For better or for worse, the music industry has a direct line into many listeners’ preferences, likes, dislikes and general listening behavior. I wonder if the lure of that data is a driver in the survival of sites like last.fm and Pandora (my two faves). Even blip.fm provides a richer view into playlists with every listen enhanced with digital words from the listener.
The holy grail for music promoters, lawyers, etc. is the ability to pick a hit. There have been plenty of data-driven attempts at predicting hits, but they have all focused on the aspects of the song. Today companies like CBS (owners of last.fm) are no doubt beginning to examine the behavior of the actual consumers.
What can we infer from the information below?
It appears that, while lady GaGa is extremely popular, she is very likely a guilty pleasure amongst last.fm listeners. They don’t want to share with the world their affinity for the GaGa.
Is her star power on the decline? Is it time to look for the next GaGa? Perhaps if we knew a little more about the profiles of the listeners, we could further understand this behavior. For what percentage of listeners are the songs unwanted?
Why does all this data exist?
Because smart and ambitious people build engaging and relevant experiences that suck people into the world of digital streaming. Even if they are not streaming, the promise of a little bit of personal insight is enough to “scrobble” their offline listening habits. Despite the kicking and screaming (and suing) we’ve heard about, these sites have blazed a trail that may save the butts of the executive branch of the music industry.
You’re welcome for the free market research data. Please please use it responsibly. Maybe even use it to increase the quality of the product you release. That would be stellar. Instead of digging through IP addresses of people who have illegally downloaded music, you can focus on these pearls.
Oh, and maybe even give us some of that data in return. The last.fm playground has myriad playful visualizations of listening data. That’s cool, but sharing the data itself would be great too.