The authors wonder how, in a situation of increasing use and access to streaming services, music preference research works. Today, when every listener leaves very detailed information about themselves - only using new forms of music distribution - it may not be necessary to study preferences. On the other hand, such research can answer the question "why" the consumer listens to a particular music, which will be a qualitative indication, more precise than the map created by the algorithm. By streaming, the user creates his or her own profile, often unconsciously, and the omnipotent algorithms transform it into appropriate music suggestions. They do this in a "diplomatic" way, making the listener feel like making autonomous choices. It is therefore worthwhile to look at the operation of the algorithms and the risks they pose.