In 2018, Spotify registered a patent through which could use information from user conversations and background noise to determine different patterns. Based on the analysis of this information, Spotify could recommend specific music, podcast or even adjust the advertising and announcements that users would hear.
The patent was approved on January 12, 2021 and has really attracted attention in all the media. Going into the detail of the patent content, this shows potential use cases for this unique technology. The most interesting part is in the possibility that Spotify would have to obtain the intonation, stress, rhythm and other aspects of the user's speech to determine from them their emotional state, gender, age and even accent according to Pitchfork.
Implementing this technology could allow Spotify to improve its advertising system and serve ads that are more relevant to users and their current situation. The patent also mentions the possibility of even differentiating the user's environment, being able to identify even if he is alone or with a group of people. Obtaining this voice information would be done through the user's metadata, not through its direct recording, which would go through a Markov model to obtain the conclusions.
It is not yet certain that Spotify will implement all the technology described in the patent. We already know that technology companies usually register many patents that later do not materialize (Apple among them and very frequently). When the Pitchfork outlet contacted Spotify to comment on the patent, they commented on the following:
Spotify has applied for patents for hundreds of inventions, and we regularly complete new applications. Some of these patents we incorporate into future products while others never see the light of day. Our ambition is to create the best audio experience, but we don't have any news that we can give you right now.
Without a doubt, a patent that will give a lot to talk about if Spotify finally decides to implement this technology. Despite "not recording" the user directly, the privacy of users and the collection of their data is again in question and, as almost always, so that a company can profit at its expense.
