About a half dozen years ago I helped a friend of mine named Tim Westergren structure a new company he was forming called Savage Beast. Tim and I had been classmates at CCRMA, Stanford's Center for Computer Research in Music and Acoustics. Unlike me, however, Tim had stuck with music after graduation, playing in various bands and composing for a number of films (not surprising, frankly -- he was a way better musician than I was). But around 2000 Tim decided to combine his love of music and his appreciation for technology to create a better music discovery engine. Savage Beast was born.
Savage Beast was built upon the premise that if you could describe a piece of music in terms of its constituent parts, you could do a better job of finding music that was similar to that particular piece of music. Tim devised a way to categorize music (working with some Stanford music professors) that he called the Music Genome Project. The Music Genome is a set of about 400 characteristics that best describe the nature of a piece of music. As Tom Conrad, Savage Beast's CTO, describes it, the Music Genome captures standard things like tempo, key, etc. but also captures more subtle things like guitar picking style or "how much cowbell" (which, of course, you can never have enough of).
Building the Music Genome has not been cheap. It can't be done by computer. It has been done by dozens of musicians over a six year period, one song at a time. The secret sauce in the Music Genome is the combination of a clever schema for characterizing music and innumerable man hours by underemployed musicians. Tim knew that there were no shortcuts in building the genome and he has never tried to employ any.
Shortly after I began talking with Tim about his ideas back in 2000, Tim and his cofounders raised some angel money and began building the Genome and the infrastructure to put the Genome to use. They initially envisioned being the back end recommendation engine for music experiences. They set about selling their concept to online and offline music stores alike. But, as it always is, selling into huge organizations like Best Buy and AOL was a big battle and took longer than they anticipated.
In the early 2000s, Tim had to make some tough choices about Savage Beast. He had to let go a number of great employees. He had to hire others on a contract basis. And he had to do a pile of the leg work himself. Tim became CEO, Chief Evangelist, VP Biz Dev, VP of Product Management, whatever it took. He remained completely committed to his original vision and to the idea that the Music Genome provided huge value, particularly as it became more complete. So he marched forward with a skeleton crew in the company trying to close deals with revenue attached that could provide him with the necessary capital to continue building his company.
I should be clear here that I have never been an investor in Savage Beast. But I am a friend of Tim's and have had many conversations with him since he started the company about his business and the challenges of building a startup. He has made a pile of sacrifices along the way. He has made tough choices when he had to. And when he closed business that enabled him to ramp up his operations he always did so in a smart and careful way. Tim viewed each dollar as his last and put them all to really good use.
Over time Savage Beast managed to land a number of the big music store deals. They are the back end to AOL's and Barnes & Noble's online music experiences. They power kiosks in offline record stores like Tower, Best Buy and Borders. And on the strength of those deals, were able to raise nearly $9M in venture financing in late 2004 and attract a really strong consumer marketing oriented CEO. With that money, Tim was able to hire back a number of the great people he had let go and make real progress on filling out the Genome.
Most impressive, I believe, is that through it all Tim and Savage Beast have remained flexible enough to morph the direction of the company to meet the needs of the market, not some arbitrary vision of the ideal music search engine. And as a result of that flexibility, they have recently released what I believe is one of the most -- perhaps the most -- interesting music listening experiences on the web. It is called Pandora.
Pandora is a streaming radio experience in which each radio station is created based upon your own choice of song or group. So perhaps you are driving to work and the song "Underdog Victorious" by Jill Sobule is running through your head. You can go to Pandora and plug in Jill Sobule or Underdog Victorious and it will play music for you that is most like that particular song or that particular artist. (I've been listening to Jill Sobule Radio while writing this and it has been fantastic.) Because of the laborious work creating the Music Genome, the music played on the Jill Sobule radio station is often surprising but always nicely in keeping with Jill Sobule's work. While listening to Pandora, you'll find yourself either completely enthralled with what will come next or pleasantly oblivious to the precise songs making up the background soundtrack to your day -- either way, the experience will make you a big fan of the Pandora service. (For a more in depth discussion of Pandora, check out this profile on Mike Arrington's must-read Web 2.0 blog TechCrunch.)
While the jury is still out on how successful Pandora will ultimately be (apparently there is a bet between Mike Arrington/Robert Scoble and Steve Gillmor on this very question), one thing is clear to me, Tim Westergren is the poster child for startup persistence. He has grown his company. He has shrunk his company. He has grown his company again. And through it all he has been an energetic evangelist, pounding on doors, living in airports, and doing whatever it takes to keep his company alive. I wish him the best of luck and hope that Pandora is a huge success. Entrepreneurs with the drive and determination of Tim deserve to succeed.