predicting hits

11/14/2005 08:33:00 AM
It seems natural that with exponential increases in computing power that predictive databases and models would be built to examine the hit potential of music and movies.

Polyphonic, a Spanish company that is doing just that in the world of music. It has created a massive database of millions of songs and looked at the ones that became hits. From this data they developed
"hit clusters", groups of songs with similar characteristics. Artists, record companies and radio stations can use this service to make sure they have music that's most likely to be a commercial success.

The system seems a little scary for many people and provides further evidence of an industry driven trend to formulaic music. The other problem with the technology is that most of its data is historic and although it can be updated, most of the characteristics of hit songs will be based on old data. It also doesn't take into account artist appeal, a significant factor these days, as it focuses only on the song.

In addition, given the iPod, satellite radio and new eclectic stations like Jack FM, consumers might be moving away from the radio based hits driven music listening experience, to something that's far more eclectic and diverse.

Despite the potential pitfalls, it's likely that industry execs will be embracing this technology in droves as it provides a safety net of sorts, in the most fickle of all businesses.

This is just one example of a trend towards predictive databases and models that we will see gather speed as business harnesses: computing power, terabytes of data and behavioral insights into giant black boxes, that will predict the success of anything at the touch of a button.
Tags:

Comments
It appears you don't have Flash installed.
Email this article to a friend
Send an email to a friend with a link to this article. Items with an asterisk (*) are required.

Your Name:
*

Your Email:
*

Friend's Name:
*

Friend's Email:
*