Archive for the ‘Nielsen’ Category

Confessions of a music nerd

Friday, July 24th, 2009

I love music and I love data. I can strum my air guitar like Jimmy Page. But I can also run regression models with the best MIT nerds. As the Who said, I can’t explain, but the two are inextricably linked. And now I have a new outlet for my musical nerdiness. As of yesterday, Billboard – and its mother company Nielsen — gives me access to its historical chart data.

Mom, I’m in hog heaven. They have a thing called the Visualizer. I can see how many weeks the Black Eyed Peas have been number 1 in the Hot 100. I can organize my very own Linkin Park vs. Red Hot Chili Peppers smackdown. Beyond the sheer pleasure of wallowing in the data, Nielsen also allows me to contextualize that data to my biography. In a feature called Soundtracks of Your Life, the site asks for a few meaningful dates in my existence and tells me what was top of the charts at the time. When I entered my birth year (1954), I was told the charts didn’t go quite that far back into Cro-Magnon territory (1958 is the first year). So I pretended to be my younger sister, but it got a little too personal when she was asked about her first kiss. I can also bet on who’ll be going up or down in next week’s chart and win some free downloads. Now, I can analyze how quickly Incubus typically rises in the charts, or predict sales of Lady Gaga as a function of her radio airplay. I’m now a gambler, on top of a music nerd. I’m tuning my R square and turning on my Chi square coefficient. How much fun can one have?

There’s lots of us out there, believe it or not. For years, we’ve wanted to peel the data off the magazine or the web site, but were never allowed to do so. Until yesterday, Billboard had only made its data accessible to professional record labels and music retailers willing to pay a king’s ransom every month. The professionals do very little with that data, it turns out, except to look at sales and show artists that they’re not as hot stuff as they think they are. I know. I worked for several of them over the years.

Music companies are not exactly stoked with analytical people. They hold the strange view that discovering the next Beyoncé is more important than devising a predictive algorithm showing when to stop investing in promotion for a particular artist. But they should do both. And we, analytical nerds, can now do some of that for them for free now that the data is up for grabs. Welcome to analytical co-creation.

The serious story here is that the best way to generate insights from large databases is not to restrict its access to a few experts, but to open it to a large number of passionate people who care about it. Billboard will undoubtedly unleash a whole eco-system of analytical nerds trying to predict artist and song behaviors. As for me, please no calls this summer. I’m testing a proprietary song-climbing pattern showing it’s time for a re-release of Pink Floyd’s greatest hits. The track most likely to be a revival hit: “Money.”