Keith Rabois is a General Partner at Founders Fund where he focuses on consumer Internet, education, enterprise, financial services, and digital health investments. Keith began his career in the industry as a senior executive at PayPal and subsequently served in influential roles at LinkedIn and as chief operating officer of Square. Join him as he shares his top lessons in building great teams.
Want to see more content like this? Join us at SaaStr Annual 2020.
Lloyed Lobo, Co-founder @ Boost.ai
Keith Rabois, General Partner @ Founders Fund
FULL TRANSCRIPT BELOW
Lloyed: Is everyone having fun so far? I need a louder noise here. We’ve got one of the best speakers in the house, Keith Rabois. Give it up, ladies and gentlemen.
Lloyed: So when they asked me to do this session, I was in complete awe. I’ve been a fan of Keith since 10+ years, and as soon as I saw him I completely forgot his entire history. And I’m asking Keith, I’m like, “Tell me what you did.” He is one of the most well-known people in the Bay Area, one of the most well-known investors. Early teams of PayPal, LinkedIn, Square, and now GP at Khosla Ventures. How are you feeling today?
Keith: Great. It’s great to be here. Happy to talk about how to build teams, because that tends to be the number one question I get asked by founders.
Lloyed: Awesome. And you’ve been on the early teams at several companies now: LinkedIn, PayPal, Square-
Keith: Started Opendoor.
Keith: From scratch.
Lloyed: Yeah. And now with Khosla you see thousands of companies and you’re advising them. Hundreds.
Keith: I don’t know about thousands, but tens to hundreds.
Lloyed: Tens to a hundreds. And you’ve seen teams from all the way to seed stage til exit. So tell us, perhaps walk us all the way back to your journey in the early days. What does a team look like for a company? What is the minimum viable team? A company’s just starting out. They have an idea.
Keith: Well, it does matter depending upon what you’re trying to do in the world what the ideal composition of the team is. So let me backtrack for a second. When I was an executive at Square, the new Khosla … founding partner of Khosla Ventures was on my board at Square, and he had this expression or adage that the team you build is the company you build. And at first, it didn’t totally make sense to me. But then it locked in my brain and it really became the operating principle for when we invest and how I invest, which is people focus a lot on products and markets and technologies, but ultimately the only thing that really matters is the people.
Keith: And really we learned this lesson at PayPal. We had an incredible density of talent, a pretty cool density of talent in one building, mostly in Mountain View. And it was our growing sort of mission, like that was the company that absolutely should have failed, but for the extraordinary efforts of a couple hundred people and really 20 to 40 people that were absolutely amazing. And so you can get distracted and in love with this awesome product or this great design or this breakthrough technology, but if you don’t have the people that can take advantage of that and leverage that particularly for five to 15 years, you’re not gonna build an iconic company.
Keith: So as an ambassador and as a board member and now as a BC fund, we obsess on the … what is the core team. And what is the core team for this specific type of project? So for example, if one wanted to compete with Elon and put rockets into space, you would want a very different team than if you were building a competitor to Facebook. There might be some Venn diagram overlap there, but there’s definitely some differences in the ideal composition of the team.
Keith: The other thing someone said to me that really resonated and has become an important lesson is you can raise the odds of success at any startup from the very early days from something like the proverbial one to 10 percent that people talk about to probably 30 to 40 percent just by changing the team composition. We’d even joked a little bit that founder’s fund should sort of rebrand as co-founders’ fund, and what they should actually be doing or what someone like them should be doing is matchmaking founders with other co-founders that have the missing DNA.
Keith: So for example, sometimes we’ll meet entrepreneurs when they’re really the two kids in the garage sort of thing. And we’ll say to them, “This is a great idea. You guys are awesome. But you really need quote-unquote co-founder. You’re missing some skill that’s indispensable to success for this company and we’re gonna help you go recruit this person, but you’re gonna have to allocate real equity” … and that’s why we call them a co-founder, or founder … “otherwise you’re not going to be able to get someone with the appropriate level of talent.”
Keith: And so in today’s world that can mean AI co-founder if you’re building a company that leverages data science. Or it can be an engineering co-founder or hardware co-founder or a business co-founder even. But what we’re doing is helping diagnose the difference, the delta between where’s the team today and what success looks like and what are the key variables that the team itself doesn’t really have good sort of air cover on. And one thing a good BC can do is have a network that may have some of these people sitting within it.
Keith: So typically when people start companies, depends on how old and how experienced someone is when they start a company, but typically they have a network that looks a lot like themselves. And that’s really good when you’re scaling part of the team that looks like yourself. So if you’re an engineer and you need more engineers, having a network of engineers is an awesome thing. But when you need a CFO, let’s say, or you need an outstanding VP of engineering or maybe a VP of sales, which is even more different to find for an engineer, you don’t know who these people are. You don’t know where to find them. You don’t know how to interview them. You don’t know how to assess them.
Keith: And that’s one thing a good BC, board member, can do is I have a network of all of those people and I sort of have done reference checks on many of them and can kinda triangulate, “Well, here’s a good one that would be a great fit for you. Why don’t you interview this person?” And so that’s what we try to do is a lot of that matchmaking with an early team and a great idea and the missing pieces.
Lloyed: Awesome. Let’s shift now to the dynamics, or rather the stages of the startup, right? What’s the smallest deal you’ve done? Because you’ve done a lot of angel investments in your time as well.
Keith: So from 2003 to 2013 before I joined Khosla Ventures, I was a pretty active angel investor in Silicon Valley. I probably invested in 80 or so, maybe 85 companies. Those would be fairly small investments, ranging from like $30,000 to $200,000. Kind of a classic angel Silicon Valley type. As soon as I joined Khosla Ventures six years ago, rarely would I invest less than 500K. Typically it’s a million to three million, and I would really call that a seed.
Keith: So we like to do seed investments at Khosla. We like to get involved as early as possible, so we like to write one to three million-dollar checks as frequently as possible, partially because we can help with the team-building and help ensure that the foundation of the team is as strong as possible rather than try to fix things later. So culture and team-building are a lot like concrete. So when concrete’s in liquid form, it’s very malleable and you can shape it really easily. And once it solidifies, it’s a real pain. Like you have to take a jackhammer. It’s very noisy and expensive. So we like to be involved before the concrete solidifies. We don’t always have that choice and we do invest in Series As and Series Bs. But I’d prefer to be writing a seed check if I can.
Lloyed: Definitely. So from those 30,000 investments that I’m more interested in, let’s reverse-engineer a unicorn. I’m sure you have one or two unicorns in there, right? So let’s reverse-engineer the team of a unicorn at 30K. Maybe you can walk through an example.
Keith: Yeah. I think typically the most important thing in your earliest possible stages, so whether as an angel investor or where you’re right at … I’ve probably invested in 15 to 20 unicorns, as a BC probably five to 10 so far. The thing that stands out is there’s at least one founder who has this ridiculous spark. You know you’re in the presence of potential greatness. It’s like finding a high school basketball player that looks a little bit like Lebron James. You can tell that the person looks a lot like Lebron James in high school. Doesn’t mean they’re gonna be Lebron James. They might get injured. All kinds of things can go wrong.
Keith: But there’s this spark and the person may be only 18 years old or 19 years old, or they may be out of some field. Like I came as a lawyer, so they may be totally off central casting. But you see this spark that you just don’t see in normal people. And for me to make an investment early before there’s a product or before there’s metrics, there’s a spark and I just am saying … I don’t even know if I buy the idea, but now this person is incredible and I’d sort of be lucky to work with them, and so that’s what usually leads to a very early-stage investment, versus a more mature Series B, which is based upon fundamentals, business fundamentals, metrics, cohorts, all the classic analysis.
Lloyed: So I don’t know if you wanna take names, but who was one such unicorn? And maybe we can walk through their journey as they grew as a team.
Keith: I’ll give you a couple non-famous ones, because the revisionist history of the famous ones I’m not sure is that useful because people have so much sort of color commentary there. A couple really more recent ones. So I funded what would typically be a crazy idea, which is a new autonomous driving startup with a very different approach. Obviously, there’s lots of people who are doing autonomous driving startups, and there’s lots of funding and funded companies there, and this was only six to 12 months ago. And the approach may or may not work, but it was very obviously to me that the founder was a world-class founder.
Keith: And I was hesitating on investing, actually, because of the crowded space and all the capital that SoftBank and Google and other people, and Uber had invested. My Chief of Staff, he’s worked for me for about two years, came into my office and he’s like this brilliant at the time 23-year-old, probably, 24 maybe. Came to me and said, “Keith, why are you hesitating? Your whole philosophy is find an outrageously talented founder and a potentially big market and don’t ask any more questions.” He’s like “That’s your explicit strategy and you’re hesitating on this investment. What the hell are you doing?”And so I had him yell at me, and I was like, “You’re right. Like, why the hell am I hesitating? Okay, we need to invest.”
Keith: Another one, actually. There’s a company called Fair. It used to be called Indigo Fair. It’s not a moderately high-profile company out of Y Combinator about 18 months ago. It’s raised certainly over 100 million dollars now. And when I met the founder, the founder had worked for me at Square, so I knew him pretty well. But I didn’t really … actually both, two of the four co-founders had actually worked for me and actually soccer teammates of mine. But they pitched me the idea, and I actually didn’t really love the idea. In fact, my direct quote was, “This is not an obviously terrible idea.”
Keith: And so they wrote that down, and they sent me an email after the meeting and said, “When we become really successful, we’re gonna put that quote up in our office. Not an obviously terrible idea.” But it turned out I liked them enough that I actually gave them a check. And then I gave them another check, and another check. And now we’re the largest shareholder in the company and they’ll be a billion-dollar company. They’ll probably be the best YC company ever, actually.
Keith: But I didn’t love the idea. I later learned to love the idea because we discovered some things in doing the idea and pursuing the idea that were bigger opportunities and more related to the specific DNA and skillset of the founders than the initial idea that seemed at a superficial level. So sometimes you don’t have to really like the idea or understand the idea even to know that the people are still worth backing and then that gives you opportunity to embrace the idea later.
Lloyed: So what was their minimum liable team like? What was that early team and how did it evolve from the first check to the next check? What were some key things that changed there?
Keith: The core team at, let’s say Fair, had worked on Square capital, so the engineers and data scientists had done the underwriting model behind lending money effectively, lending money to micro-merchants and merchants to allow them to grow their business. So they learned how to use data to underwrite long tel businesses in the real world and a lot of what Fair does today is underwrite merchants. So what Fair does is it tells very distributed businesses, gift shops, et cetera what inventory they should stock and take some of risk so that if you’re running a local business you don’t have to think, you can just stock these mugs or something like that and guarantee they’re gonna sell.
Keith: So they had the data science capabilities in underwriting the risk of inventory and the long tel risk of merchants failing. And then they had the product-design chops from the core part of the Square business and the logistical competency of having run Caviar, which is a food delivery business that Square had acquired. So some of the core team had also run operations for Caviar. And obviously food has to be delivered perfectly on time or it’s very soggy. So the operational chops plus the underwriting ability, plus the product design chops were pretty ideal for this business, actually.
Lloyed: And how many people were they? Like two or three?
Keith: Oh, it was two to four.
Lloyed: Two to four.
Keith: Really three to four.
Lloyed: And then how did that team evolve between the next round of funding? What were some of the key next hires that they made? And how should we think about it as sort of seed-stage founders. I’m a seed-stage founder. I have an idea. I got some early traction. I’m not sure I’m at product market fit yet. I got maybe 100 people who like it who might be paying for it. Anyone I pitch to tells me, “well, show me more traction.” How does my team evolve? What are the people I need to hire in the beginning?
Keith: I think I would plot … I’d take a whiteboard and plot the key risks in order of degree of difficulty. So every company has two or three or four, hopefully no more than five really core things that you have to conquer and solve to be successful. So Elon wants to go to Mars and he almost surely has written out on a whiteboard somewhere or a piece of paper all the things that have to happen to put a human on Mars. And so you wanna decompose each of them and put a name against each of those five things, and that person you better have conviction is about as good in the world as you can get to solve that specific problem. And if you don’t have a name or you don’t have conviction around that person, you better go find one.
Keith: So that’s the key thing is … It’s like, okay, we’re here and we wanna go there. What are the two to five most likely problems that we’re gonna encounter? How do we address them or validate them as fast as possible? And do we have the right person who’s like what I call the DRI, borrowing from Apple, directly responsible individual, that’s responsible for delivering that? And if not, I gotta go find that. The earlier you have the three to five people mapped against the three to five most important things, the more likely you are to succeed.
Lloyed: And so maybe one of the things in the early days is I need to get the produce market fit, or maybe I need to get the two million ARR. And then you identity I’m missing someone on the go-to-market or growth side. How should one thing about those early hires? Say you were an engineering founder like myself, and you got one engineer, one UX person. You’re missing that whole element of growth or go-to-market. How do you think about that?
Keith: Well, certainly for a consumer product without consumer adoption, there is no startup. Like, that is always the number one risk for any consumer product, period. There is no substitute for that. 99% of consumer startups just are not gonna succeed because they cannot interrupt normal people in the middle of their day and convince them to change their behavior to put basically their new product on the home screen. So any consumer has 24 hours in their day. They already have a lot of commitments. They have to work, they probably sleep hopefully eight hours. They have a job, probably. They have friends, family, and other commitments.
Keith: So the only way you can be successful as a consumer startup is you have to rearrange a lot of people’s lives. You have to get them to substitute from something they’ve been doing for years and substitute into your new product. That’s pretty rare and very rarely works. So until a consumer startup proves that it can do that at least in some micro scale, there isn’t any value creation. So I would start with, truthfully, I would start with … I wouldn’t even launch the company without some degree of insight of how I’m going to … I wouldn’t launch a consumer company, to be specific, without knowing how to do that.
Lloyed: On the B2B side, probably similar learnings, because without-
Keith: B2B, you can brute force. B2B, you can absolutely brute force to a certain level of traction, which is you can hand-hold a certain set of customers and potential customers into solving their problems and hope that you can extrapolate from that and make it replicable. You can definitely isolate from your own experiences of products that are broken or market inefficiencies and triangulate. You cannot brute force a consumer startup.
Lloyed: You can’t. So let’s stick to the B2B side of things, given most of us are probably in the B2B space here. When’s the right time to hire … a few things. One, a VP sales. One, a COO. One, a CFO.
Keith: All slightly different and they do very by business type within the general, let’s say, enterprise businesses. Even within software-based, SAS-based enterprise businesses. A VP of sales is a misnomer. You need someone needing sales. Obviously you’re gonna not have a direct, complete self-serve kind of direct model, which is atypical that that works at scale. So you typically have someone quarterbacking sales, and that person doesn’t have to be a VP. The person is typically actually more like a sales manager, maybe even a sales hero where the person’s doing a lot of the sales or all of the sales, closing all of them him or herself.
Keith: And then you’ll eventually have two to five people running around, and they need someone to organize them and quote-unquote manage them, so you have a micromanager or executive. But you don’t need a VP of sales. You need someone who knows how to sell and can close deals. And then once you start showing that you can close deals, then you may replicate this and you may replicate it by GO or industry or some other dimension, but you don’t need an exactly … You may need someone who can get the appropriate attention of a decision-maker.
Keith: So if you’re selling five-million-dollar databases, you’re probably not selling to a very junior person in the organization. So you probably need a sales representative who’s capable of holding his or her own with a comparable buyer. So you are identifying who’s the buyer, what’s the price of a product, and making sure that whoever the decision-maker is, you have a representative of the company that’s appropriate for that level of sophistication and seniority. So maybe the CEO may need to do the sale because the only person who’s able to hold his or her own with a very senior person in a larger organization is the CEO. But that’s okay. Then you learn how to repeat and you replicate.
Keith: CFOs, you don’t need in most businesses for a very long time, until you’re roughly 10 million dollars in revenue, partially because most CFOs are only gonna add about 10% value. So we take the revenue line, multiply it by 10, that’s about the value creation level that they can handle. So one million times 10, you’re gonna be underwater in terms of what you pay a CFO.
Keith: There are some businesses, so I’d say Opendoor is one, Affirm is another, that use capital as a source of oxygen. So we take debt, we lend money to people for buying asset, in Opendoor’s case, and then we get more of it back. People need to give us that oxygen to be able to do this rapidly. So in companies that use a lot of oxygen or have very complicated capital structures or arguable very low margins, CFOs and VPs of finance, and VPs of finance and strategy can be very valuable very, very early. But those are rare.
Keith: Same thing with a general council. Typically you would not hire a general council until much later. That said, Stripe’s eighth employee was their general council. Given the regulatory complexity of what Stripe aspired to do in their ability to simplify, a desire to simplify some of the underwriting and regulatory requirements, having a general council in the first eight employees was brilliant, and Stripe probably wouldn’t exist today without the efforts of that GC in the very, very early stages. So depending, again, depends on what you’re trying to achieve when you need certain executives. But typically you would defer a general council or a CFO.
Lloyed: And then you were COO at Stripe. When did you come into Stripe, and was it the right time?
Keith: Square, but people get them confused often.
Lloyed: Square, of course. Great.
Keith: It’s the water, too. Just perfect. Anyway, I was the 20th employee, which is actually pretty early for a senior executive of any type, let alone a COO. Typically, you’re better off at that stage hiring operating executives until there’s too many seams or decision-making complexities that you need someone to roll up various trade-offs to. So growth versus risk, et cetera. Customer support and product. Things like that. Sales and product, sometimes. But the company needed a business-minded, somewhat literate financial services person to complement Jack’s product and engineering and design chops.
Keith: And so it was more of a complementary … find a complement, and the COO role was the right way to do it sort of in terms of being attractive to people. But typically … I have a whole different speech on when you should hire a COO and when not. Typically you’re talking 50 to 250 employees before that’s a likely move.
Lloyed: Cool. Got a couple minutes. I got a couple of questions for you. One, as you’ve grown … Now, you’ve come across this team. You’ve funded it the first check out of 30K written at a $500,000 seed round. They’re at a Series A or B stage. What does that team makeup look like, apart from the revenue. Say 10 million ARR, you wanna fund them in A round. What is the key team look like there? And then a follow-up question to that is internal promotions versus external hires. How do you think about it?
Keith: Well, since we have a minute and 15 seconds, I don’t think I’ll be able to answer both very well.
Lloyed: So let’s take the latter, maybe.
Keith: The other one. Internal promotions, external … It depends on the velocity. So every individual in a company has a growth curve and every company has a growth curve. You can only promote people internally if their personal growth curve is exceeding the company’s growth curve. So perversely, the faster a company grows and the more iconoclastic the company is, the less you can do internal promotions. People just can’t learn at this curve. Fortunately, most companies, even very mostly successful companies grow at more like this curve.
Keith: So you have a chance of keeping the people that are growing faster than the company and promoting them, grooming and mentoring them. Certainly less risk in the sense of cultural transformation, but you have to have a way of teaching and then trying to do that successful. You probably wind up at best with a ratio of 50/50, 50% internal promotions. That’s very healthy if you can do that ratio, 50% eternal hires. That keeps most things working pretty well, but it takes a lot of work to be able to survive at that ratio.
Keith: On the team-building, again, by a Series B, the metrics are kind of speaking for themselves. And so you can identify what parts of the team sort of need upgrading almost empirically. Series A, you’re still taking a leap on a story, a narrative about why we’re gonna change the world, we’re gonna conquer the world. And what are the gaps in that story is really where the hiring comes in? Like, what doesn’t seem credible? Why should I back this one team to change this whole industry, and where do I feel uncomfortable? And that’s what I would look for.
Lloyed: I’m gonna ask one question that’ll probably take 10 seconds. As COO at Square, what were one or two things you did to encourage that internal promotion? Or how did you do that?
Keith: Yeah, so we basically got people mentors. When I joined the company, we had 17 engineers of the 21 of us, and none of them had ever managed a single human being before. So they were all individual engineers. They’re all reporting to Jack, which is obviously difficult for the CEO, and we had to figure out how to get engineering managers. So we recruited a mentor to take our five highest-potential engineers and over a year turn them into engineering managers. So that’s how we survived.
Lloyed: And by the end of it, how many people were each of them managing?
Keith: Each of them did very well, so they each had teams of three to 10. One of them eventually became a director of engineering and managed the whole thing. But there was no other choice, really. We tried to get an external hire, but it’s like doing organ transplants. You’re subject to so much risk.
Lloyed: Awesome. Well, we’re out of time, but give it up for Keith Rabois.
Keith: Thank you.