is part of a series of articles on personalisation. Read more.

The topic of personalisation is at times complicated, sometimes sensitive and can often be defined in any number of ways depending on who you ask.

Our aim is for every touchpoint a user has with Spotify to be in some way personalised to them as an individual in a specific context. This aim is the beating heart of our growth strategy. I want to share some insights on how we do this and what marketers might learn from it.

As a starting point, Gustav Söderström, Spotify’s Chief R&D officer, defines our general approach to personalisation as “self-driving music”. Music is like a mirror. It provides a rich and textured data set because it’s personal and uniquely reflects how you’re feeling and what you’re doing. And with that, every time a Spotify user presses play to a song, shares a track, or adds a song to a playlist, our data (which we call our streaming intelligence) gets smarter. Add scale to this and you have a self-driving personalisation strategy.