Slides on skipping and listening patterns from study of Norwegian streaming service WiMP Music
(from presentations at by:Larm 2013 and 2014). See more results at www.cloudsandconcerts.com
Arnt
Ola
Xin
Johannes
Pål
Beathe
Marc
The following have been involved in the research project: Anne Danielsen (project leader, U of Oslo), Arnt Maasø (lead on steaming study, U of Oslo), Beathe Due, Johannes Bjelland, Kenth Engø-Monsen,Pål Roe Sundsøy, Wenche Nag (Researchers, Telenor Group), Yngvar Kjus (post doctor, U of Oslo), Anja Nylund Hagen (PhD, U of Oslo), Marika Lüders, (researcher, SINTEF), Research assistants Ola Løvholm, Xin Jian, Marc Casanovas, Linn Jakhelln, Gro Kirkeby, Erik N. Strutz, Hanne Tråsdahl, Ragnhild Toldnes, Inger Helseth, Helena Zarifa Pedersen and Ada Sandnes, and master students Henrik Sanne Kristensen and Magnus Indregard.
For more information on the streaming study, contact Arnt Maasø, ! University of Oslo
Prime time for discovery
(normalized)
Streaming Playlist creation
Skipping
Search Many skip and search when exploring music and creating playlists. And many skip and search during weekend parties.
06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01 02 03 04 05
24 hours
Data: 9 weeks 2012
New Artists Discovery Search! 7 %
Number of New Artists
From Search From Playlist
Playlist! 26 % Unknown Source! 67 %
More than 40% of ‘playlist discovery’ from editorial lists 6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 0
1
2
3
4
5
Figures here only show artists which are streamed for the first time by each user.
Time of day for streaming of new artists Data: 9 weeks 2011 and 2012
Mobile skipping ratio during 24 hours 0,7
Stationary
Mobile
Skipping ratio
0,6 0,5 0,4 0,3 0,2 0,1 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 0 1 2 3 4 5 Skipping hours 2013
Data: 9 weeks in 2013
Number of streams
PC/Mac without skips PC/Mac with skips
Mobile without skips Mobile with skips
album orienented mobile use
On a mobile it is more common to stream songs in the sequence they appear on in an album, compared to when listening on a desktop. At the same time, a lot of tracks are skipped within an album on mobile.
0 %
10 %
20 %
30 %
40 %
50 %
60 %
70 %
80 %
90 %
100 %
Percent of subsequent streams from the same album in 2013 Data: 9 weeks in 2013
Different tracks over 9 weeks Tracks streamed once per user Tracks streamed several times per user 30 % 25 %
14% or tracks! streamed more than once,! account for 72% ! of all streams
20 % 15 % 10 % 5% 0% Percent of tracks
Data: 9 weeks in 2013
Average streams per album track number for five genres
Number of streams per track
(normalized) Pop Rock Alternative Classical Jazz
Streams decrease with rising track numbers in an album. Pop has the larges fall in streams with rising track number, while classical music varies the least.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Album track numbers from 1 to 18
Data: 9 weeks in 2011
Variation per listening session Average
Median
8 7
7
6
6
5 4 3
3
3
2 1 0
1
1
Unique devices Unique artists
The average length of a listening session is 60 minutes. The median is 30 minutes.
Unique albums
*) A listening session is defined here as continuous streaming without a break of more than one hour.
Data: 9 weeks in 2012
Affordances of streaming media Affordances of the listening media and design of user interface are important to understand streaming modes (e.g., mobiltiy, session length, search, skipping, favorites, genre patterns). The tempo increases, in a way, for better or worse. In a way it’s great not to be tied to an album which turned out to suck, after you’ve spent 90 minutes listening to it. At the same time... you miss out on that [experience], ‘cause you only play the three most popular songs. (23 year old man, 2010)
Fast forward is a good thing. The skip button is a great invention.
(50 year old man, 2010)
I’m old fashion. I still like listening to whole albums. […] The biggest difference [with streaming] is that you might not listen to it again, whereas before you would have a CD collecting dust, which you later realized at a party or whatever, that «Fuck, did I buy this?! Why?! (38 year old man, 2012)
Quotes from focus group interviews by Clouds & Concers
Context and patterns of everyday life Playlist use and daytime patterns indicate individual ways of streaming. Yet, everyday habits influence listening greatly (as with most media).
If it’s raining, I might listen to jazz. It’s like, depending on the weather.
(23 year old man, 2011)
At our place – I live live in a shared flat with two other – there’s always music playing, really. We all have subscription for Spotify or WiMP or both. (24 year woman, 2012)
Quotes from focus group interviews by Clouds & Concers