On June 29, Pandora Internet Radio founder Tim Westergren treated Richmonders to a meeting at Virginia Commonwealth University’s Grace Street Theater to discuss the radio’s future.
The meeting allowed attendees to ask Westergren questions about anything Pandora, from the company’s humble beginnings to suggestions on how to improve ease of use on the radio website. These “town hall” meetings started in Austin, Texas with only six people; now over 200 meetings later from Biloxi, Mississippi to Boston, Pandora’s town halls have become so popular that entire rooms and theaters fill up with listeners eager to hear what’s in store for Pandora.
“For us they’ve become this really invaluable way to just talk directly with Pandora listeners,” says Westergren. “I am really interested in hearing your [the listeners’] thoughts, questions you have, criticisms you have, Pandora ideas you have for us.” Pandora was originally launched in January 2000 as a music genome project, the most comprehensive analysis of music ever attempted.
“What we do is understand songs musicologically, and then by marrying that with some mathematics, kind of build this tool that connects songs based on their musical content,” explains Westergren of the project.
Pandora’s team of 50 musician-analysts have been listening to music one song at a time, dissecting and analyzing every single detail of each song. “We have this musical taxonomy. It’s a list of about 400 musical attributes that sort of collectively describe the songs,” says Westergren. “[What we do is] essentially dissect melody and harmony and rhythm and vocal performance down to the basic building blocks.”
Westergren’s original music genome idea failed initially, but that didn’t stop him from using his project with a different purpose. “When I got out of school I knew I wanted to be involved with music,” reflects Westergren, a Political Science graduate from Stanford. After working as a “manny”—a male nanny—for five years after graduating, Westergren became a working musician, writing music and traveling across the country to perform.
“I became pretty intimately acquainted with the challenges of being a working musician,” he says. “There’s one really hard problem for musicians, which is, how do you get exposure? Only a small number of artists get on broadcast radio. And radio really is a key to professional longevity. So we took this huge database we had built, [the] music genome project, and decided to turn it into a radio.” Thus the birth of Pandora Internet Radio in November of 2005.
Pandora’s team of musician-analysts still dissects songs based on 400 musical attributes as they had when the company was first launched. The goal behind this extensive analysis now, however, is to create playlists based on musical proximity for its 30 million registered listeners, and its constantly growing number of unregistered listeners.
Westergren believes that the beauty of Pandora and the reason that it’s successful lies in the method of catering to its listeners based on musicological proximity and not on popularity.
“It’s not a popularity contest,” he says. “It’s one of the only systems that, in terms of how it picks songs, is blind to popularity. So when you type in an artist or song, Pandora doesn’t know if that artist [or song] is famous or not, so it really is a truly level playing field.”
Listeners at the Richmond meeting all expressed the desire for more granular feedback when rating songs picked for them by Pandora. Instead of just giving a song or artist a simple “thumbs up” or “thumbs down,” which is what they can do now, listeners would like to say why they gave a certain rating, saying things like “I like the vocal harmony on this song, I don’t like this singer, I like the melody,” and so on.
This idea of providing granular feedback is something that Pandora plans to do in the future for its listeners. There are also new technological applications for Pandora in the works, such as car accessories in conjunction with the iPhone, as well as incorporating Pandora into DVD and Blue Ray players with Bose.
“I realized, wow this is what the world is going to be like,” says Westergren about these advancements. But changes for listeners aren’t the only goals of Pandora. “We’ve gotten pretty ambitious with what we think Pandora can do,” Westergren says. “We’re intent not just on redefining radio from top to bottom for listeners, but also to really make an impact for musicians.”
Westergren has received a lot of feedback from artists who have debuted on Pandora and have become successful from this platform. “Artists will say things like ‘My iTunes sales have spiked,’ and it corresponds exactly to when they were added to Pandora,” he says. “Or they’re at shows and people are saying they heard about them on Pandora—that’s happening a lot. So I think we’re beginning to see the ripple effect of it. I can’t say that this musician attributes his or her success to Pandora, but I think we’re kind of creating a bit of rising tide.”
Because of this feedback that he has already received, Westergren only sees continued success for musicians in Pandora’s future. “Our goal is to create a musician’s middle class. I will be happy if someday Pandora will have reached the point where you graduate from college and you say to your mom, ‘Mom, I want to be in a rock band.’ And she says ‘That’s a great career.’ We’ll have laid the foundation for a viable profession, which I believe is there. The audience is there for it, and certainly the talent is there for it.”
The Greater Richmond Grid, Volume 1, Issue 1, Summer 2009
