The popular app reveals how powerful their data set really is.

In a new talk, Shazam VP Cait O’Riordan broke down some aspects of the popular app’s data set showing just how much information they can gather from their users and how they can predict which songs will hit the Billboard #1 over a month in advance. In the video, O’Riordan shows what they can assess simply based on how many people use the app to identify a song, when they use it, and at what point in the song they use it.

For example, she uses the weekend spikes on one electronic artist to show how they are gaining popularity in clubs, even if they have little radio presence. Later the analytics from multiple hit pop songs are laid over each other to reveal their similar attention-grabbing formulas. O’Riordan also shows how the “baking soda” hook of O.T. Genasis’ ‘Coco’ got users to whip out their phones and Shazam the song.

It’s a fascinating video and you can watch it below. After this it’s not hard to see why the company was recently valued at over $1 billion.