Frank Chen, Partner at Andreessen Horowitz: ‘Are you a mercenary or a missionary?’
Total read time: 10-15 minutes
The 10th episode marked the first VC on the program! Last Wednesday I went to the office of one of the leading Silicon Valley venture capital firms, Andreessen Horowitz, on Sand Hill Road in Menlo Park, CA to talk to a partner at the firm - Frank Chen.
Before joining Andreessen Horowitz, Frank was the VP of products at Opsware. Prior to that, he was a director of client services at Loudcloud. Before Loudcloud, Frank worked for as a VP of products & UI design at Respond.com. He also served as a director of product management at Netscape, a product manager at Oracle and a senior technical writer for GO Corporation.
We set up an interview with Frank’s assistant back in February over email. And two months later here I was sitting at the conference table at A16Z next to him:
- Let’s go back to your college years. What was the trajectory which led to where you are now?
- I was an undergrad at Stanford, and at Stanford, I did a degree called Symbolic Systems. It turns out there’s been a lot of very famous entrepreneurs to come out of that program. Reid Hoffman, who started LinkedIn, was a year ahead of me. Marissa Mayer, who’s now the CEO of Yahoo!, was a Symbolic Systems major. Symbolic Systems is computer science, psychology, linguistics, philosophy. There’ll be a quiz after this. By nature of it being an interdisciplinary program, you’re always looking at the same problem in lots of different ways, and I think that’s why so many entrepreneurs come out of this problem - which is it’s not narrow. You’re always thinking about, for instance, how do people learn language from the point of view of a philosopher, a linguist, a cognitive scientist, and computer scientist? It was great training.
- Was Go Corporation the first company you ever worked at?
- It was my first full-time job. I had a couple of summer internships at Apple and IBM. In fact, I did my summer internship with Reed Hoffman at IBM. Think of two startup guys inside IBM. It was like we were total fish out of water! Go was my first full-time job. In 1992, Go Corporation, for those of you who don’t know it because most don’t, we tried to build the iPhone. In 1992, it was 15 years too early, literally. This is one of the perils of being a startup person - which is you never know when the right time is for your startup.
- What are your responsibilities now as a partner at Andreessen Horowitz?
- Let me back up, because venture capital is probably a little mysterious to most people who are watching this video. A venture capital firm really only has 3 jobs. Job one is to find great companies. Job two is to pick from among that set the companies that you really want to get involved with and invest. Three - it’s to help them grow to be the biggest possible companies that they can grow. That’s the mission of the firm in a nutshell. My job is primarily on steps 1 and 2, which is finding the companies, and then evaluating the opportunity set to make sure we make the best possible investments.
- This is actually one of my favorite questions. What does your typical day as a VC look like?
- A day in the life of a venture capitalist. Let me divide up the pie chart. 30% of my time, I’m meeting with an entrepreneur for the very first time. It’s the first time that they’ve ever been with the firm, and they’re pitching their idea. 30% of my time, I’m spending time with excited entrepreneurs who are absolutely convinced they’re going to change the world. It’s just a naturally exciting part of the day. 30% of the time, I am going deep on a company. The questions that I’m trying to answer are: Is this the best possible company to invest in, in this space? Do we want to place a bet in this space? What is everybody else doing in this space? What are other startups doing? What are big companies doing? We’re going deep on a particular company, and that’s 30% of the time.
The rest of the time, we’re thinking about technical trends. We’re meeting with big companies. We’re really trying to understand the ecosystem that we’re in, which is, if you’re going to invest in a systems company, what’s VMware doing? What’s NetApp doing? If you’re going to invest in a social network, what’s Facebook doing? What are the other companies in this space doing? We need to make sure that we’re on top of what’s happening in the rest of the industry.
- This is straight from Mark and Ben’s original vision for the firm. When they setup the firm, the firm is an answer to the question, ‘What did I wish I had gotten from my venture capitalists?’ Between the 2 of them, they raised over $1 billion of private and public money to build their very, very successful companies (Note: Marc Andreessen has found Netscape, Ning and Opsware; Ben Horowitz was the VP at Netscape and CEO/co-founder of Opsware). Along the way, they raised a lot of money from venture capitalists, and the venture capitalists would promise them all these things that you just described: great engineers, great customer leads, great help along the way, PR help, the whole 9 yards.
The fundamental ‘ah-ha’ was if you’re actually going to do that and help portfolio companies, you have to hire a team to do it, which, as I’m telling it to you now, you’re like, “Duh. Of course you have to hire a team. How else could you actually scale to provide that?” This is innovative in the venture capital industry. By numbers, we’re roughly a 70-person firm. 50 of those 70 people wake up every morning thinking, “How do I help the portfolio companies get big?” The vast majority of the firm doesn’t do what I do. We don’t think about deals, we don’t think about valuations, we don’t think about, “Is this the best possible investment?” We think about, “Who do I need to introduce portfolio company-X to industry player-Y? How do I help them build big companies?” It’s 50 out of the 70 people.
- Which verticals, for example quantified self, which is really hot now; payments, enterprise are you the most excited about? What are some examples of companies Andreessen Horowitz invested in, in each of these verticals?
- We actually take a different approach to thinking about investments. Rather than saying, “I’m excited about quantified self. I’m excited about internet of things. I’m excited about payments.” By the way, we’re excited about all of those spaces. What we do is we say, “We want companies who are fundamentally changing the world with software.” We want highly-technical founders because software’s hard to write, and as you watch big companies struggle to write good software, you see that every day in your own experience; it’s hard to write really, really good software with a great user experience. We want the founders who really want to build big companies.
Those are really our guiding principles. Is it a software-eats-the-world company, that’s our investing timeline, and is it going to be a big company led by technical cofounders? If it hits those bullets, then we want to look at it, regardless of the space that they’re in. There are some venture firms who will, at the beginning of the year, draw up a Bingo card. It’s got 12 spaces on it, and each space is basically a market segment: Internet of things, payments, quantified self. They’ll absolutely drive their entire investment agenda using that Bingo card, which is, “I am going to find the number 1 company in that space and I’m going to invest in it.” We definitely do our homework. We want to understand what’s happening in the universe, but we also want to let the great entrepreneurs tell us, “I am defining a company where it’s not even on the Bingo card. You never even thought about it.” The reason for it is we’re in the outlier business. When Google raised money in the mid-1990s, people thought it was the strangest investment of all-time. 20 search engines, completely commoditized, no path to monetization, late entrant to the market, and it was crazy. Look what happened. It was an outlier that basically said, “Here is how I tie a world-beating search engine to the revenue growth engine of all-time.” It wasn’t until Google made that possible that people saw.
A funny story. That same year, Kleiner Perkins, which was one of the early investors in Google, also invested in a company that, frankly, they were more excited about. If you go back and read the press coverage, you’ll see much more coverage about this other company which they funded in the exact same class as Google, and that other company was Segway. It was going to change the urban landscape and revolutionize personal transportation as we knew it. It’s very hard to predict the future. We’re in the business of trying to predict the future. It’s an uncertain business.
- You let serendipity happen, and some companies just define unique category.
- That’s exactly right. You let the great entrepreneurs with the big idea and just indomitable thirst to win tell you, “I’m going to define a company in a new market space.” We love hearing those stories.
- In your extensive career as an executive and VC, which entrepreneurs did you work with, and what specific things did you learn from working with them?
- Wow. I’ve had the great privilege of working with great entrepreneurs in my entire career, starting at Go Corporation, the CEO was a guy named Jerry Kaplan. He wrote a very famous book in our community called ‘Startup, a Silicon Valley Adventure’, where he talked about his journey at a startup. The biggest lesson I learned from some of the early companies that I worked with is it’s really hard to predict timing, which is when is the right time where you can get critical mass for your idea? Now we know the world wants smartphones. Everybody wants a computer they can carry in their pocket, that’s available to them every day. Why did that happen only with the iPhone and not in 1992, 1995, 2000, or 2002? All companies who have tried; and it’s really hard to predict when the right time is.
Another great entrepreneur that I worked with is the Horowitz of Andreessen Horowitz. He started a company called LoudCloud, and essentially, we were trying to build Amazon Web Services in the year 2000, which was a little too early to try to build that.
- 4 years before.
- Yeah, 4 years before Amazon started. Actually, it was many more years than that. Ben and team engineered a fantastic outcome for that original business idea, where we morphed into an enterprise software company called Opsware, and sold the company to HP for a little over $1.5 billion, so it was great, great outcome for people. Fundamentally, we were trying to build something like Amazon Web Services, where you could just plunk down a credit card, and then this Cloud infrastructure would take care of scaling for you and automatically add more resources as it was needed.
The thing that I learned from Ben is just you’ve got to have such high conviction and passion around your idea that you’ll stick through the hard times, because we had more than our share of near-death experiences at LoudCloud, Opsware. People were predicting our demise. We almost did run out of cash, but Ben just had such conviction and courage to just say, “I’m going to stick through it.”
The advice that I’d give entrepreneurs for having watched Ben do that is, you got to have high passion around your idea. There are 2 kinds of entrepreneurs in the world. There are mercenaries and there are missionaries. The mercenary says, “I’m going to get rich, and starting a startup is my path to do it. I don’t really much care what the idea is, but I know that if I have 20% ownership big outcome, then I’m going to be rich.” Then there are the missionaries. The missionaries are like, “It’d be great if I got rich along the way, but I am so convicted about this idea that I have no choice but to build this company. I can’t help but to think about it; I have to do it. The idea’s grabbed me. It won’t let go.” Those are the entrepreneurs who are going to stick through the hard times.
As everybody says, “No initial business plan survives first contact with the customer.” You’re going to be surprised in strange ways. The business will go through zigs and zags. It’s never a linear path. Because it’s hard, it’s strange, and it’s not what you predicted, you really want the entrepreneurs who are like, “I just got thrown a curve ball, but that’s okay. I’m going to stick with it until we see this thing through.” That’s the great lesson I learned from Ben.
- Wow. I remember Ben speaking at Startup School; he actually mentioned some of his near-death experiences at Opsware (Note: Quote from that talk, ‘During the post dot-com crash days, I slept like a baby. I woke up every 2 hours and cried.’). Really cool to hear from you, as well. This question might be oversimplification, but still, let’s give it a try. Old startup conundrum: what is your advice to young entrepreneurs, to go for growth first or to go for revenue first?
- It’s a great question. My overall advice, and I’m going to echo the answer that I gave to the previous question, first, absolutely be convinced that this is an idea that you want to commit 5, 10, 15 years of your life to. One of the big mistakes that I see with entrepreneurs is they think, “I’m in and out in 2 years. I’ll get bought, then I’m out on easy-street.” Very few companies have that easy a time. There are always exceptions. We were blessed with an investment in Instagram, for instance, that got a big outcome with Facebook very, very quickly, pre-revenue. Blue Bells like that happen, but you can’t plan for that; you can’t plan your startup around that idea. You really want an idea that you are so passionate about that you’re going to go the distance on it.
Then the question is, “Should I attract a big audience or should I try to get revenue along the way?” I have to divide the world into two here. In consumer land, a lot of the big successes in our industry went growth first, revenue second. You look at Larry and Sergey at Google; their mission was to make the world’s information accessible. Their first product in that space was a world-beating search engine. They resisted the temptation to monetize that thing for a long, long time; long past when the investors were like, “Time to let the revenue go. I see the great search engine. Where’s the money?” They were getting all kinds of pressure to monetize the results of the search engine, but they took their time.
Mark Zuckerberg is doing the exact same thing with Facebook, which is they got to over 0.5 billion users before he turned the corner and said, “Now it is time to monetize.” In our industry, the biggest outcomes have been ones where the entrepreneurs had such conviction around their idea that they pushed off monetization for awhile.
In the other side of the world, on the enterprise side of the world, I would not encourage that at all, which is letting your customers use your product for free can give you all kinds of bad signals. It’s possible that they love your product and they’ll pay you a lot of money for it afterwards. It’s also possible that they’re paying exactly what it’s worth, which is, as soon as you start charging for it, they’re going to all vanish. With enterprise companies, I encourage entrepreneurs to think of how they will charge for the product pretty early in their lifecycle. The last thing you want to do is paint yourself into a corner, where as soon as you launch a priced plan that everybody leaves.
- That goes along with David Heinemeier Hansson’s approach who is really big on charging from day one: “Just put a price on it. That simple!”
- That’s exactly right. He’s done a great job with his business, and we’re real great admirers. He’s not a big fan of venture capital because he’s one of the few entrepreneurs who were able to build a big business out of bootstrapping it, charging customers from day one. Hats off to him.
- His company 37signals has a pretty famous blog called, ‘Signals vs Noise’. Talking about blogs, which blogs do you follow?
- It’d be hard for me to give you a list. I probably read 150 blogs on a daily basis. When Google Reader announced that they were end-of-life-ing, I had to go cry in a corner for literally hours. I’m auditioning some replacements. The nature of my job is that, the way Mark described it to me when he called me was, “Your job is to know everything.” I was like, “Great. I get to go to a job where everyday I’m the village idiot because I don’t actually know everything.” It’s my job and my team’s job to be on top of the tech trends of our day. We read the broad, horizontal ones like TechCrunch and PandoDaily. We read the highly-vertical ones. A great one in mobile, for instance, is Horace’s blog, Asymco, which is, he’s writing about mobile. I read 150 blogs a day.
- Some social news sites like Hacker News, obviously?
- Yeah, Hacker News is the big part of it, and there are great services that aggregate a lot of them.
- Right. Well, you have time until the end of June to find a replacement to Google Reader.
- That’s exactly right.
- Developing on this readership angle, which, say, 5 books changed your life, and why?
- That is a great question. I don’t think I got to 5, but I got 4 for you. Let me walk through them and this will say a lot about me; maybe not so much about career or trajectory for success. One of the books I remember vividly reading as a junior high student was J.R.R. Tolkien’s ‘The Hobbit’. It was just such a vividly rendered, alternate universe. There’s always been a longtime affinity from the computer hacker community in the mid-1970s when this whole industry got started, and science fiction and fantasy writing. For some odd reason, they’ve always gone hand-in-glove. That really had a deep impact on me, thinking about, “Wow. Here was a world where this author just created a world out of whole-cloth. He just imagined it in his head and he rendered it in such vivid detail that it sucked you into that world.” I think a lot of computer scientists, that’s exactly the appeal of computer programming for them, which is, “I can create a world in which people are sharing, or people are playing a game and they’re so immersed in it that it becomes a world in and of itself.” That’s ‘The Hobbit’.
Second book talks about the geek nature. I was one of those classic programming-since-a-kid people, and a lot of people that I meet these days have exactly that story: “I learned to program on an Apple II and I’ve been doing it since I was 12.” I was exactly one of those kids, one of those nerds, but I actually learned to program before I saw my first computer. I read ‘Programming the Apple II’, a book by Apple Developer University; cover-to-cover before I had ever seen an Apple computer. I was just fascinated by the idea of controlling a computer and getting it to display things on a television. Obviously a huge impact, and loved it. Eventually, I did get a computer, and you realize, “Programming is a lot harder when you actually have a computer and things need to compile.”
Another book that had a huge impact on me, and this was actually required reading for Ben Horowitz’s staff. Ben was my first manager at Netscape; that’s where I met him. He had his staff read a book by Andy Grove, who is the long time CEO of Intel, called ‘High-Output Management’. It’s basically Management 101, and it’s exactly what you would expect from Andy Grove, who was not only a great strategist but a great Operations Manager. For the discipline that Intel had to keep on making Moore’s Law happen so that all the magic that software guys do after the fact, he did it for a decade and a half, and it was just amazing; smaller and smaller nanometers, more and more precision, better and better yields. That’s the world of Andy Grove. My job to the industry is to make Moore’s Law happen. He took all of that process engineering know-how and he wrote a book about management. It’s exactly the book you would expect from a process engineer. It literally is the nuts-and-bolts of management. How do you have a one-on-one? How do you do a performance review? Should you focus, as a manager, most of your time on your top-performers or should you spend a lot of your time trying to fix your laggers?
- 80/20 analysis.
- Yeah, exactly. Andy’s answer to that, as an aside, is, “Absolutely, you need to spend time on your top-performers. Do not try to fix the people who are underperforming, because what will happen is you’ll get 10% improvement out of somebody who’s underperforming, but if you spend a lot of time with your top-performer, you’ll get 10X productivity because they’re just that good.” It’s exactly the same insights that he had; maximizing semiconductor yields coming off the line, and he applied it to management. A big impact on me as a manager as I thought about, “Gee. How should I allocate my time?” That’s the book #3.
The last book I want to share is a book outside of the tech industry. It’s a book by Rick Warren, who’s an extremely popular pastor. He pastors a church in Southern California, and he wrote 2 bestselling books. One was called ‘Purpose Driven Church’ and the other is called ‘Purpose Driven Life.’ ‘Purpose Driven Church’ is basically about his journey as a pastor. It’s his startup story, and every pastor starts, basically, with no congregation, no audience, and they have to get to the point where they have critical mass, and then they hit the viral loop, just like a startup. ‘Purpose Driven Church’ is basically his story. If you read it with the lens of, “I’m going to learn startup lessons from this,” you will learn startup lesson after startup lesson because the psychology, the methods, the thinking, the ups-and-downs, the emotional roller coaster; it’s exactly like any startup book you’ll ever read. Doing it outside of your domain gives you a fresh insight. In general, we find human beings; it’s much easier to be an editor than a writer. It’s much easier to find fault with somebody else’s stuff than your own. It’s a lot easier to be a coach than a basketball player.
- Appreciating art versus creating it.
- Yeah, that’s exactly right. When you read about it in somebody else’s domain, I think it just makes the lessons easier to draw for some reason. That book had a huge impact on me as well, watching Rick Warren build his church. Today, it’s one of the most popular churches in the country.
- Thank you for the recommendations. Taking the more personal development, personal growth theme now. In which areas (sports, productivity, learning foreign languages, work/life balance, you name it) would you like to improve?
- That is a great question. In my job, it’s my job to stay up-to-date, so I’m constantly learning new technologies and new technical trends. If we’re evaluating platform-as-a-service companies, then I better understand who’s winning the race, Python or Ruby. I’m fairly constantly taking classes from Udacity, this is one of our investments that make college classes online. At any given time, I’ll be taking 2 or 3 different classes just because I need to stay up-to-date, because it’s impossible for me to evaluate a startup opportunity unless I understand something about that space.
- Which course are you taking at the moment?
- Right now I’m taking a class by a Duke University professor called ‘Irrational Behavior’.
- Dan Ariely! I love his book (Note: ‘Predictably Irrational’).
- Exactly, Dan Ariely. It’s just fascinating. He’s an amazing thinker, and he’s at the vanguard of this interesting trend. For 200 years, the presumption in economic theory has been that we’re all rational actors, and as rational actors looking to maximize our own outcomes, that’s how society’s organized and that’s how the capitalist system is going to work, and that’s how basically we get the self-regulating system out of personally rational actors. Dan’s basic premise is, “No, not so much. We are not rational actors. Our biases that are built-in thinking about money and the time value of money, all of these things are actually not rational, at all.” We have deep biases that lead us to very, very strange and counter-intuitive outcomes, and you have to understand that at the core of economic theory, which is that we’re not rational actors; that our biases get in the way and we’re always making suboptimal decisions. Just a fascinating class.
- Do you schedule certain time intervals to take online classes? Say, 5:00 to 6:00, for example?
- What I try to do is to keep my Fridays clear for learning and not looking at startups, not evaluating startups. That’s a rule better-observed in the breach rather than the observance, but I basically try to have Friday as a learning day, so that you have enough time to really engage on something that you’re trying to learn.
- Talking about schedules. Paul Graham contrasts a maker’s schedule with a manager’s schedule in one of his essays. What are your thoughts on this? Do you have any routines you developed to be more productive?
- Learning Friday is exactly a manifestation of Paul Graham’s idea that there are these 2 fundamentally different rhythms to culture, and I think he absolutely nails it. There was another very famous book written by an IBM software engineer called ‘The Mythical Man Month’ (Note: Fred Brooks). It talks about how you can’t hurry software. You can’t take one person’s task and say, “I’m going to put 7 people on it,” and expect 7-times productivity improvement. You’re just going to make it worse, because the fundamental rhythm when you’re programming is you have to really get into it. Any interruption to you working on a hard algorithm or a hard data structure is going to set you back in a huge, huge way.
People have this experience with their computers all the time, although less these days than in the early days when computers didn’t have a lot of memory. You’ve probably experienced using your laptop and all of a sudden the laptop becomes very slow, it becomes unresponsive; the fan starts spinning, the hard drive is spinning all the time. That happens with your computer when its context switching between programs, you’ve got Word, Excel, PowerPoint, and you’ve got a browser with 17 tabs on it. The computer is doing its best to keep all of those processes alive. What happens is its run out of memory to do that, and its solution to that problem is it’s going to use your hard drive as memory. These days, we have solid state drives so this is not as big a problem. In the old days, it would be literally using your hard drive which is 100 times slower than memory. The operating system was paging things in and paging things out and your experience as the end user is, “My computer is now completely unusable. I’m going to have to reboot it.” That happens to people, too.
The equivalent is when you’re programming; you do not want to be interrupt-driven. You do not want another task getting swapped into your area of attention. This happens with the creative people, too. It’s not just the programmers, the graphic designers, or people who are thinking deeply about business strategy. You just need contiguous time to think through the problem, and any interruption is going to be nearly fatal to your productivity. Learning Fridays is a way to combat that.
By the way, if you ever wonder, “Why do computer scientists stay up late?” I bet you every good programmer you know is on the noon to 3:00 a.m. in the morning cycle, rather than the regular cycle. I am fundamentally convinced that they learned in computer science labs when they were undergraduates, that if they needed contiguous time and they needed to, to grapple with the hard problem sets and to write really creative programs, you needed uninterrupted time. I’m convinced that that’s why computer scientists skew late because from 9:00 p.m. to 3:00 a.m., that’s the golden hour. That’s when you can get in the flow and not be interrupted when you’re doing creative computer science work.
- Absolutely. Finally, I guess my final question for the episode will be: what’s the single most important trait for an entrepreneur to succeed?
- I’m going to go back to the answer that we had been teasing a little earlier, which is a lot of people talk about product-market fit. Your job as the entrepreneur is to find a product that people want and engineer the fit to be as close as possible. We’re convinced that in addition to product-market fit, we think a lot about founder-product fit. In the first hour that I spend with an entrepreneur, getting to know them, getting to know their story, I literally spend half an hour on their background. I want them to tell me their life story. I literally go back to, “Where were you born? What experiences in your life led you to this point where you are now the right entrepreneur to attack this problem?” I want to understand the fit. I want to understand why. I want to understand the motivations.
This idea of founder-product fit; are you passionate about this idea? Did it come out of a series of life experiences that feel genuine and organic, rather than manufactured? Ultimately, what I’m trying to get to is, “Are you the missionary or are you the mercenary?” The most important thing for me as an entrepreneur is, do you have the deep-seeded passion? Do you have to solve this problem as opposed to, “I could have been a product manager. I could have been a startup CEO. I think I’m going to get richer this way.” The most important trait for an entrepreneur is that they have a set of life experiences that has left them with the deep passion to solve the problem that they’re trying to solve.
To readers: which Frank’s insights did you find the most intriguing? What questions do you want to be answered in the future episodes?
P/S: For the previous episodes of ‘Princeton Startup TV’, check out www.princetonstartuptv.com