May 5, 2014
Building Companies that Devs & DevOps Teams Love… And Avoiding Expensive Mistakes
Orion Labs founder, Chef Co-Creator and former Amazon Dev Ops 'master of disaster' Jesse Robbins speaks to Heavybit members about the expe...
In episode 16 of Venture Confidential, we’re joined by Jerry Chen of Greylock Partners. Chen recounts his experience from managing and marketing product at VMware to investing in companies including Docker and Gladly as a partner with Greylock. The pair discuss what it takes as an investor and board member to help make a good company great.
About the Guests
Jerry Chen joined Greylock as a Partner in 2013 and is as a board member for companies including Docker, Blend, and Spoke. Chen focuses on investments in cloud and application infrastructure and new enterprise applications. He previously served as VP Cloud and Application Services for VMWare and was an associate with Accel Partners.
Peter Chapman: Jerry, welcome to Venture Confidential.
Jerry Chen: Thanks for having me. It's not really confidential, is it?
Peter: Welcome to "Venture Broadcast." Yeah.
Jerry: Perfect, thank you.
Peter: You know, I don't have exact numbers but I think our listenership is small enough that it's still semi-confidential at this point.
Jerry: Got it.
Peter: You graduated from Stanford with an engineering degree, spent a couple of years at a consulting company, a couple of years in private equity, went to business school and then VMware for nine years. Why VMware?
Jerry: Couple reasons. One, it was the smartest people I found. In my career I've always followed where the talent is. After Stanford, went to Bain because the smartest people where working at Bain in San Francisco then. It was in 1995-96. Netscape just went public. It was the dot com days and we were just attracting a lot of smart people.
And then after that, a couple tours in New York in private equity. A tour in venture capital during the dot com days, and then when I was at business school I did a internship for VMware and they were the smartest engineers, smartest business people, smartest product people. Regardless of the market industry or the job title, I just followed the smart people.
Peter: Tell me about those early years. VMware during this time seem to just be a breeding ground for future talent. We've seen so many cool leaders emerge from this decade at VMware. What was it like working there?
Jerry: It was amazing. I guess I was lucky. I helped create a couple product lines there, Cloud Foundry in later days and early days when they called it VDI, or Virtualized Desktop Infrastructure.
I think with VMware it was, A, a magnet of talent. Like I said, smart people attract smart people. And then B, the competition against Microsoft and other competitors like Oracle or Sun in the early days was a crucible. It was forced us to be ruthless about prioritizing our product features, our engineering, our go to market sales resources.
Because when we're doubling at head count and doubling in revenue every single year, some days you think it's easy, right? Sometimes, in many ways from the outside, it's like, wow, these guys are growing so fast that anything they do works well. But it's not by accident. Part of the market is sucking up inside the proverbial tornado, if you will.
The way you get to that exponential growth is, leverage on your people, your time, your talent and your resources. Every dollar, every hour you spend everyday, you got to make sure you're getting the most bang for your buck. And that's the difference between good versus great.
When you're early in your career you have a hundred decisions you make in a week or a month. Ten matter.
Ten decisions matter. So the difference between good versus great is knowing which 10 decisions to spend time, focus on and get right. Because the other nine decisions, either, one, someone else will make the decision, two, you can fix it later, or three, it's a rounding error.
So just like when I ran product, there's only 10 features of any product that matter. You get those 10 right, you're great. And any job running product, running sales, running engineering, if you get the right 10 decisions and get the right answers, you're going to do better than most.
Peter: How do you figure out what those key product decisions are?
Jerry: A lot of experience. I mean, IQ and intellect helps but early days you overcompensate for lack of experience, I know I did, by just working my ass off, really hard on all hundred decisions. And when you realize you're working 18 hours a day, seven days a week, either coding or running product or running sales, you're not working smarter, you're working more. Not even harder.
So partially it's experience. You learn what 10 product decisions matter. Partly you get mentorship, so you seek out folks inside the company, outside the company, that can give you some advice. And three, study what's worked in the past, what's worked in analogous businesses and other markets.
The hard part of studying the past or studying other products is understanding what are the right lessons to learn and not the wrong ones. What I mean by this is oftentimes as investors or product people or founders, we see success, we see failure, and we're pulling the wrong lesson. We're learning the wrong pattern or the wrong anti-pattern, and the truth of the matter is, oftentimes the driver for success is something a layer or two layers deeper.
It's not necessarily "the right channel," it's what the product feature was. Was it the right move versus competitor, was it the right pricing thing. And oftentime you stop, oh yeah, they price it correctly or they got the right developers to edit correctly. That's not true at all. Sometimes there's this second or third question you get right.
The difference is don't stop when you have what you think is the right answer. Ask one or two more questions.
If you're always asking these questions as a product person or as an investor or as a founder, you'll probably get the right answer eventually.
Peter: I'd love to walk through this. Maybe we can talk about a feature or a product you launched at VMware that had an obvious lesson and a less obvious lesson.
Jerry: Okay, we'll start with early days when they called it a Virtualized Desktop Infrastructure. I joined VMware '03 as an intern and then '04, '05, '06, I ran their enterprise desktop business. We had a product, a workstation which early, or it was VMware's first product. Developers loved it for QA, for testing browsers, in ran Linux on Windows and Windows on Linux.
We tried to create a product called ACE, assured computing environment, basically a workstation with security features. Encrypted VMs, your basic locality, so we do a secure hash, we can move the VM around. We had VM lifecycle, so it would time out and self destruct. And that product never really took off.
We couldn't understand why originally because we thought of all these features customers wanted. They were asking for them. At the same time, we had a bunch of other customers using VMware as a server product, it was called ESX Server at the time, now vSphere.
There's the bare-metal hypervisor that were hosting Windows virtual machines, desktop VMs, on servers and using a remote desktop protocol to connect to these VMs. And we were like, why? I mean, were they just hosting these VMs because of the cheaper licensing for legacy applications?
We could have just stopped there and said hey, it's just another use case for the server product. But I spent a lot of time talking to the customers asking what was the driver, why were they hosting virtualized desktops on VMware? Why weren't they using Citrix or some other product to host terminal sessions?
It came pretty clear that, one, for a bunch of desktop applications and for all the idiosyncrasies that Windows has, hosting individual desktops made more sense for app compatibility. And also people wanted their own workspace, their own VMs, their own world that Windows especially lent itself, was designed around a single user, not multiple users.
By asking the next two or three questions of these use cases, we basically said, there's actually a market for hosting desktops on the server, we called VDI, Virtualized Desktop Infrastructure. We built that to all business and Amazon years later is now hosting workspaces on their Amazon cloud in much the same way.
It could be easy to say hey, let's stop the first question or two. But if you dive deeply talking to customers and talking to your engineers and understanding how the product was used, you actually found value that was not obvious at the time.
Peter: You had a brief stint in venture. You worked at Accel for a couple years before business school.
Jerry: That's correct.
Peter: And then a decade later, you come back to venture. I want to talk about both bookends here.
Peter: What were those two years like?
Jerry: It was the weirdest time to be in VC. I joined Accel in 2000, the peak of the dot com mania. Term sheets were going out the door, there were pet sock commercials on the the Superbowl and it was the peak of the dot com mania, if you will.
Companies were going public, no revenue, losing tons of money. It was amazing. And then I saw the crash. When the Nasdaq, the market crashed. I watched a bunch of companies, start-ups that had great ideas or great technical teams, couldn't get funding, went out of business.
For me in that two year period, as an associate, I was able to see what it was like at a super bull market and a super bear market. Which is great because you learn both success and failure in investment business, both wild success, for almost no reason, and then failure for almost no reason because the market's bad.
At the same time, at Accel I was able to work with some really talented investors. Peter Fenton and Theresa Ranzetta and Jim Goetz were all there. I had worked with Theresa and Peter at a previous job in consulting and so that relationship brought me to Accel. And so with Breyer, Wagner, Schwartz, Patterson, that whole team there was a bunch of great investors. I was surrounded by smart people and I saw this incredible bubble and deflation of the bubble.
But after that, as a young associate, I was 24, 25, trying to invest or revise founders you realize as someone whose never had a real job, never shipped product, how can I be a good investor? How am I a good picker of companies? How can I be a good board member advisor, and why would you cross the table, Peter, pick me as your board member?
I realized the VCs and the people at the board meetings I went to that add a lot of value, actually had experience shipping product.
And so I went to business school, spent some time learning the language and grammar of business there, and then joined VMware for almost 10 years where I was able to see that rise from 250 employees to 15,000 and from like $15 million to $5 billon in revenue and to ship tons of product.
I understood product, I understood engineering, I understood go to market, understood partnerships. That would have made me a better executive no matter what I did, but also I think makes me a better investor and a better board member.
Peter: You're telling me that your 10 years at VMware was just your own way of getting ready for venture?
Jerry: Yes and no. I would say no job prepares you for VC, VC prepares you for no job. Because you never go to VMware for 10 years or any number of 10 years and assume you want to go to venture.
Because after I left Accel and joined VMware, I'm like, "Hey this is great. I love shipping product. I'm never going to do venture capital again." You don't take a job if you just want to be in VC, and anyone who's curious about it, you can never predict these careers.
What are the odds you do two years at one VC firm, spend a decade at a company, had to go public in the New York Stock Exchange and then join another great VC firm? You can never predict these things from the beginning. But I got lucky. You know, I got lucky working through great people at Accel, great people at VMware and super fortunate now to work with amazing partnership at Greylock which I hope to spend the next 10, 15 years in my career with.
Peter: What brought you back into venture?
Jerry: It's a great question. I think my partner Reid Hoffman calls it "tours of duty." And as you think about your career, I think about my 10 years at VMware in two major tours of duty. The first working for Diane Greene creating the desktop business. The second tour of duty, the last four or five years was working for the second CEO, Paul Maritz, creating Cloud Foundry and the whole applications platform team that had vFabric, Spring and a bunch of other products in it.
At the end of that tour of duty Paul spun off, we spun off a bunch of my team to Pivotal. A new CEO was coming in, Pat Gelsinger. He's done a great job and that's kind of the logical bookend. I did two tours of duty and the question was, do I want to do a third at VMware and do something else, which I could've done.
But I said, you know, after 10 years, I wanted a new challenge. So I took some time off. And during that time I traveled, I re-established some hobbies, talked to a bunch of startups, talk to a bunch of VC firms and I never intended to go back to VC but I had a relationship with a bunch of venture capitalists over the years.
I got close to Accel folks early, but in particular, I got close to the Greylock partners and I knew Bhusri in particular. Aneel was a general partner at Greylock, also he's founder and CEO of Workday, one of the definitive enterprise SaaS companies out there.
Aneel and I connected, and the way to describe Greylock is a very special place. All the partners around the table are operators and product people. Reid Hoffman created LinkedIn and PayPal, David Zee was early product person at Excite@Home before he became a VC and invested in Facebook and Pandora and LinkedIn. Aneel was a product person at PeopleSoft and then he created Workday while partner at Greylock.
It's kind of like, my partner John Lilly would say like, "your tribe." You know, it's the DNA. We all think about the world similarly. I think Aneel described it where we're entrepreneurs with a VC checkbook, and that's how we think about things.
That team felt right for me and it felt like it was a partnership that I want to be part of that I could get practice and practice VC, the job and the career and the art, if you will, the way I wanted to. Which is basically being thoughtful, being a great partner to founders, investing early in enterprise software companies and helping them grow from zero to hopefully multiple billions of dollars.
Peter: What were those early days at Greylock like for you? You had a pretty early investment in Docker, right?
Jerry: That was my first investment. Yeah, in 2013.
Peter: Yeah, great choice.
Peter: We've spoken a bunch on this podcast about that first investment, which as I understand it, is often one of the most stressful investments in a VC's career. How did you land on Docker?
Jerry: I would say two things. First, the partnership's super supportive and the way they think about it is, don't overthink your first investment. Think about your first four or five investments.
Think of it as a basket of four or five projects, because if you think of it like your first investment, you're going to overthink your first. It's not going to be your last.
The best way to think about it or get over that fear and anxiety of your first investment, which is still going to happen because it's your first investment, period, but if you think about, look, this will be one of four investments, one of five investments I'm going to make over the next two or three years, some would do better than others. Then you place less emphasis, less one-itis if you will, on the one deal. So that kind of was a great mindset going into it.
Secondly, as I looked at Docker, two things. One, I was familiar with what Solomon was building, which was similar to what we were doing with Warden containers at VMware Cloud Foundry was really inspired by dynos at Heroku. Obviously that was all built upon a bunch of what Google did with Linux at Google, and Solaris Containers and on and on. So it's a long history of these technology patterns, if you will, studying which happened in the past.
I remember I called up Ben Golub, the CEO of Docker at the time, Solomon was founder and CTO. I said, "Hey Ben, I'm Jerry at Greylock and I want to talk about investing in Docker." And Ben said to me, it was November 2013, he's, "I'm going to Hong Kong for the OpenStack Summit next week. Are you going to be there?" I said, "Yes, yes I am."
Hang up the phone, basically bought a flight and a ticket to Hong Kong next week, chased down Ben and the Docker team in Hong Kong at OpenStack Summit and convinced Ben that he should raise money from Greylock and yours truly. And so that's how my first deal happened. Basically in finding a great technology like Solomon and chasing down the CEO halfway around the world at a conference.
Peter: I love the persistence. You've had a pretty consistent track record of enterprise technology at Greylock. Notable to me is Docker and of course, Cloudera. You have a penchant for investing in businesses built on top of open source?
Jerry: I say that what I'm looking for, Peter, is you're always trying to ride a wave bigger than yourself, bigger than your competitors, right? And what I mean by that is what I'm looking for as an investor is just changes in distribution, more than anything else. So you think about, real quick before I get into open source, mainframe to client server, you're selling software not to mainframe but to a PC.
From client server PCs to the browser, now we're selling software through a browser. Now browser to cloud mobile, now we're selling software at ads from in the cloud or on a screen that you're holding on your smartphone.
What's most disruptive for a startup is riding new changes in distribution that the incumbents can't, is hard to follow. Because incumbents, they're the victims of their own success.
They're used to selling software through a PC channel or a browser channel and so these new channels or new methods of distribution are really powerful.
Open source, be it Docker or Cloudera or all the other projects that we've worked at Greylock, Greylock actually was the first investor in Red Hat back in the day, my partner Bill Kaiser. Open source is a new way of getting software in bits into the hands of their customers. And that's a wave we believe in and ride much like we believe in mobile for consumer or for other software deployments.
And so if you think about what I've done at Greylock, both on the infrastructure side and cloud side, think about how open source changes how you get software. How the cloud changes how you get software like databases, or security software. And invest in company called Cato Networks that's software defined, wide area networks. Which is incredible company because it changes how you think about purchasing networking from location to location.
Or, I think about some of the SaaS companies invested in, be it Gladly, Spoke, or Blend, they're all cloud-hosted. So, again, changing how your consumer software and they're changing how you get software from you as an enterprise to your end-user. And an end-user's going to be a consumer.
The case of Blend, it's mortgage application software hosted in the cloud. Or it could be an enterprise IT buyer like Spoke or it's just going to be your CIO or VP of IT. I think a lot, Peter, about distribution and channels. How do you get technology in bits in the hands of your users? That could be a developer, it could be a an IT buyer, it could be a CFO, it could be a VP of Sales.
Peter: Okay, so when you think about open source, you think about it as a distribution channel. Like, this is a really cost effective way of distributing yourself.
Jerry: Open source is one of the ways. It's A, distribution, to reduce the friction in cloud awareness and trial. For sure. Number two, it's also leveraging called the wisdom of the crowds or the wisdom of the developers. So instead of having a single company or developer driving your roadmap or driving your decisions, you can read the signals of the open source community.
In terms of going deeper on these questions, it makes it easier for you to kind of figure out where you are customers want to go and ask those questions, because you have a whole community leading you. So on one hand,
it just reduces friction across the board. Reduces friction on trial, adoption, reduces friction on customers giving feedback. Because the open source community will lead you.
Peter: You touched on a theme you wrote about a while ago, the three enterprise buyers. I'm going to mis-remember the name of the article.
Jerry: Something like triangles, yeah.
Peter: The triangles, right. Jerry's triangles, in which you say, "These days when you sell to enterprises, you're selling to three people. You're selling to a CMO, a CIO or a CTO, and a CFO." Give us just a brief summary of how these three players work together in an enterprise buying decision.
Jerry: So as you guys know me, Peter, you realize I think of the world, apply frameworks at everything, from investing to your startups to how I run product. And when we had three buyers is really just even a layer deeper, there's three systems of record in any company.
So your company, from your pizza stand to your Fortune 100 auto manufacturer, there's really three systems of record that make that company: your customers, your employees, and your assets. Those are the really the parts of the balance sheet, that's your company.
As a buyer or a seller or a startup, you're really selling to the CMO, the VP of Sales, right? So who owns your customer record. Which is basically Salesforce or CRM or trying to own everything pre-sales or post-sales. But the customer record is who you're selling or who you care about.
Your employees, is you're selling to the either your IT person or VP of HR, VP of Talent, it's like you're selling a system record around your employees. So that's everything from hire to fire. Recruiting, training, bonuses, etc. Managing your employees because that's the asset for any knowledge economy.
The third thing you care about is your assets. That could be IP, it could be physical assets, it could be financial assets if you're an investment bank or a commercial bank. And so when you think about those three records, those are where a lot of software companies and value have been created. Like HR software, Workday. CRM software, Salesforce. Financials, SAP and Oracle, or ERP and Oracle, right?
So as an investor and as founder you want to break down, okay, if I want to build a really valuable company, I either play in one of those axes of this triangle or navigate the intersections between these.
Peter: So the two investments we talked about, Cloudera, Docker, are both in the third prong of this triangle?
Jerry: IT buyer.
Peter: IT, assets, helping companies manage infrastructure. Tell me about a play you've made in one of the other legs of this triangle.
Jerry: Sure. On the sales side, I invested with a company called Gladly that's talked about B2C customer support, business to consumer help desk. So today, I think about more what channels you have.
When your plane is delayed, you either tweet at United or American Airlines or Delta. You maybe text or maybe call them, maybe go to their Facebook page and the customer experience across a bunch of these retail channels is broken.
Because one, it's ticket-based or problem ticket-based, not Peter- or Jerry-based, and it's just fragmented. Gladly is basically trying to create two things. One, an omnichannel, the customer support experience. It's like, you're not a number. You're Peter. You're Jerry.
Number two, it's trying to apply some level of machine learning or intelligence to basically solve a bunch of these problems for you. Because a lot of these questions from a delayed flight or tracking a package either have been asked a million times or could be automated. So Gladly is trying to be a modern delight for the customer. And JetBlue was our first customer trying to fix that airline problem because, I think you know, that's one we've all felt firsthand.
Peter: Great. I want to bring it back to Jerry's career. When I think about you sort of re-entering venture, series B in Docker makes a ton of sense. You have a decades worth of experience in enterprise software, you're familiar with containers. This is straight up your alley.
Tell me about the decision to venture into a slightly more foreign domain. I'm imagining that you could've just made enterprise software bets, enterprise IT bets, for your entire career. What made you want to explore other forms of enterprise software?
Jerry: I think it's a combination of where my interests grow and lie. There's a bunch of great technologies and how AI's changing application software, SaaS, and how mobile is changing and how big data is changing it.
And number two, here the great founders kind of lead you everywhere. So a lot of VCs say, "Hey, I'm going to do a deal in the space," or, "I care about this space." The truth is, we are often led by the best founders. When a great founder walks in your door, you're going to know it and you're going to spend time with him or her.
But I think what happened for me is I understand technology from my decade at VMware and then I spent time like at Cloudera, so I understood the value of data. That also said, it's how data and infrastructures changing, how they build applications.
And so when you combine my knowledge of Docker and VMware, plus the early experience at Cloud Era and how a big data is changing how you have insights in customers that leads to machine learning AI, you realize, look, at the end of the day, IT exists to support a business process.
It could be order to cash, hire to fire, right? It is a business process. You don't build IT for IT's sake. What I was seeing through these investments where the processes, the workflows in a company were changed dramatically by one, going cloud, going cloud-native how you build those applications, and then how to use data.
One of the insights was I was seeing how these new applications were being built. Not just big CRM applications but, customer intimacy, support applications, IT help desk applications, vertical SaaS companies attacking construction, an investment I made in Rhumbix, or healthcare or real estate, and when you see where the develops are going, you see how data can change these applications that lead me to these investments in Gladly, Spoke, Rhumbix and Blend.
Peter: So you're pointing to an invisible theme across your investments that you've written about, which is data and machine intelligence as a defensible layer.
Jerry: As a moat, yeah.
Peter: Yeah. You wrote this great article last year that changed the way I think about data, called "Systems of Intelligence" in which you posit that companies are moving away from IP and high switching costs as moats and towards data and the flywheel effects you can create from data and machine learning.
I'd love to walk through an example here. Can you talk to us about how data allows a company like Rhumbix to build defensibility?
Jerry: Sure. I think this framework, the system intelligence came from a conversation we're having with smart founders and we're talking about how, I think at the time my blog was The New Moats, and we're sitting around saying, wow, the old moats' defensibility, like economies of scale or secret IP or deep IP, are getting harder and harder to sustain.
The reasons were open source and cloud are changing how you monetize your IP. And so selling with a widget, you're selling a service for the widget. And number two, the economies of scale are so huge now because Amazon and Google and Azure allow anyone to deploy an application quickly and they want to compete against Amazon itself. That's a huge order.
So all of us started thinking where the moats are moving towards new areas, and historically the two areas where, on top of scale and IP, people built what was called a system of record, a database to just talk about the three systems of record, you know: people, assets, and customers.
Or systems engagement. The user experience between your developer or your user and the application that was Windows. Or the browser wars which I lived through. That was super stressful for Microsoft because they thought Netscape and the browsers were going to be the system of engagement for software going forward.
Then more recently mobile, right? Android versus iOS. Mobile became the system engagement in how you consume applications, and obviously that lead to the smart war battle we're still going through now.
But in between the system engagement and the system of record lies what I call the system of intelligence. Which I think is a very fertile place we're a startup to play in because one, it can either build a data set in a vertical, like construction like a Rhumbix and no other, no one else has. Google's not going to have the data that Rhumbix has around construction and productivity.
Or the data blends are going to have your more softer applications. Or you build healthcare verticals, real estate verticals, construction verticals. Data that the incumbents aren't going to have. If you're building an app that identifies cats or dogs, or "hot dog or not hot dog," you're not going to do better than Google or Facebook.
But if you're building a construction vertical software like a Rhumbix or Blend in real estate or healthcare vertical that you've done as well, that's a natural place where a startup can create a moat.
The other way to create a moat or a system of intelligence around data is crossing multiple systems of record.
Which, you see them as maybe not intuitive at first, but really is when you dig deeper because Salesforce is going to build a system of intelligence on their own CRM. Einstein, right? Or Oracle tried to do that, to have their own applications on top their own data. But there are certain categories that are problems inside a company that it can build intelligence across multiple databases, multiple applications, you can actually create another moat.
That lead to my investment at a company called Spoke that's doing kind of IT help desk or internal ticketing across IT HR facilities. So, asking questions like, "I need a new computer," to, "What's the best coffee place nearby?" to "How I can contribute to my 401K?" That's three or four different systems of record that I need to query right there. So talking three or four different applications, that's difficult. That's friction for the end user.
Spoke will create a friction-less system to talk to all three databases and give you your answer. And so when we find a way to either go deep in a vertical where you're building on data no one else has: construction, financial services, healthcare, etc. Or go anywhere across databases or across some records where no one incumbent is going to dominate, that's another defensible place.
It's natural for a startup to play across multiple incumbents. Otherwise you might as well just buy public stocks in Workdays, Salesforce, ServiceNow, etc. Because they're going to own their own siloes. They had to go across siloes to find a new defensible mode.
Peter: Either you're going across siloes or you're picking a niche.
Jerry: And going deep. So I think those are the two logical ways around system intelligence. Go deep on a market or vertical or a niche, which I think is a logical evolution of the cloud market. A bunch of horizontal apps now vertical apps.
We're seeing that from a, look, self-driving cars is an expression of a vertical niche, right? Logistics in transportation and collecting data around maps, driver data, etc.
I mean, you look at all the startups in the self-driving cars space, and then comes Waymo or Uber, They're collecting data with cameras and maps that other players aren't, that Ford doesn't have or GM doesn't have. Those incumbents are trying to catch up.
So again, that's an example of a niche where you're collecting data other people don't have or you go across verticals, across siloes, across other natural barriers.
Peter: I want to shift the conversation back to you a little bit. You've been at Greylock for almost five years now in your second tour of venture duty. What's changed in those five years?
Jerry: It's interesting, Peter. Every year I think about how to get better at this job. And there's different ways to try to get better at a job. It's, how do I find great investments and great companies?
What thought leadership can I put out in the world, like new moats or system intelligence that would attract founders? Or how do I sell myself in Greylock better? How do I influence founders better when I'm on the board? And then how do I work better with other investors around the table?
So, I think every year there are different aspects to the job you can work on. I think the core investing judgement's tough because the investment lifecycle is so long.
Five years, eight years, 10 years, for these exits so you can try to improve upon your judgement every year trying to see what investments you passed on that you should have leaned in or what investments do you do that you probably should have avoided. I think a lot about that.
I think a lot about the mechanics of doing the job better as an investor, as a board member, as a mentor to the founders. Then I think about different spaces. This framework of system intelligence didn't come out of nowhere, it came from me analyzing the market and thinking about what attributes would make a great company and not just a good company, but a great one.
Because the hard part of this job is there's a lot of good companies out there that you can invest in and, as a partner at Greylock, I'm lucky enough and fortunate enough to meet a bunch of great founders and good companies. But we're looking for that exceptional company that can be 10, 20, 30, 40, hundred billion, right? The LinkedIn, the Facebooks, the Workdays of the world, and those are special. And so really trying to separate the great from the good, is hard.
Peter: You mentioned four fundamental skills of venture capital. You said marketing, marketing both yourself and the firm. Finding and diligencing investments, helping founders, and working with other folks on a board. I'd love to pick one of them that you've been focusing on and sort of talk through your own educational experience. Is there one of those that's been particularly prominent for you recently?
Jerry: The past couple years, Peter, a lot of the joy in learning has been the last two. It's how to be a good mentor, advisor, board member, to the founders and executives, and how to be a good partner around the table, not to just my partners at Greylock but also the other investors within a company on the board. Because it's all influence.
As a board member, you don't run the company. You don't tell the CEO what to do, you give advice. You ask questions and you realize that every founder and every executive you work with reacts differently.
The way you communicate with one founder is going to be different from another founder. A three-time founder like Shlomo Kramer, and how I work with him on Cato Networks is very different then a first-time founder who dropped out of this MIT PC program to start a company. Because they have different sets of experience and different things you need to learn.
The past couple of years I've been learning how to work with different types of founders at a different stage of their career. And that's been a joy because investing's great. It's intellectually challenging exercise to think about trends, technologies. But it's really the joy of working with these teams and companies and helping these organizations grow, has been a lot of fun.
Peter: What does it mean to be a great board member? What sort of behavior do you appreciate in your fellow investors?
Jerry: I think as a board member, and I think when I was at VMware, if I were a founder running a team, what would I want from my board? The key is make sure you don't lose the forest for the trees. Is, let's figure out in every board meeting and any given quarter,
there's only two or three issues that I think I really make or break the company, that you should be thinking about.
The board's job is not to get in the weeds of the product detail or HR organizations. They can give you advice, they can ask questions on the product reviews, but at the end of the day, there's really two or three questions at every board meeting that should be on top of mind.
The way I like to be a board member or a partner to the founders I work with is like, "Hey, ask us two or three questions." Challenge them, challenge the founder, him or her to think about them, and push them.
Our job is to worry about risk, what could go wrong, what could go right. Our job as board members is to push to be more aggressive or pull you back when you're being reckless. But really it's, what are the two or three questions?
And these go back today first framework I said, good versus great at your job or success is what 10 features matter in the product, what 10 questions matter at your job as a CEO day in and day out. As a board member or investor, what two or three questions matter right now? And if we can answer the right two or three questions the rest will take care of itself.
Peter: We've heard a lot of this podcast about other relationships between investors and founders, what it means to be mentor, how to influence a founder, how to adapt to your working style to fit that other founder.
We haven't heard a lot about relationships between investors, and I'd love to pry into this a little bit. You mention that something to be good at as a venture capitalist is partnership, working with other investors. What are you trying to cultivate in that realm?
Jerry: It's interesting because some people say VC is a lonely job because you're making the calls, you're not making the calls, and they also often say it's interesting because venture capitalists are this weird frenemy relationship, to compete with VCs to win a deal but then you work with them on the board.
The way I like to approach it is, every investment I work on, you have a team. And to make it less than that team is you and the founding team, but also you and the other investors. And so when I join a board, I have another person join me. You and me, Peter, that's my team for this company, this new co we're working together to make new co successful.
So when we're invested together, we're actually on the same virtual company, the same team. That's the attitude I try to think in that room, in that board, with that founder, solving that problem together. Then on another company, I could be trying to win another investment, another series A deal, and we're competing. This is how the industry goes. But it's interesting to compartmentalize project to project, company to company.
By being a good partner to another investor and to the CEO, you talk about the founder-investor dynamic. But also it's like, hey, how do you and I as two board members in the same company communicate? We have the same objective, make this company successful, how you and I communicate.
So, you and I had the same concerns because you and I have the three, your top three concerns, are just my top three concerns then something's wrong. And so, either I'm not seeing something, you're not seeing something. So let's have that conversation.
We're the most powerful as an advisor to the company if we think the top three issues are the same.
Like hiring new VP of sales. Like increasing the product velocity, hire a new VP of engineering or do a different partnership. But if you and I don't have the same three priorities, the same three questions, we're going to be less effective as board members. But mostly if we're seeing different things, great.
You're closer, you see other aspects. You have a different relationship with the founders than I do. And I'm going to learn from you. And I try to learn from my other investors. I learn from other board members.
My first board was Docker. I invested be hidden Peter Fenton who I knew from back in the day, and he did Docker at Benchmark. And Peter's a fantastic investor and fantastic board member. And so just learning from him on the Docker board has been educational for me.
I'm always trying to learn from the folks around me because they have a different experience set than I do and they're going to see different things. So I think that's what it means to be a good partner. We're all the same. Why would you use a founder, pick three VCs that all thought the same, look the same, taught the same, talk the same?
You want diversity on your board, but also you want a functional board. People that can actually communicate and work with you.
Peter: So there's a balance between diving into disagreement and finding the alignment that you need to be a good board.
Jerry: I think you want all things, a creative tension. You want your product head and your engineer head and your company to have a creative tension. Or your sales and your engineer to also have a creative tension. Because I think that tension, like all things, creates the best result.
Just like as a founder or CEO, you probably want folks around the the table to help you make the company better. And that's going to have tension between the founder and the investor, because they're going to push you, they're going to ask a bunch of questions, and you want investors that all see different parts of your business.
From a technical side to the business side to partnering side. You want that tension, at least it's not disagreement or fighting, but you want people to be pushing at all aspects of your business because there's multiple muscle groups you need to exercise. And when you have different board members working at different muscles of your company, then you're going to be stronger.
Peter: Is there a particular muscle that you tend to focus on?
Jerry: Oh gosh. I think the muscle varies in the lifecycle of the company. So I think when I do a seed investment or a series A investment, it's two or three individuals, PowerPoint, and maybe some code.
I think there it's hiring, that muscle. So it's team building and it's finding the right customers for early product-market fit, iterating on the product and hiring the right people. That I think you know, call it a series-B inflection point where you're about to go to market then that's all about exercising go to market muscle.
Pricing, packaging, channels. What is the right marketing message? What's the right way to build a good market organization? And then if you're lucky enough at a series D, series E in a company scaling, there's a whole bunch of other issues around, upgrading the team, upgrading your products, how do you become a truly global organization?
My partner Reid Hoffman also has a podcast called Masters of Scale, where it talks about the different phases. I think goes: family, village, nation, etc. Because there's different challenge to different phases. And so I'd say there are different muscles when you exercise a different phases of the company.
I was lucky enough to see multiple phases, several of them, couple hundred people to 15,000 at VMware. And my work at Greylock, I've worked from two founders, a single founder, to becoming a series D like Blend where there's 500 hundred employees and the good investors and the good board members know which muscle you should work at at any given time.
Peter: I ask all my guests the same question in closing, which is, what do you wish you knew going into this that you know now? What advice would you have for your younger self?
Jerry: If I could talk to myself five years ago when I first started out at this tour of duty with Greylock partners, to realize that what works for me is different then what works for other VCs.
So, how I build a relationship with the founder and get conviction on investments and build that trust and invest in that company is going to be different from other partners at Greylock and other VC firms. And to realize what you're good at, what your strike zone is and just stick with it and don't worry about everyone else.
Peter: Jerry, thank you so much.
Jerry: Thanks for having me at Venture Confidential.
Peter: Where can our listeners find you and who should be getting in touch with you?
Jerry: You can find me at Greylock.com, website, or Jerry@Greylock.com just to email me. Or I think @jerrychen on Twitter, you want to see my blogs, musings, and random comments on what's happening in the world and that's probably a good place to start.