User Behavior: Spotify

Spotify premium allows users to pick exactly what song they want to listen to whenever they want to listen to it. If you chose to download the song you can even listen to it when you’re not on WiFi without using data. In this blog post I will analyze scholarly articles that researched the understanding of users behavior when using Spotify (Zhang, Kreitz, Isaksson, Ubillos).

During the time of the study the premium feature allowed researchers to study long-term user behavior as well as making comparative studies of user behavior on desktop and mobile devices. A premium account was required to do any research on a smartphone client. The data from this study was collected directly by Spotify from 2010-2011.

pic2

From research it was observed that there is a strong daily pattern and significant variation of hourly arrival rates. The session arrival rate is lowest around 2 a.m. and increases sharply up until 9-10 a.m., which the researchers define as the morning peak. After the morning peak there is a drop during the mid-day lunch break time during weekdays. After the lunch break the rate increases again and the daily peak is reached at roughly 6-7 p.m., which they call the evening peak.

It was found that the morning peak of mobile session during weekdays is often one hour ahead of desktop sessions, and it’s believed to be because people are using the Spotify mobile app while commuting to work and then once arriving to work using Spotify on their desktop. This data is observed for the weekdays, and the weekend effect explains that both the morning peak and lunch break dip disappear and the commuting effect also disappears since it is the weekend, when people typically don’t have work.

pic1

This information is important in understanding people’s behaviors about when they listen to music through out the day. The researchers believe many users launch Spotify to have background music playing at work, and this is why there is a morning peak of session arrivals, which match up with the session length. They know this is valid because during the weekends the morning peak session length disappears. They found that mobile sessions are much shorter than desktop sessions. The usage pattern between desktop and mobile sessions is dramatically different.

Spotify users favorite time of day to use the service is from noon to midnight on any given day. For mobile users the favorite time of day is from 3 p.m. to 6 p.m. and for desktop users it’s 6 p.m. to 9 p.m. They believe this is because many Spotify users tend to use desktops rather than mobile devices after they arrive home from work. I wonder if this would still be true today, 5 years later, with the use of mobile devices spiking. It would be interested to see this data tested from 2016-2017.

Pic3.jpg

Unfortunately there is not much else known about behavioral patterns for Spotify users. Since Spotify is still relatively new little is known about the behavioral patterns of its users, or in other music streaming systems.

The researchers concluded that there were session arrival, session length, and playback patterns that were exhibited daily. It’s safe to say that people who have Spotify premium (paying for it) tend to use it every day of the week in a pattern. Users also have a favorite time of day in which they like to access Spotify. Overall these findings help expand our understanding of user behavior in Spotify, and stipulate new insights of user behavior in other music streaming services.

To my best knowledge there have been no studies on the impacts on society by Spotify or of music-streaming services. The effects found by these researchers were neither positive nor negative but instead just quantitative results of users listening behaviors.

All information found from: http://www.ds.ewi.tudelft.nl/fileadmin/pds/reports/2013/PDS-2013-001.pdf