Russ: This is the BusinessMakers Show heard here and seen online at theBusinessMakers.com. And now it is time for the BusinessMakers Flashback, and staying in tune with that theme of the new world of radio, we are going to revisit an interview of back in 2007 when we had Tim Westergren, founder and CTO of Pandora, on the show. We entered the discussion where I had asked Tim to tell us about Pandora.

Tim: Well, it's a personalized radio service, so it allows people to quickly and easily create radio stations that play just music they like. It's currently largely used online, but it's beginning to spread offline as well. And it's based on something called the Music Genome Project, which is this enormous sort of musicological study that we have been working on for the last eight years that allows us to connect songs musicologically and allow for really sort of efficient personalization.

Russ: Okay. So obviously, we needed to know more about this Musical Genome Project.

Tim: So the Music Genome is a musical taxomity. So it's kind of like musical DNA. It's a list of literally close to 400 musical attributes that collectively describe a song. So it's all the details of melody and harmony and rhythm and instrumentation and vocal performance, etc. And we have a team of musicians, about 50 of them, that for the last 8 years also now have been listening to songs one at a time and scoring, sort of analyzing every song along each one of these attributes, one by one, sort of collecting musical fingerprints.

And when you start a station on Pandora by typing in a song, for example, what happens is we look at that song's musical DNA and go searching for its musical neighbors in this database that now numbers over half a million and use that to create stations.

Russ: Okay. And stations mean that you can create multiple stations. And, you know, a station would sort of synch up with that first initial input a song or a group?

Tim: Right. So the most basic use is type in a song, sit back and let it run, and it'll explore that musical neighborhood, whether you start with a swing tune, a folk tune, a blues tune, you name it. But if you'd like, you can actually lean in and interact with the station. And that's, of course, one of the great benefits of Web access is the two-way conversation. So you can actually start giving songs feedback, thumb up and thumb down. And as you do that, what happens is that station hones its musical sort of definition based on your musicological input. So it's translating those thumb up and thumbs down into a musicological feedback.

Russ: And getting smarter about -

Tim: Exactly.

Russ: - what you want to hear.

Tim: Yeah. I mean, the idea is really to understand what you like about the input. So if you're a Rolling Stones fan, what is it that you like about them? Is it the more rocking stuff or is it, you know, Ruby Tuesday? You know, what is the style of theirs that really appeals to you? Now, as you use your thumbs, you'll get more and more homed in on it.

Russ: Okay. After hearing this real cool musical site, I wanted to know how in the world Tim made money off of the operation.

Tim: Well, there's a complicated answer to that. I mean, we've tried many, many different ways over the years. You know, we are eight years old as a company, and we've been through several different business plans, including licensing the GM to other companies. We're now, of course, business to consumer and have been for the last couple years. And our business is all abut advertising, so we support it with visual advertising and are able to do that because it's such an interactive experience.

Russ: And so does that mean that it's free to all users?

Tim: It is indeed. You can pay for it if you want to get rid of the advertising, but it's primarily an ad-supported, free service.

Russ: So next, I thought it was real interesting that Pandora is now really getting traction, but it's actually already eight years old. So I ask him to tell me about that.

Tim: So we launched it in January of 2000, which right there says a lot.

Russ: Yeah, that was an interesting time.

Tim: In March of that year, raised our first venture capital. And the idea of the company back then was we would be a technology licensor, so we would build the Music Genome and deliver it as a recommendation tool to other companies. And we chased that business model for about four years, and we followed the entire arc of the dot com bust, which we basically hit right square on.

We work with portals, we work with online retailers, then we moved to brick and mortar stores and delivered it through Web-enabled kiosks for a couple years, just trying to sort of scratch and claw our way through the sort of desert of those few years.

It wasn't until 2004 what we sort of turned our attention to radio. And what we really did was repurpose the musical asset that we have for a different application. And at that point, after we had spent, you know, a couple years not getting paid, not paying salaries and going through all of the sort of dot com collapse adventure -

Russ: So then I wanted Tim to tell us a little bit more about this licensing approach and the challenges of getting this model right.

Tim: The trouble was the companies we were working for, we were licensing to were all going through struggles of their own because the music industry was and continues to be in turmoil. So I wouldn't say that we were anybody's darling. We had a couple key deals, one with AOL and one with Best Buy that really kept us afloat during that time though.

Russ: And since you've transitioned to this new way of delivering your product, did you discontinue those relationships?

Tim: We did. We put them to bed right around when Pandora was launching.

Russ: Okay. And Pandora was launched in 2004?

Tim: 2005. So we - what actually happened was in early '04, we raised our first large round of institutional investment, which was the first time we had sort of the latitude to think more than just two weeks ahead.

Russ: Meaning that prior to then, it was really kind of angel and friends and family kind of investing then?

Tim: Yeah. We were really truly broke. We weren't paying salaries. We faced frequent eviction notices. We had the whole - you know, we were against the wall.

Russ: A textbook startup.

Tim: Yeah. And we were able to coax a bridge investment here and there to keep the lights on, but '04 when we raised that money, we paid back an enormous amount of back salary and then had a - sort of a chance to look forward strategically for the first time.

Russ: And was it easy for you to raise the venture capital?

Tim: Well, put it this way. The yes that I got from Walden VC who invested in '04 was my 348 [Laughter]. Yeah. It was insane.

Russ: And lastly, I wanted Tim to tell us a little bit more about the company. It seems like the staff there is probably a pretty unique staff. So I said, how many employees do you have these days?

Tim: We have about 115 and of that 115, there are 50 musicians that comprise the sort of music analyst group that are doing the Music Genome work.

Russ: It almost seems to me like it would become a very appealing place for a serious musician to want to go and work.

Tim: Well, this is a job I would have taken in a heartbeat back when I was in a band. I can tell you that. I mean, it's a perfect job for a working musician. They work four to five hours a day, they have complete flexibility. They can come and go if they have to go on tour. They get full health insurance. I mean, it is the dream job for a working musician.

Russ: Okay and that concludes our revisit with the interview of Tim Westergren, founder and CTO of Pandora. And that wraps up this mornings BusinessMakers Flashback. You've been listening to the BusinessMakers Show, heard here and seen online at theBusinessMakers.com.