There's an incredible focus on these two points -- targeting traditional radio listening and advertisers -- at Pandora. People on the outside who refuse to consult the company completely ignore this. They discount Pandora's dominant share of Internet radio listening (more than 75%), increasing share of total radio listening (now at over 7%), surging mobile listener hours and monetization and a growing local advertising infrastructure.
Nobody has come in or can come in, not even Apple, and erase this with one fell swoop.
The comparisons between Pandora and Spotify need to, at the very least, slow down. They're not as meaningful as most people make them out to be.
On-demand has been around forever. There's a time and a place for it. Always has been. Always will be. Same goes for radio. It's not a matter of one or the other.
Spotify has a multi-million songs library. That's great. You know the song you want to hear and chances are you'll hear it. It's not all that different from firing up iTunes or digging into your personal physical music collection.
At last check, Pandora has a catalog of around 1,000,000 songs. The typical surface scratch puts one in the win category for Spotify, but again, that's a thin, misleading and ultimately incorrect analysis.
When you go to Pandora, you start your radio station with a seed. Pandora takes the attributes of that seed and builds your personalized station. It's a fluid process. I have seen the Music Genome Project live and in action; it's an incredibly deep and complicated, human-organized system.
Type in "When You Were Young" by The Killers (I'm going to see them Friday night at Madison Square Garden!) and Pandora fires up a great live version of Coldplay's "Clocks" followed by MGMT's "Kids," which I like, but could thumb down to get "Starlight" by Muse.
You don't search a song on Pandora because you want to hear it -- you go to iTunes for that. You search a song on Pandora because you want it to curate a radio station you will like. One that's not a cookie cutter effort from