Janko Roettgers notes that Pandora users are slightly more likely to report that owning music is important to them, and then argues:
The company didn’t make any data available about people who pay for a streaming subscription, but I wouldn’t be surprised to see significantly lower interest in music ownership amongst users who pay for unlimited access.
That’s why it’s smart for Apple to invest in iRadio. The goal is not to kill Pandora, but to actually bring that type of radio service to more users, and keep them from switching to a full-blown access model. In other words: It’s not about Pandora, and all about Spotify.
I’m skeptical that the data Roettgers is using, which comes from the NPD Group, means much of anything in this case; while streaming service users may have much lower interest in owning music, I doubt that Pandora users would report higher interest because they use Pandora. Instead, it may be self-selecting—people who care less about owning music (or care more about selection) might choose streaming services, whereas everyone else will use non-random access services like Pandora and purchase songs and albums they really like.
That said, this sort of service could help compete with Spotify, Rdio et al., but not because it will provide similar functionality to them. Rather, it could make iTunes more useful. iTunes’ Genius auto genre and playlist-making features are actually incredibly good, so using it across iTunes’ entire library of music could be terrific. Embedding that in iTunes, iPhone and iPad would be great, and would give people more reason to stay on the iTunes platform.
What it won’t do, though, is convince people who highly value unlimited access to replace Spotify or Rdio with iTunes or to stick with iTunes over those services. They’re simply not substitutes.