Pandora, the Future of Radio, Part One: Interview with Founder, Tim Westergren

Interview with Co-Founder and Chief Officer of Pandora by June and Rodger Caldwell

Imagine a day in the life of the founder of an online radio company spending most of his time traveling the country meeting his users face to face. Not just paying lip service to listening to his users, I had the adventure of an interview with Tim Westergren by cell phone while Tim was driving around Athens, Georgia en route to a meet-up with Pandora listeners! This guy is the real deal.

Welcome to Pandora, simply the future of radio. Try it for awhile, and traditional radio will seem awkward, outdated? you will get impatient with it! Just log onto www.pandora.com, pop in the name of a musician or band you like, and you will get a customized flow of DJ sets based upon similar artists, each ?DJ set? focusing on a different quality of your favorite music. Pop in a few more bands, and your personal ?DJ? will incorporate them too. When you like something you key in a thumbs up, if you don?t you key in a thumbs down, so your DJ gets to know you better and better. Your DJ will continually throw in new obscure music you never would have imagined but that fits right in with what you like, as well as oldies that also fit right in. You can set up separate customized ?radio stations? just by grouping together a few bands you like, your personal Pandora DJ will do the rest for you.

This seamless, intuitive, customized experience is made possible by the ?Music Genome Project? which is about 40 musicians studying countless songs, artists, and bands and characterizing them by about 400 attributes, including the ?soul? of each song. They keep adding more and more music as guided by the almost 2 million listeners Pandora has attracted within 6 months without even advertising. I chatted with Tim about how it all happened, and what we can expect in the future.

June: Thanks a lot Tim for giving us a peak into the inner workings of the future of radio!

Why don?t we start out by having you describe in your own words what Pandora is about.

Tim: Pandora is about making it very easy for someone to listen to, enjoy and discover great music. The key there is that it?s very easy and you discover a lot of great music with it!

June: In setting up this interview, I came to realize you travel quite a bit! What are your travels about? Are you expanding the scope of Pandora?

Tim: I?m doing a bunch of things. I?m first meeting with Pandora listeners. We have meet-ups wherever I go. I had one last night in Atlanta, we are having one tonight in Athens, where we post an invitation for people to sit down with me & just talk about music & Pandora and what they like and don?t like to get firsthand feedback from people. Interacting with Pandora listeners from all over the place becomes a great information gathering opportunity for us. Wherever I am, that?s part of the anthropological research that informs the Music Genome project growth.

June: I?ve just been using Pandora for just a couple months and it?s expanded my music interests incredibly. You do indeed make it very easy.

Tim: That?s good. That?s really a big objective. You?re not alone in that. A lot of folks are finding new stuff. That?s entirely what we are trying to do.

June: It?s not what I say I think I like too. I like that about it. It plays what I like but didn?t know it.

What new features can we expect coming up?

Tim: I think one of the things that we are most excited about the pending inclusion of Latin music.

June: That really opens a lot of new doors.

Tim: That?s really just the beginning. Our objective with the Music Genome project is to make a real universal collection so that it encompasses many different languages & cultures and can even help people jump across cultures.

June: Ahhh! That?s interesting, so if you?re listening to one type of music, another type of music from a different culture that has the same qualities might start popping on?

Tim: When we analyze music we don?t actually bucket music into any kind of arbitrary genres or arbitrary boundaries. So if we have a rock band from Japan that sounds like Counting Crows then you?re going to hear it.

June: So the Music Genome project, was that created for the purpose of creating this type of radio or was it created for another purpose?

Tim: No, I had no idea what it would become.

June: What was the birth of the Genome project, what was the goal of that?

Tim: Originally I thought it was going to be just a recommendation tool. That sites would use it to sell music so if you?re a music retailer that someone could be shopping on your site for music we would be really good at recommending other albums or artists that you should buy based on what you were purchasing. It would be a recommendation service. We actually pursued that business for quite awhile.

June: So how did it meander its path to Pandora?

Tim: Umm?quite circuitously. Our initial plan was to license it to places. We licensed it to AOL, Tower Records, Borders, Barnes & Nobles, and Best Buy, We were sort of building an interesting but small business model based on that. Then we moved into the retail space for Best Buy we did some in store kiosks for them, but what happened was that it wasn?t turning out to be a very healthy business for us. It?s very hard to make a profitable business when you?re beholden to a small number of very large customers. We kind of chased that for a while and then we were struggling as a business.

We were out of money, trying to get financing and we really had our heads down in the weeds. Then in 2004 we were able to raise a round of venture financing and what that allowed us to do was stop and take a breath for a second and consider where we should really be taking this thing. What happed in those intervening years were pretty significant things.

Broadband had become a much more ubiquitous phenomenon. Music on the Internet requires a nice healthy amount of pipe and doesn?t work so well on dial-up. Once that had flipped over & broadband had become much more ubiquitous online audio suddenly looked much more like a mass-market product than it had before. What also happened was a sort of clearing up of the licensing landscape. For a long time there was a lot of uncertainty about what licenses were going to look like from the record companies for online web casters. That had kind of cleared up so we knew what we were getting into. So lo and behold, the thing we had built in all of that time, the Music Genome was perfectly suited to be applied to the task of building great playlists! So we went for it.

June: One step suggested the next step, it sound like.

Tim: Right the path kind of revealed itself to us in a way.

June: Sort of like Pandora!

Tim: Exactly, but it took a lot longer.

June: The Pandora space is just so friendly and intuitive. It?s not so much what you say you like or even think you like it lets the music & your guys (Pandora?s) research create & direct what you hear. Its feels so relaxed and intuitive. Is that by design?

Tim: Oh yeah, Absolutely. One of our mantras when we did his project was that the interface would be simple. The problem that we thought needed solving was that you needed something that would make it easy, quick and painless for people. We were determined to master to that objective. That?s what we were all about. Can you hang on for one second?

Tim asking stranger on the street from his car: I?m looking for the Blind Pig Tavern.

Stranger: Right across the street.

Tim: And is there a record store there too?

Stranger: Record store?

Tim: The best independent record store in town. Mixed Lots or Mix up or???

Stranger: Umm?well there?s One Street Records downtown.

Tim: That?s it!!

Stranger: One Street Records its close about a mile down Thomas. You?ll have to find somewhere to park and it?s in the middle of downtown.

Tim:Hey, thanks a lot! Sorry about that June!

June: No that was great, you?re in the street. Looking for The Blind Pig Tavern? Now we know you?re out there doing the work yourself. That?s proof your not just sitting in your office saying ?Oh yeah?we talk to people all the time.? So, back to the interview!

June: I?ve read that you have 30 people analyzing music.

Tim: We have 40 now.

June: 40 now, but still analyzing all the music in the world. When I do the math it just seems to be impossible!

Tim: It is impossible to analyze all the music in the world and we knew that. We had a choice, which was, you can build a system that can handle millions of songs and do it very quickly and be very scalable. But to do that you have to take shortcuts. You can?t do it in the deliberate detailed way that we are or you can do it deliberately and with detail and not you cannot get the same amount. Our theory is that when it comes to music the problem that needs to be solved is not to make 15 million songs easy to navigate but to make the smaller subset of the best music phenomenally easy to navigate.

In North America there are about 5 million songs in some form of commercial existence and we have about 400,000 of them so let?s say we have about 10% of that for arguments sake. I think most people you ask about Pandora they would say it has a massive catalog just from using it. It?s kind of hard to stump it. I think that tells you that our theory is right. In reality the amount of music that enters the public consciousness and the amount of music you need to give people all of their starting points & allow a big amount of discovery is not as big as people think. Once you have a couple million people using the service they can guide you.

June: Is there a Pandora Music Festival or Pandora label on the horizon?

Tim: I would love to put on a festival, but I don?t imagine us becoming a label. We don?t want to be in a position where there?s the perception among our listeners that we have a bias towards a particular catalog. I think trust these days is very difficult to get, very easy to lose.

June: How is co-founding a new company like giving birth?

Tim: Not a bad comparison. Well it?s grueling, certainly. Physically, mentally it?s extremely demanding and it takes an extreme amount of tenacity to see a company all the way through. You have to be very resourceful & very tough. People ask me what are the most important attributes to have or what makes for a successful start-up. I think tenacity is the most important part of that. You have to will your company to succeed. I think one way it does resemble childbirth is that you need a partner. I think that anyone wanting to launch it company is well advised to found it with somebody. It?s very hard to go through the life cycle of a startup without someone to share your burdens, your worries. My first piece of advice to someone starting a company is don?t do it alone.

June: That?s a good point. A lot of times we see the finished product and think you just woke up some day and magically it just fell into place. Never seeing what went into it.

Tim: I guess you also go through a lot of change. You go through these changes & you have to be very adaptive & willing to embrace change.

June: And then when it?s born you have your child with a mind of it own to deal with. You can?t just walk away from it.

Tim: Then it learns to crawl and that?s all she wrote!