It's the same cost/benefit calculation I point out below.
For services like spotify (and I'm specifically referring to the part of the service where you can choose to play a specific song), there is cost to an artist, in lost sales. Meaning, people who would normally buy a cd/track, and who do not since they listen to it on spotify. I think people can agree this cost is not zero - some sales definitely are lost - but they'll argue the magnitude of the cost.
There is also a benefit in terms of discovery. This is the streaming revenue, plus the sales that come from discovery. I think people can agree the benefit is above zero - discovery does happen - but they'll again argue magnitude.
The question is whether the discovery benefit outweighs the cost. There are widely varying opinions on this, and it can really only be answered with data rather than by someone's strong opinion.
It's also tricky question to test because I think the counterfactuals may be impossible. How do you simultaneously test the success of a song on Spotify, and the exact same song not on Spotify? Maybe an artist with two equally popular cds could put one on Spotify and not the other and see what happens for a rough approximation, but I'm not sure. I'd love to hear more thoughts on ways this calculation can be approximately measured.
Spotify isn't really optimized for discovery. Service like YouTube, music blogs, hypemachine, are much better for this and their UIs aren't designed to replace your music player either, so listening to music on YouTube isn't going to replace buying (or some other way of getting the music file) the song.
For services like spotify (and I'm specifically referring to the part of the service where you can choose to play a specific song), there is cost to an artist, in lost sales. Meaning, people who would normally buy a cd/track, and who do not since they listen to it on spotify. I think people can agree this cost is not zero - some sales definitely are lost - but they'll argue the magnitude of the cost.
There is also a benefit in terms of discovery. This is the streaming revenue, plus the sales that come from discovery. I think people can agree the benefit is above zero - discovery does happen - but they'll again argue magnitude.
The question is whether the discovery benefit outweighs the cost. There are widely varying opinions on this, and it can really only be answered with data rather than by someone's strong opinion.
It's also tricky question to test because I think the counterfactuals may be impossible. How do you simultaneously test the success of a song on Spotify, and the exact same song not on Spotify? Maybe an artist with two equally popular cds could put one on Spotify and not the other and see what happens for a rough approximation, but I'm not sure. I'd love to hear more thoughts on ways this calculation can be approximately measured.