A Pandora’s Box of Start-up Expertise

by Dru Wynings. Average Reading Time: almost 49 minutes.


Thank you so much, Tina. Thank you for inviting me. Originally, our CEO, Joe Kennedy, had been scheduled to be your guest today, and he sends his regrets. And I’m going to do my best to fill his shoes. When Tina and I talked a little bit about kind of what we might do here today, we went a little bit through the history that she just enumerated. And we thought maybe it would be interesting to talk a little bit about all the things that I’ve done over the years. And so it’s been fun for me preparing a little bit. I’ve been trying to distill down some of the stories from all of those places and try to understand what it means. And the first thing that amazed is that I’m actually in my 20th year of working in technology. I got my start in 1991 working at Apple on the Macintosh. And for 15 of those years, I’ve been involved in what you might call a start-up. So I’ve really made a career out of working for start-ups. I’m a product guy, which means that all of the jobs I’ve held have somehow been related to designing, engineering, delivering the solution that the company that I worked with was providing. And that’s really kind of fundamental to my DNA, so all of the stories that I’ll share are kind of from that perspective. Over those 20 years I’ve worked for six really interesting companies and hundreds of brilliant people, and so I thought I’d just share some of the lessons that I’ve learned. And as I went through that process, I realized that there were sort of seven key lessons which, I guess, is slightly less than three a year, which seems kind of disappointing on balance, but… So I thought we’d spend a little time in talking about those lessons and use them as a tool for a little bit of story-telling about my background. One thing as I talk that I’m sure as you have people come through and talk to you about entrepreneurship in general, many of the people that come and stand before you are founders. Well, I’ve never been a founder, so I’m going to tell lots and lots of stories about small companies and the things that they do, but it will always be from the frame of an employee. And oftentimes I was an early employee. I’ve been fortunate enough to have leadership roles at many of the companies, but I’ve never done the founder thing. And I think, you know, I’ve had an unbelievably fun and rewarding career and I think sometimes we forget that you don’t have to be the person who kicks something off to be able to have a really exciting experience. So, to start here, I think — as I thought about all of the work that I’ve done over the years, by far the most important thing that I’ve learned is that I succeed when I’m passionate about the company, the product, some aspect of what it is that I do everyday. And for me, this really all started when I was…well, I was 15 years old when the original Macintosh came out in 1984. And I remember like it was yesterday, watching the Super Bowl and seeing that 1984 ad with the runner come on, and kind of being teased by this promise that Apple is about to do something that would make 1984 not like 1984. And as soon as they started to tell the world a little bit about what they intended with the Macintosh, I was rapt. I mean, I spent the entire next year poring over every piece of documentation, and I still have these tattered product brochures from that era that I looked at endlessly. And that really began my love affair, not just with computers and the whole world of technology, but really specifically with Apple Computer. And it’s not at all an exaggeration to say that I went to college so I could work for Apple. And so I went to the University of Michigan, studied Computer Engineering, and focused all of my energy on trying to get myself into a place where I could be an intern at Apple after my junior year, which is about the time that they started taking interns. And the summer before I went to Apple, another guy from the — not from the Engineering College, but from the Computer Science program, and I was a Computer Engineer and there was a bit of like snobbism about Computer Science versus Computer Engineering at Michigan, and this guy ran the computing lab for the Computer Science program and I was one of the geeks that ran the computing lab for the Computer Engineering program. And so there was a real rivalry there. Nerd fight kind of thing. And I remember thinking, “Oh, here’s this guy and he got this internship at Apple a year ahead of me. Oh, he’s not that impressive. He’s just this Computer Science guy,” and blah blah blah. Well, turns out that that guy was Tony Fadell, who later invented the iPod. So be careful of who you’re dismissive of. But fortunately enough for me, the next year I sent a letter off to Apple saying that I really wanted to be an intern. And I remember this letter. It said something like, “I wanted to change the world through technology.” So I always wore my passion for these kinds of solutions on my sleeve, and frankly, I think I got the job that summer because of this goofy, over-the-top letter that I sent. Not even email. It was an actual typewritten letter. And I had a great experience at Apple, which we’ll talk about a little bit later. But it was, in large part, a wonderful experience because I so loved the product and I so loved the company and I was so passionate about the entire thing. And I tell you, I’ve done lots of things over the years, different kinds of companies and different kinds of product categories. Some have been successful, some have been catastrophic. But the ones that I look back on with the most enthusiasm, I have the best memories from, I’m the most glad that I was involved in, are the ones that I was really passionate about. And one example is Pandora. So the company that I work for today is the internet music service, internet radio service, Pandora. And I’ve been there for the last six years, so, a pretty big chunk of that 20 years, I guess. And the story that led me to Pandora was that I had in 2001, after the bubble had burst — we’ll get to talk about that, too in a little bit — I was looking for work and there wasn’t anything in consumer software. There was no new internet companies. It was kind of nuclear winter, if you remember that period. It was incredibly difficult to find any kind of work. And I found this enterprise software company that was just getting going. It was founded by the guys who had started WebLogic, and so they had this great pedigree around creating enterprise solutions. And I had spent three-and-a-half years there. And as much as I loved the people, I wasn’t terribly passionate about the product. It was, you know, a harbor in a storm for me while the entire industry kind of redesigned itself. And towards the end of that period, I started thinking about wanting to get back into doing something for consumers, something that really connected with my own personal passions. And I started thinking about, like, what are the problems in my life that I’d like to see solved with technology? And I had just gone through the process of ripping my CD collection. I have like a thousand CDs. I have been passionate about music since I was a little kid. I have always enjoyed nothing more than getting someone in front of my stereo and playing music for them. And I started thinking about, ‘Well, gosh, there’s this opportunity now that we’ve kind of encoded all of our music digitally to do interesting things with it. To make sure that people know about new music when it’s released, or introduce people to other people who share similar musical interests, or find new music that you will love based on the things that you’re listening to today.” And so I started thinking about these ideas, these ideas of putting together some kind of a music destination on the Web that would make recommendations and connect people together. And in talking about this a little bit, I got introduced to this little company in Oakland called Savage Beast. The company had about eight full-time employees and they were looking for…well, the job description on the website when I met them said, “Engineering Manager.” By the time that I arrived for the interview, they have decided to upgrade that to VP of Engineering because I think they were looking at my resume saying, “Well, if we’re going to hire this guy, he probably, given his background, wants to be a vice-president of our little eight-person start-up.” And as much as I appreciated the upgrade in the title, the reality was they needed an engineering manager. They needed somebody to hire a bunch of engineers and to manage its schedule and to hit the date and repeat over and over and over again. They had a pair of brilliant founders, including somebody who was a technologist and had a great technology vision. And I got the job offer from them…and turned them down. And I turned them down because it didn’t feel like the right career opportunity for me. I didn’t really want to be the engineering manager. I wanted to have a much larger kind of purview with respect to the product. And I didn’t want to work for — I wanted to work for a company that was doing something for consumers, and Savage Beast was making just recommendation infrastructure for other brands. And so it was like a B2B solutions. They’d sell this recommendation engine to, like, Best Buy, and Best Buy would use it in kiosks in their stores. And so I thought, ‘Well, gosh, it’s not the right kind of product exactly. It’s not the right kind of career move,’ and so I turned them down. But a funny thing happened. Every day, I thought about this little company and this great group of people and this opportunity to use the start that they had, which was something they called the Music Genome Project. And the Music Genome Project is this magic thing that they’d been developing for four years. And the idea is that they have a team of musicians who came to the office everyday, listened to music one song at a time and analyzed each piece of music along as many as 400 different musicological dimensions. So not just, are there guitars, but what kind of guitars, how are the guitars played, what role does the guitar play in the entire kind of overall mechanics of the song. And by doing that, the company ended up with a kind of fingerprint for every song that flowed through this process. And they had been doing it for years and so there were tens of thousands of songs, now almost a million songs, that have gone through this process. And I was so fascinated by the potential for that, and I just couldn’t sleep at night thinking about how to tie together some of the ideas that I had about this consumer service with this incredible technique of the Music Genome Project. About a month passes and they call me up and they say, again, “You know, we talked to some other people and we think really you might be the right guy for the job.” And I said, “That’s funny. I can’t sleep at night because I’m thinking about this so much.” And I said, “Are you still kind of focused on this B2B thing?” and they said, “Yeah, we’re still focused on this B2B thing.” I said, “Are you still looking for kind of a glorified engineering manager?” and they said, “Yeah, we’re still looking for a glorified engineering manager.” And I said, “You still want me to take a 40% pay cut?” and they said, “Yeah, still want you to take your 40% pay cut.” I said, “Sold!” And that was definitely a moment of just following my passions for solving this problem and trying to come together with this team to figure it out. And a kind of miraculous thing happened. I arrived at the company, and two weeks later, Joe Kennedy, who’s our CEO, started, and the company raised a little bit of money. And suddenly we had a new mandate to create a consumer service rather than this B2B thing. So I’m, like, “God! Halfway there. I’m halfway to the job that I really wanted. Now we get to build a consumer service.” Another few weeks goes by and my phone rings and it’s the founder/CTO/technologist product guy who had hired me and he said, “You know, I’ve got a new baby. I’ve been at this for four years. I’m really kind of a B2B guy. I’m going to leave, and I think you should take over running the product organization.” So in just six weeks, I went from accepting a job description I didn’t really love and working out a product that wasn’t a perfect fit, to suddenly I had exactly the job I wanted and exactly the product opportunity. And ultimately we spent about the next six months doing the work that led to Pandora, which I’ll tell you a little bit more about in a minute. But lest you think that my enthusiasm for passion-driven decision-making when it comes to career choices is limited just to the things that at least seemingly have a happy ending — things are pretty happy for Pandora these days — let me you one more story about being passionate. How did this happen? So I had worked at Apple and then I had worked on this video game You Don’t Know Jack, and somehow I got it in my head that the hard computer science must not be in computer software. It must be in backend infrastructure. And so I had an opportunity to be the first employee at a company called Relevance Technologies. And Relevance was working on a kind of matching search technology that would allow you to search through huge databases of documents and pull things out based on meaning. So I came aboard and led a small team there that was responsible for all the front-end components of this. And after about just a year, we sold it to Documentum, which is a great outcome for everyone. Documentum was this big successful East Bay company, had all kinds of resources. They had been traditionally a client server company and they really needed people who were thinking about the internet and the Web, and they brought us in for this very strategic goal in mind. And I spent about a year there trying to figure out enterprise software, trying to get excited about document management for big corporations. I mean, the kinds of things Documentum does is like makes software so that Boeing can manage all of the regulation-related paperwork for creating safe aircrafts. Like, it was very, very hard for me to personally become invested in this problem. And my phone rings, and it was the CEO of Berkeley Systems, the people who made You Don’t Know Jack, which is one of the great kind of experiences of my career to that point. And she said, “We’re putting the You Don’t Know Jack team back together and we’re going to attack something in the internet space,” and this was when eToys had just gone public and was worth $9 billion or something. This was a crazy, kind of frothy time. And notably, I had just gotten a new puppy. And I loved this puppy. And this puppy was like the absolute center of my life. I mean, I had done nothing but work for the first many years of my career and suddenly it just felt like I had a whole other reason to get out of bed in the morning and it was this puppy. And so this CEO, Julie Wainwright, says to me, “And the company is going to be for people with pets. It’s a $28-billion e-commerce opportunity. And we want you to come and run Engineering. It’s called Pets.com.” Yeah. I nearly destroyed the U.S. economy by myself. And I had said that Relevance was about a year old when it got acquired by Documentum. You guys all understand how stock options and these things work. I had a bunch of unvested shares in Documentum. And I said, “Keep your shares. I’m going to go make the internet safe for pet food,” and went off. And I try not to do the math, but it was millions of dollars that I walked away from when I was 28 years old to be able to go pursue this experience with Pets.com. And it was a disaster! I mean, it was a glorious disaster, right? I mean, the whole company from beginning to end was 18 months. We founded it, we got it funded, we took it public, and we drove it off the cliff all in just 18 months. And I’ll tell you, I still count it as one of the great experiences of my career. I learned so much. I worked with the most amazing people. If you really wanted to get me going, we can have a conversation about whether or not it’s fair that Pets.com is the poster child for all the excesses of that period, because mostly what I remember is working my tail off trying to make that company work. And it was definitely many, many, you know, 120-hours weeks, and a lot of bad decisions along the way. But I don’t regret that for a moment. And I’m certain that I wouldn’t be at Pandora today doing the things that I’m doing if I hadn’t been at Pets.com. And that experience is absolutely foundational to everything that I’ve done since and probably everything that I will ever do in technology. And so, again, I think it’s one of those examples where I think if you follow your passions, good things happen. And I don’t intend to spend 15 minutes on the remaining six lessons. But let’s talk about another one that I was fortunate enough to learn early. So I went to work at Apple. And Apple is this really remarkable place because it is absolutely filled with the smartest human beings that I have ever encountered. I don’t know how they let me in, but it was a really great honor to be able to work with these people. But I will tell you something about Apple, at least in terms of how it affected me. Apple is a place that cherishes Renaissance thinkers, people that are good at lots of things. The people that are most celebrated in the company culture can write software and design user interfaces and have brilliant marketing insights and have a great eye for design. And so in the four years that I was at Apple, I wrote about 500 lines of code and dabbled in a million things. I was miles wide and about a quarter-of-an-inch deep when I finished my time at Apple and decided to move on to something else. I wasn’t great at anything. And I went to Berkeley Systems and together there worked with a team that created this video game, You Don’t Know Jack, which was a little bit of a phenomenon in the late ’90s. It was kind of this irreverent quiz show-party game. And Berkeley Systems was completely different than Apple from the standpoint that everybody knew exactly what their job was, and your job was to be great at it. And in my case, it was to run the little development team that wrote all the software for You Don’t Know Jack, so the sound mixers and the gameplay and the graphics engine. But it was not my job to design the gameplay or to design the look and feel of the game or to write the questions or to figure out what the packaging should be like or what to go to market. I was welcome to occasionally chime in with a suggestion or a thought, but it was very, very clear that the thing that was going to make me successful at Berkeley Systems was being the best software engineer that I could possibly be and to lead that team really effectively. And so I really focused on that for my years there. And thank goodness, because I wouldn’t have been a software engineer, not really, if it hadn’t been for that experience. And so I think the lesson for me of that period is, it’s really important to be great at something and to know what that thing is and to really invest in it. You’re going to have all kinds of opportunities to go wide, but look for opportunities to go deep. At least for me, it made all the difference. So let’s see… You know, one of the things, having done start-ups for 15 years, that I think has been true of every moment of success that I’ve ever been involved in is that focus is maybe — if you can take anything away from this, I would challenge you, as you think about your ideas for the companies that you’ll undoubtedly start, focus on a small number of things and be really, really great at them. Again, kind of this idea of ‘be great.” You know, I have mentioned that I worked for this company, Kenamea, in the sort of 2001 to 2004 period. And that was an incredibly challenging time in the market for any kind of software solution. I mean, there was no opportunity to do anything for the consumer, really, and there were very few opportunities to do interesting things in enterprise software. But we were trying at Kenamea. And some horrible things happened beyond just the economy. We had tragedies like 9/11. And, you know, that particularly impacted Kenamea because we were developing software specifically for the financial services industry. That was going to be our main go-to market strategy. We were going to take… You know, this company was so esoteric that I don’t think I’m going to bother talking about what specifically we did other than say it was financial services-oriented. And the week that we were set to launch was the week of 9/11, and, of course, that entire disaster, the epicenter was in the heart of the American financial industry. And so we were faced now with an industry in disarray and a really challenging time generally in the economy. And we did the natural thing, which was we bobbed and weaved for three-and-a-half years. We tried four different products and five different strategies and we would go and pitch an enterprise, and they would tell us about their problems-du-jour and we would run home and try to figure out how to write software that would solve that problem. We weren’t focused at all. And I think you know, candidly, Savage Beast, the company that was the precursor to Pandora, suffered from that a little bit as well, and it was the same period. And I think it was a period where we had few choices other than to search for markets when so few dollars were changing hands around technology. But Savage Beast had this Music Genome Project, this kind of miraculous technique for matching music together, and they were going to market and talking to everybody that sat somewhere in the value chain between consumers and music. So they went to Best Buy and Borders and Tower Records and AOL. And they said, “Tell us about your problems,” and Best Buy says, “Well, Borders is killing us because the have listening stations in the stores. We really need some solution for that. Could you help us build a kiosk so people can listen to music inside a Best Buy store before they buy it?” And we would say, “Yeah, sure.” We didn’t really know anything about kiosks, but, “Sure. We could do it. Could it do music recommendation also?” And they go, “Yeah, it could do music recommendation.” And so we’d come back and we’d write software. And we even went to IBM and had them design special touch-screen kiosks, and the head graphic designer on the team had an industrial design background and he designed the bracket that we used to mount it onto the CD stands. I actually went and physically installed them in 15 Best Buys somewhere in the Midwest. And there is something sort of magic about Best Buy at three in the morning, but, we don’t have time for that story. And so the company was really, again, kind of putting its feelers out in lots of different directions. And one of the really transformational things that happened when we kind of re-visioned the company in 2004 is we said, “We’re going to stop doing this. We’re going to pick something that we’re going to be great at and we’re going to focus on that like a laser beam.” And in the case of Pandora, that was to become the world’s best radio station for connecting you with music that you’ll love. And we have stayed very, very true to that. I mean, we’ve never thought of ourselves as a website, we’ve never tried to be a social network for music, we’ve never gotten involved in music videos. The list goes on and on and on of all of the things that are constantly trying to distract us from this mission of being the best radio solution in the world. And we’re kind of constantly reminding each other of, like, it’s still important, even with 50 million users and a pretty good story right now, to stay focused. There’s so much left to do. And so I think focus in all that you do in start-ups is incredibly important. The next one is to be agile. This is a word that’s thrown around a lot in software development these days, but I think it’s a pretty good descriptor of…one of the best opportunities you have as an entrepreneur in a start-up environment, and that’s that larger, more established entities can’t move as quickly as you can. And success and scale and size all fight against this. And I would encourage you, for the longest possible time, as you sort of start your own companies, to look for ways to stay agile. Pandora has been…well, the company’s been around for six years focused on this radio problem. We’ve been in the market for five years. There’s all kinds of momentum. We have about 180 employees these days. But we’re still very focused on staying agile. And I’ll be a little bit specific about how we do that: we don’t have any long-term plans. We don’t have any medium-term plans. I know what we’re going to do for about the next three months, and beyond that, I can wave my hands a little bit, but I can’t tell you anything specific about what comes next for Pandora. And that’s very much on purpose. It’s not for lack of ideas. I mean, we have a list of ideas that’s a thousand lines long. But we resist the temptation to take those ideas and put them on a calendar for the next year or 18 months. And believe me, there are lots of dimensions of the business that would benefit from us doing that. We are an advertising-supported company, and when a big consumer package goods company like Procter & Gamble comes to you and says, “We want to run an ad campaign with you, Pandora,” well, they’re going to talk to you about an ad campaign that’s going to run nine months or 12 months from now. And so what they want to know is, “What can we be doing with you in a year that will be novel, that will be new?” and I have to tell them, “I have no idea.” They don’t like that. So that’s just one example of many where people are going to want you to make long-term commitments to plans. But there’s so much benefit in resisting that temptation. I’ll just give you an example from the last three months, and you can tell me whether or not you think a year ago we should’ve been committing this quarter to a bunch of things for Clorox. So, in January, we had the opportunity to partner with Apple for the iPad launch, to be on stage with Steve Jobs at the iPhone OS4 announcement to show off background processing, to launch with Ford in the Ford Fiesta for 2012, and to partner with Facebook to be one of the sort of exemplar applications for their new open social graph product. All of those opportunities dropped in our lap in January and had to be done in April. So if we had–and, you know, my product organization is about 40 people, and we can’t do anything like everything that comes our way. And so if six months ago and nine months ago and 12 months ago, we’d been committing one little thing at a time to some long-term plan, I promise that we would be busy doing a whole bunch of sort of things that are no longer important to the business. So I would encourage you to look for ways to stay agile, particularly around product decisions at the last possible moment, because you’re going to constantly be accumulating new information about your business. And that really is one of your great advantages against larger companies. They might have more resources, but they don’t have the ability to react as quickly as your young company will. The next one is, be decisive. My favorite Pets.com story is a little bit about being decisive and a little bit just funny, so I’ll share it with you. Julie Wainwright was the CEO/founder of Pets.com, and I had worked with her at Berkeley Systems on You Don’t Know Jack. And Julie is a force of nature. She is a brilliant woman and incredibly hardworking and very decisive. And when we were launching the company, we wanted to do something to grab a bunch of attention. And we decided that the way we were going to do that is we were going to have a big sale when the site launched. And keep in mind, we didn’t have any infrastructure at all for this company. The website had been under development for about three weeks. We had this little tiny warehouse that was in downtown San Francisco. If you know anything about logistics, the worst place in the world to put a distribution center is in downtown San Francisco. I mean, just trying to get the trucks in and out were logistically impossible. But we decided that we wanted to have a half-off pet food sale. And I remember sitting in the conference room, talking with Julie about the business rules that go along with such a thing. There’s a whole bunch of decisions that you have to make about, like, well, a sale like this, how is it going to work? Who’s going to be able to buy it? How much food are they going to be able to buy? And so on. And that list of questions was long. We were all stressed out, there was way too much to do, and I think many of us in the room were using all of the unanswered questions as kind of a passive-aggressive way of saying we’re not ready to launch this company. And I remember somebody saying, “Well, what’s the business rule going to be for how much dog food you’re going to be able to buy?” And Julie, just like that, said, “You are allowed to buy a thousand dollars worth of dog food at 50% off. You can’t buy any more than that.” I should mention one other thing. The system we had for applying shipping charges to our orders was also incredibly immature. For example, we didn’t know how much any of the products in the catalog weighed. So, all we could do was offer — well, of course, this was, you know, 1999, so of course shipping is free. So, “Yeah, we’d ship your dog food to you in the U.S. Postal Mail for free. And then if you wanted to upgrade to UPS it would take two or three days. We could do that for $5 for your entire order. And if you wanted it delivered overnight, we would do that for $15 via Federal Express.” Now, it turns out that a thousand dollars worth of dog food is about a metric ton. So we’re — it’s a true story. So we launched this thing, and it’s not 24 hours until my cell phone rings. And it’s a guy in our distribution center. And he says, “Tom, I’ve got an order here on the pallet, getting ready to get picked up by Federal Express where they will put it on an airplane and send it to Hawaii.” “It’s a thousand-dollar order, almost to the penny. It weighs about a metric ton. And it’s going to cost us $15,000 to ship it.” Happily, another decisive moment for Julie was that she had said that another one of the business rules is going to be that you couldn’t buy food in this program if you were a dealer. And it turned out that the person who had made this order was themselves a pet store somewhere in Kauai. So we were able to, under the terms of the sale, cancel the order. And, you know, I think that — the important part about that, it wasn’t a catastrophe. The phone rang, we didn’t spend the $15,000. Even if we had in the grand scheme of the $100 million-plus we lost, it wouldn’t be the end of the world that… The favor that Julie did the company in that moment, and in many, many moments throughout its history, was that she was very decisive. She knew precisely what she wanted. And if she wanted to sell live fish, by God, we were going to sell live fish. Which is another story I don’t have time for. Compare this to the time that I spent at Apple. The company, as I said, was full of brilliant people, but I was there in 1991 to 1995, when Steve Jobs wasn’t around and when arguably the company lost its way a little bit. And there was a joke while I was there that I think sums up what was wrong with Apple in that era. And the joke was that Apple was the only place in the world where a vote of a thousand to one is a tie. And clearly they’re not voting anymore, or there’s only one vote that matters or something. But that can happen, right? You get a bunch of smart people in a room and you all kind of agree on some path, some solution, something that’s going to move the company forward, and then some smart person says, “Oh, you know, if we do that, here’s the downside,” and all the other smart people go, “Oh, that’s right. That’s true. Let’s just have another meeting.” And so we were in this kind of analysis paralysis at Apple, and it made it really, really hard to ship anything to end users, particularly on the software side. And it meant we tried to do a lot of these kind of big kitchen sink releases where we tried to kind of boil the ocean and solve every single end user problem all at once. So be decisive. Let’s see… This one is really kind of about going to market. And it’s really born principally out of my experience at Pandora. People ask a lot how we were able to grow Pandora to its current listenership. And the truth is, I don’t really know. We don’t have a marketing budget. We tried to build the best product that we could. We tried to be really focused. We tried to be good at just one thing and put all of our energy behind that. But I don’t know what made it resonate with people. But I can tell you that Pandora spreads by word-of-mouth. And I literally mean that. I don’t mean viral marketing. I mean words and mouths. People sit around the Thanksgiving dinner table and say, “Hey, I found this thing that I love called Pandora. You should really check it out.” And we noticed this phenomenon very, very early. And we decided to really embrace it and to try to have the tools that we would use in that process be words and mouths. So we’ve really tried to foster a direct and genuine relationship with the people who love Pandora. Early on, this meant, instead of a launch party — and having been through the dot-com era myself, when somebody said, “Should we have a launch party?” my first though was, ‘Oh my God, no’ because that seemed like such a symbol of the excess of that era. But as we talked about it, we came to this idea that we would have just a little get-together, kind of an informal gathering of folks in San Francisco who had discovered the product. And so we put a post up on the company blog, which was a very in thing to do then. We said, “Come meet us for a few beers at this bar in downtown San Francisco and talk to us about the music you’re discovering on Pandora.” And so we did that and about 30 people showed up. And it was just this magical night for us. We probably spent 50 bucks on pitchers of beer, but we met these remarkable people. And we came back the next day and said, “Wow, that was really kind of magic. How can we do more of this?” And somebody said, “Well, we all travel a fair bit and go and meet with potential partners or whatever, and maybe what we should do is like when I’m in Boston next, we’ll put another little message on the website that says, ‘We’re going to be in Boston. If you want to come out and meet us for a coffee, we’re going to be at this coffee shop.'” And so we started doing that. In particular our founder, Tim Westergren, really embraced this role, and pretty soon he was purposely making trips so that he can invite people to have coffee with him. And those grew. The first one in New York, I think, was five or six people literally in a Starbucks. I think the last one we did in New York City was almost 800 people in a big auditorium. But the format is basically the same. Tim sits down on a stool in front of whoever shows up and they talk about Pandora. He tells a little bit of his own personal story. And Tim has unbelievable stories about what it took to start this company and to get it to where it is today. And then he takes questions from the audience, and people stand up and say the most remarkable things about their love of music and what they’re discovering on Pandora. And they’re kind of magic nights, and they come from this desire for us to get out there and just genuinely interact. They’re not scripted. They’re not talking points. There is not an agenda behind what we do other than just this desire to genuinely connect with our listeners. Another dimension that’s born out of similar thinking is that when you sign up for Pandora, like most companies, we send an automated email to you that says, “Welcome to Pandora. Here’s three things you probably want to know about this service.” But we have a couple of things that are somewhat unique in my experience. One is, the email comes from Tim, our founder’s email address. It’s really his email address. We send 50,000 or so of these a day. Not only does it come from his email address, the email actually says, “Please write me back and let me know what you think of the service.” And so every day, for the last five years, hundreds, some days thousands of people, write to Tim. And he tries to respond to all of them. And when he can’t, he has a handful of people who help him. But they don’t respond pretending to be him. They say, “Hey, I’m Lucia. I picked up Tim’s email. I just wanted to write you back.” And he has these unbelievable exchanges with people. And I’m convinced that these emails are the gasoline that we throw on the fire of the kind of word-of-mouth behind Pandora. That you can — oh, and just some of them are priceless, right? I mean, there are people who write back and say, “Oh, sure, I can write you back,” and whatever, and he responds, “Hey, how are you doing?” And I love the ones where they say, “Oh, you played ABBA on my whatever station. How could you play ABBA on my station?” And Tim jumps right in. He says, “Well, was it a good musical match for the station?” “Well, yeah, but it’s, I don’t know, because it’s ABBA,” and then they go back and forth. By the end, the person — no, Tim actually has a real story about this where the punchline of the story is the last email says, “Oh my God, I like ABBA.” So, you know, take those opportunities to engage with… Well, I can’t believe I sign up for these new services all the time. I have to check out what other people are doing,” and the first email you get from them, the first thing you see, before you even open it, is the email–the address is, like, donotreply@forshizzle.fr, or whatever. And it’s, like, “Really? Why would you not want people to reply?” So, you know… And this whole philosophy goes back to the whole kind of Cluetrain Manifesto thing of, you know, people recognize a human voice. They try to be human. When we put copy on the website, we spend all of our energy trying to pound the marketing speak out of it. It’s like, “I like that little turn of a phrase, but how do we rewrite this? How would you say it at Thanksgiving dinner? How would you talk to this person like they’re your friend as opposed to someone you’re selling to?” It’s really hard to do, but I would encourage you to make that investment. All right, last one, and this is somewhat ironic since I just spent the last 45 minutes talking about myself. Be humble. If you spend six years of your life being decisive, focused on being great, being focused, working hard under incredibly stressful situations, the only way to come out of the other side of it with any friends at all and any sense of joy in how you spend your days is to be humble. Hire people that are wonderful, that aren’t jerks, that realize that when you find success it’s not about you. I mean, the success at Pandora had nothing to do with me, despite my 45 minutes. It has to do with our listeners, it has to do with the wonderful people that we’ve hired. I think that there is, can be in Silicon Valley, a kind of cult of personality things that develops around this companies, and our founder Tim, who is the face of the company, is the most affable, humble person that I’ve ever worked with. And he’s a constant sort of inspiration to me to kind of try to remember that we’re out here trying to solve problems. It’s not about sort of creating a legend around you as an individual. It’s around helping people do things that they could never do before. But by all means, if you get a chance to come talk about your career, do it. It’s fun. So that’s what I prepared for you today, and by some miracle it’s almost exactly 45 minutes. And my guess is that you have lots of questions. And we touched on lots of things, but I certainly came here to talk about Pandora specifically, if you’d like to hear more about that. I’m sure somebody came to hear more about Pandora. Yeah? So you guys have to hire and pay actual musicians to go through all the music? We may be — we’re one of four organizations on the planet that hire and pay musicians, which is part of why it works. It’s like us and Starbucks and… Yeah, so we have a team of about 25 part-time musicians who do the Music Genome analysis work. And many of them have been with the company almost as long as I have. It’s a very coveted job because it has…well, frankly it pays well, has great benefits, and it’s incredibly flexible and fits the lifestyle of a professional musician really well. Many of the musicians who work for us are touring musicians themselves, and so they have this flexibility to come and work for us and then to go off on the road for a tour for several months and come back. So there’s a long waiting list of people to get in. And the reality is, though, in the grand scheme of all the costs associated with running Pandora, the Music Genome Project itself is kind of a drop in the bucket these days. Yeah. You mentioned your experience at Pets.com was foundational for your later career. What was it about that experience that was so different? You know, one of the neat things about Pets.com was that our principal investors, the first $50 million, if you can believe it, came from Amazon. And Amazon had this incredible focus on customer service. And I think living through that 18 months, I learned all kinds of things. I learned lots of ways not to do things. I worked really hard. I had a team that worked really hard. I managed through a really scary period. But I think one of the big things I took away was, like, really focus on your customers. You know, care about them, understand that you’re there to serve them, and the folks at Amazon really instilled that in all of us. So I think I walked away with a tremendous sense of kind of how to put the customer at the center of everything that you do. Yeah. Did you say that there is a trade-off between being focused and being agile? Because you were talking about at the original company when the Best Buy opportunity came up and you guys followed that path. And even though it strayed from your original path, it could’ve been kind of the next big thing in the way the company could’ve progressed in the future. So do you think that you should always stay true to the core competencies of your company or explore the opportunities that come your way? You know, I think… I think it’s really important to know what you want to be the best in the world at, and you want to be kind of agile within that momentum. You know, we wanted to be the best in the world, and still want to be the best in the world at kind of introducing you to music that you’ll love through this really easy, serendipitous mechanism. There are opportunities that come our way all the time that have potentials to be big markets. One example is talk radio. Music radio is just a fraction of the overall story. Talk radio is certainly a huge category, too, and there’s probably interesting personalized things to do in talk radio for Pandora. And it may be something that we do at some point in the history of the company. But right now, we know pretty clearly that we are a music company, and we’re really focused on doing that well. A great way to stay focused within a vision is to find the metrics that make your successes seem small. And for example, we have 50 million listeners and 21 million on mobile. We have 47% of internet radio, but we are like 1.2% of all radio hours in the U.S. So I really try to focus my team on that 1.2% number and remind them that without changing anything, without deviating at all from our focus, we’ve got 98% of the market still in front of us. And Clear Channel has 20%. So it’s not like there aren’t examples of companies that get really big within that category. So that helps. Yeah. Can you talk about the ways that Pandora went about iterating on how it monetized the service? Because I think that’s probably going to get a lot of questions. Yeah, there’s a story there for sure. So, when we launched — and this is a chapter of the company that we don’t talk about a lot, because it wasn’t very long-lived. When we launched… So the company raised some money in the spring of 2004. That’s when Joe joined by coincidence, that’s when I joined as well, and we spent about a year kind of exiting the kiosk business and the API business and all the B2B stuff that we’d done and figuring out what kind of consumer software company we wanted to be. And it was then about a year after that funding had come in that we were able to launch the first version of Pandora. And so as a consequence, we had spent a fair bit of the money that we had raised. And music services, you may have heard, are very expensive to run. They have high licensing costs, they have bandwidth costs, of course the overhead of the operation itself. And many of us, a bit battle-scarred from the dot-come era which was so focused on “get big and monetization will come,” decided that we wanted to launch the company with this idea that it could sustain itself relatively well from the beginning. And we decided that the only real path that was open to us was subscription, because you can’t really build an advertising business when you don’t have any audience. Advertisers buy reach that they know you’ve got. So there was not going to be any short-term monetization opportunity around advertising. That was always part of the sort of medium-term plan, if you like. We thought maybe we’d go for about a year and then start turning advertising on as an ad-supported free option. So we had a spreadsheet — you know, the one, if I was going to do eight, I’d say have a spreadsheet –that described how the business was going to work, what the costs were going to be, how many people we need to convert to subscription to kind of ramp towards profitability. And we launched things and it worked out exactly as we imagined it would. We’d converted people to subscription at pretty much exactly the same rate that was in the spreadsheet. Except there’s one key difference, and that was people found the service much, much, much, much faster than we imagined. So rather than there being 50,000 people who found Pandora into the first month, half a million people found Pandora in the first month. And when you’re going to churn off 98% of 50,000, it’s a pretty different feeling than churning off 98% of half a million, when ultimately you want to get to this reach play that’s advertising-oriented. And so we started looking at a model then that was advertising-supported much, much, much sooner. And about that time, we were looking at raising a little bit more money. And one of the potential investors started modeling out this opportunity to run Pandora as an advertising-supported business sooner rather than later, and they came to the conclusion that that was really the company that they wanted to invest in was this advertising-oriented company. With the benefit of their investment, we were able to accelerate our plans. And I remember after we closed that round of funding, it was a couple of days before a board meeting and we all knew that the newly expanded board was going to want to hear about how you’re going to turn Pandora from a subscription service into an ad-supported free service. And I had thought about this a little bit and put some plans together on the product side and had come back to the rest of the management team and said, “I think in three weeks, we can take this company that we had built entirely as a subscription service and turn it into an ad-supported service. We need to get an ad server, we need to make it possible to register in the system without having a credit card on file,” all of these things. There was a bunch of work to do. But I was pretty proud of myself! I thought, ‘Three weeks? That’s pretty fast!’ You know, we weren’t supposed to do this for another six months. I sort of was patting myself on the back as I walked into the board meeting. And we rolled out this plan to them and the board said, “Wait a minute, isn’t there a free trial already?” I said, “Well, yeah, there’s a free trial,” and they were, like, “Well, why don’t you just make the free trial last forever?” I was like–well, I of course had 15 reasons why that was not a good idea, not the least of which is we didn’t have any ad-serving infrastructure at all. But they were, like, “But you’re not going to have any advertisers.” Well, that’s a fair point. And so I said three weeks and I remember literally they said, “We think a nanosecond is the right answer.” So I went back, sharpened the pencils, and figured out how to do it in three days over a weekend. And three days later, we launched ad-supported free Pandora without an ad server. Two things happened. It’s the one time we were caught flat-footed from a scaling standpoint. I was sure that because we had this 10-hour free trial, that word of mouth was already really strong on Pandora. “Hey, try this thing. You don’t have to do anything. There’s no commitment.” And turns out, free? Really, really different than 10 hours free. You know, it was like that ad where the dot-commers were sitting around watching the orders go click-click-click. That Monday, things got really out in front of us for a few hours. But that was sort of a good problem to have when you quickly got it back under control and have been growing at a pretty remarkable clip since. But then, the phone rang in our CEO’s office, and it was Apple. Apple said — this is three days in — “This is great that you’re ad-supported free now. We’d like to be your first advertiser.” This was November of 2005. “In fact, we’d like to buy out all of your ad inventory for the months of November and December.” So Joe says, “Oh, it’s so great that you called,” comes into my little office and says, “Apple wants to be our first advertiser.” Well, that’s funny because we don’t have an ad server.” Or a website that was designed with any place to put an ad. And he said, “Well, do you think we could do it? I’d hate to turn Apple down.” And I said, “Yeah, we could probably do it. I mean, how hard could it be? We’ll hard-code their ad onto the homepage.” And so we did that. We hard-coded their ad onto the homepage, in like the worst possible place ever. If you saw screenshots of Pandora from this period, it’s like, “Why would you put an ad there?” because the whole site was not designed for ads at all. But then, advertising is — particular the company that is sort of meticulous about their marketing images as Apple, they changed the campaign a lot over the course of that two months. And so they would call and say, “We have new creative for you. Can you traffic the tags in your ad server?” And we’d have to push out an entirely new version of the website to our 50 servers or whatever. And they called all the time. And it was like just one fire drill after another to change these ads. So, yeah, we became an ad-supported service and really focused on ad-supported solution sort of exclusively for about four years. And then we started talking about other monetization opportunities and I started thinking about, well, is that old thing we tried with subscription, and it really wasn’t that interesting because only 1% of people subscribed. I was, like, “Wait a minute. How many people are on the site everyday? Oh, wait. A lot!” You know, 1% of a big number is a pretty interesting number when you multiple it times $36. And so we have a pretty robust subscription business now that we’ve been focusing on in addition to our advertising business for the last eight months or so. It’s still… It might have been 5% of our overall revenues last year, it might be 10% of our revenues this year. Advertising will always be the principal monetizer for Pandora. And in many ways, I think it will always be where we put a disproportionate amount of our energy because that’s where our audience is. You know, 1 or 2% of our audience is going to subscribe. It’s great that that generates 10% of the monetization potential, but what we really care about is how do we make 150 million people experience Pandora? And you do that through ad-supported free. Great. I want to thank you so much for a fabulous event. This was terrific! Thanks so much.
Tom Conrad, Pandora – A Pandora’s Box of Start-up Expertise.