Vinod Khosla on How to Build the Future

My name is Sam. Today, we're talking to Vinod Khosla. Vinod is the founder of Sun Microsystems and Khosla Ventures. He's been involved in the creation of dozens of billion-dollar companies, and I think he's one of the most interesting thinkers that I've ever spoken to about how to build an ambitious company and team, and everything else you need. So thank you for taking the time to talk to us today.

"Great to talk about it. I want to start with the very beginning and how to think about the idea and the mindset for a company. One thing you've said before that I really love is that there's a huge difference between a zero million dollar company and a billion dollar company. Maybe you could start with just explaining what you mean by that."

To me, when you set out on a journey, your mindset determines who you bring on board, how you approach it, what you set up, what deals you do, or which investors you gather. In a zero million dollar company, you're thinking a certain way to tactically achieve a small, short-term goal. In a billion dollar company, you start building from day one the company and the people you'll need to build the company.

One of the things people seldom realize when they're starting up is that you don't ever plan what you're going to do; you build a plan to make to plan. Who helps in that planning? As you plan iteratively, as you evolve your strategy and your tactics, that team, which I call the kitchen cabinet of a company, is the essence of what your company will become.

So one of my favorite tweets I like to send out is, "A company becomes the people it hires, not the plan it makes." That's grossly underappreciated and is the biggest difference between the zero million and the billion dollar company: the initial people you hire and your experience. It is the initiative people you hire, but also how you approach the initial tactics.

My other great analogy is: if you have a large vision, you're climbing Mount Everest. It's never a straight line. Nobody's climbed Everest in a straight line. You get to base camp, you get to camp one, camp two, camp three, camp four. If you have the right approach, you keep your vision – which is Mount Everest – but you're flexible about tactics as things change, as you zig and zag, when you pivot. These are all things on the way to staying with the vision.

Now, you can also do the same tactics without worrying about the vision. My big beef with a lot of investors is they want revenue; they want to meet the plan, as opposed to collect assets for this larger ascent to Mount Everest. You can clearly set up base camp where you get revenue stability, cash flow, breakeven, and the ability to raise more money. But if your goal is to get to Everest, and you still get the revenue, you might have 20 million, 50 million, or a hundred million of revenue, but it doesn't help you get to Everest.

You can take a little longer, a little harder, to get to the base camp that lets you get the resources to keep the journey to your vision. There's a huge difference, and in the team is the biggest difference, but there are also strategic differences when you're dealing with investors. In my view, investors matter a lot in this because you make short-term versus long-term trade-offs.

What percent of investors in Silicon Valley do you think are good long-term company builders? As I get in a lot of trouble for saying this among friends, I think 90% of investors add no value. My assessment is that 70% of investors add negative value to a company. That means they're advising the company in the sport of team building to find entrepreneurs. They're advising a company when they haven't earned the right to advise an entrepreneur.

Some of the junior people here, when they ask me about this, I say, "Hey, at this other firm, people, young people are going on boards." I say, "You haven't earned the right to advise an entrepreneur." It's unfair to brand someone just because you got an MBA and joined a venture firm. It doesn't mean you're qualified to advise.

Now, the biggest piece of it – not the only way – is: have you built a large company? Have you gone through how hard it is, how uncertain it is, how traumatic it is to go through? Just this morning, I was talking to somebody about how many times we almost worried about making payroll at Sun, how we almost went bankrupt. Plenty of times, more than two.

There was a period – there was a three-month period where we almost went bankrupt because we had a hardware problem: monitors we bought from Philips were breaking every 30 days. I think there was probably a month or two when I, the earliest I went home was 3 AM, and the latest I was back in the office was 7 AM. Unless you've got rent, you felt that gut-wrenching decision. You can't advise not if you haven't been there.

Now, I hate board members who sit in a board meeting and say, "Oh, can you improve quality?" and then five minutes later say, "Can you ship faster?" and five minutes later, "Spend less money?" And they've never gone through really hard trade-offs and how uncertain that is. If you add more people and increase your burn rate, are you going to improve your chances of getting something out? These are very uncertain, very hard calls.

The biggest thing an entrepreneur deals with is which risk to take. When you take a financial risk of running lean, when you take an engineering or marketing risk of a feature-poor product, when you take a risk, do you want to take? It's like whack-a-mole. Ambiguity is so hard to deal with, but it's the essence. It's frankly one of the areas where entrepreneurs don't think about what they actually need; they pick the wrong people.

If somebody's never dealt with decision making under ambiguity in a big company, they're not qualified to help you. One of the things I hate about board members is when you're making these decisions about which risks you want in a stressful environment, the thing you most want as an entrepreneur is a board that you feel is calming you down and supporting you, not adding to the stress.

Most board members, while you're doing that, just tell you, "Oh, you're going to die," like your company is doomed. The thing I hate the most is when an old board member of mine used to send me press clippings of competitors all the time to make a point. You really just want someone who says, "You've got to take a hard risk here; it's a tough decision."

There are times when panic is the appropriate response in a company. One thing that I think I've noticed is that entrepreneurs that are working on hard, ambitious companies really struggle with figuring out who to trust for what advice. So how do you think about that?

"My big advice, and the first piece of advice I gave Joe Kraus when he was starting Excite, was the single hardest decision you'll make is whose advice to trust on what topic." The answer, you know, if you're 20 years old, do you ask your dad or your friends? Or if you're facing a marketing problem, do you ask somebody who has done marketing at IBM? They've never dealt with things where the market isn't established; it's incremental year to year, 5% improvement is what they're shooting for. They're not qualified to invent whole new markets or whole new approaches to markets.

Those are really hard decisions, and that's where nuanced advice comes in. For instance, which employee would be better? You know, the funny story that's actually not known very much is Scott McNealy when he started Sun. He started as a VP of manufacturing. I didn't know that; almost nobody knows that.

At one point, I said, "Scott, you've got to become our VP of Sales," because of certain types of behaviors. But his background was in manufacturing, and you have to make those tough calls. It's a very non-intuitive gut call. So you have to make those.

But back to this issue of whose advice to trust and on what topic: this is the single hardest decision, and I think the right investors can really help you. I had an argument just last week with another co-investor in a healthcare company. They wanted this healthcare person who had never dealt with change beyond 2% a year. I'm like, experience doesn’t matter; the rate of learning matters. First principles thinking matters. Pick for the best athlete, not the person who was the most established wide receiver who knows how to run one pattern and one pattern only.

"But there are two things I want to follow up on, so we'll get to the athlete in a second." We'll keep running into things.

"So, I agree; it's really hard to know whose advice to trust. But if a 20-year-old entrepreneur comes in here with no work experience, and you decide to back them, you have to give him or her advice and say, 'Here's how to do this.' What do you say? How do you know whose advice to trust tactically?"

I look at not what entrepreneurs are saying, and we always have this debate inside of them too, but how they're thinking about the problem in first principles thinking. If you give them a brand new problem, I'll often say, "Hey, if you were doing this other startup, how would you approach it?" If they have to think from scratch on a brand new problem – and by the way, it's a great interview question – that they’ve never dealt with, they don’t have experience with. How do they approach it? It is probably the best indicator of how fast they will learn.

If I can pick between lots of excellent candidates learning how to trust different people's judgment, including learning who to trust and which people to trust, this becomes a nuanced thing. I'm sure you've noticed one thing: we look at YC stuff in the three months that they’ve been there. What's their rate of change? If we've had multiple points of intercept, that may be a stronger indicator than any other single story. That's number one.

Right, that is my number one by far. So we always say, "How fast did they evolve their plan? Changed their plan?" What I'm saying to other investors is, "Stick to your plan or are you executing your plan?" I'm the exact opposite: "How fast are you evolving your plan or changing your plan and learning?"

Building that team that can adjust brings up a related issue we should talk about. Since I want to focus on people: when you hire a VP of Marketing – and I've said this to you before – one functional question is can he do marketing? But that's not the most important question.

If he's one of the top five people in the company, the most important question is, "What are the questions he will ask? How will he make the CFO better or the VP of Engineering better through the questions they ask?" This prompts this kind of thinking, which then leads to a better kitchen cabinet – the people you coalesce around your dining table when you have a really hard, ambiguous, uncertain decision to make.

How do you evaluate that in an interview? You're interviewing a VP of Marketing; you want to know if you're going to make the CFO better. Well, how do you probe that? I'll often say, "If you were to sue people, often say, 'Here's what happened at my company,' and I'd say, 'If you were CEO and you had to make a different set of decisions, what would you have them think under those circumstances?' Or if you're doing this other startup."

One of my favorite questions in the startup world is: "If I gave you ten million dollars today, what startup three startups would you consider, and what are the reasons you wouldn't want to invest in those?" You suddenly get how they think about a new problem, which is what you face every day in a startup.

Yeah, it really is true. I think one thing everyone underestimates when they start a startup is just how little of the problem they've already thought about and how much more is going to reveal itself every week. So I often tell entrepreneurs a business plan is completely irrelevant other than to judge how somebody's thought about a problem, not what they're going to do.

Speaking of that, how much do you expect a founder to have figured out early on? How ambitious should a founder be? Like, how ambitious were you when you started Sun? How much of Sun had you figured out, and how much did that vision stay true to what happened?

Sure. I'd say I was very ambitious. I'd done one other startup before that, which was also pretty successful: Daisy Systems. That went on to go public in the 80s and raised a hundred million dollars, which was a huge IPO in those days. So I was very ambitious, but because I was much more passionate – and passion is an important ingredient we should talk about, especially in the team – I was passionate about what I wanted to get done, not about the returns.

I liked founders who are very ambitious, mostly because if they're trying to build a $5 million company, they'll hire a team to build a zero million dollar company. If they're ambitious, they'll hire a team to build a billion dollar company.

So, among the first 15 people at Sun, we hired Eric Schmidt, who went on to run Google. I didn't know he was going to be that capable. We hired Carol Bartz, who ran Autodesk and then Yahoo. We hired Bill Joy, who wasn't part of the initial founding team; we recruited him as a founder after the fact. We hired so many other people: guys like Tom Lyon and Bob Lyon, each of whom have started billion-dollar companies.

Larry Garlic, who nobody knows now, started Remedy, which became big. Any vector that chimes himself has started so many companies. There was such an incredible talent pool there. So this I really want to dig into.

There's only a small handful, but the way you just want to take one small diversion: when I met Andy Bechtolsheim, he was in Margaret Jacks Hall at Stanford. He said, "Why don't you license the technology for ten thousand dollars?" Right? He had licensed it to six other startups at ten thousand, which, in the 80s for a graduate student, was a shitload of money.

In fact, one of those companies, Sibling, was funded by Kleiner Perkins, and John Doerr was on the board. So, they took the license. I said, "Andy, I want the goose that laid the golden egg. I don't really care about the golden egg because it will be irrelevant in a couple of years." I didn't know why, but part of it was I just loved interesting people.

So, I gave him half the equity just to join, and then I did a sales job convincing him to drop his PhD. Best decision I ever made; best decision he ever made. But it was a hard sales job to convince him. I sort of say in recruiting: "A no is maybe, and maybe is a yes." That's sort of my job, and I get very disappointed when I can't get a yes.

“How long did it take you to get him from a no to maybe to yes?” A couple of months, but Bill Joy took six months because I also had to convince him to drop his PhD. So, two people dropped their PhDs.

The very best people I've ever recruited in different ways in my career have all taken at least months to recruit. Yeah, it takes time because they always have something great to walk away with. The people who don't have something great to walk away are probably the people you don't want.

Absolutely, yes. You know, especially when you're thinking beyond dysfunctional. The people who do their jobs well, whether it's marketing or database architecture or whatever, or thinking linearly, if they're broad thinkers – which is what's key to that kitchen cabinet that helps you evolve a plan – you need people who are so full of ideas, they're always triaging down to the thing they can do.

People like that always have great opportunities, so it's hard to get them to join as an employee because they can start their own company, and you chase them forever.

So this is actually what I wanted to go to next. There are a handful of companies that have been able to get those people in the early 10, 20, or 30 employees that could go start their own company and that go on to later – PayPal’s the famous example, Sun's another one. What did you do at Sun, and what has happened in other companies where you've seen this, where people build this phenomenal early team that goes on to be wildly successful?

Well, it's always about – you know, entrepreneurship is a funny thing because vision is impractical. If you're reasonable, you won't do unreasonable things; it's just by definition. So if you have great managers, good process people, they will work against allowing the company to become great. They'll take a great idea and turn it into a good one, executing a decent idea at lower risk that's more reasonable and more sensible.

If that's okay as a goal to have, if that's your goal, great. But if your goal is something unreasonable, something ambitious, really visionary – something that can change the world – then you have to take the other approach: get this think tank of unreasonable people together, and below that layer, the reasonable people who micro-optimize within the macro ideas that the kitchen cabinet comes up with.

How do you think about the equity that it takes to attract these kinds of people?

I see this as a major problem nowadays. People aren't allocating equity widely enough. I think among the first three or four founders at Sun, we kept less than half of the common, which was just – the total was something like 25% to 27% for the founders, in equal a slightly larger chunk for everybody else we would hire. Then investors had a minority, but a significant minority, so it was like 40% for investors us after the round, something like that.

In retrospect, that was a very good idea. When last year my son started his company, I said, "Keep 15% for yourself instead of 45." He could have done either number. Try and hire one or two people at 15%, even though they're coming later, even though they didn't come up with the idea. But that would be incredible resources, especially magnets to attract other people or bring essential skills.

Can you say what you mean by a magnet to attract other people?

So if you believe a company becomes the people it hires, then your key task becomes attracting the people. There's also using them productively; that's a management skill. But attracting people becomes who finds you attractive, and selling depends on magnets.

Bill Joy was an incredible magnet even back then, right? And even though open-source didn't exist the way it does today, people wanted to work with Bill and Andy. And even if Bill didn't do a day of work, he was worth it because he helped them attract Eric Schmidt.

I don't think Eric would have come work for me as a 25-year-old other than I had some convincing power of why this was going to be large. We discarded, for example, the notion that which most investors said, "Why don't you be a graphics add-on, Terminator Codec Access?" and that was an established known market.

There was no graphics terminal, there was a company in Utah called Evidence in Southland that built graphics terminals. Those are easy. You need to distribute computing, and nobody had heard the term when we released NFS. There was no distributed file system in the world, and we open-sourced it. People first said, "Who needs distributed computing?" and the second question was, "If it's important, why are you giving away all your intellectual property?"

Thinking on the specifics doesn't matter; thinking nonlinearly about it is what matters, and that's what the team enabled.

So, full circle back to equity: as much as leaving 30% of the pool to non-founders. Neil took 15%, he recruited in his other startup, Keira; he kept 15%, hired 15% as a co-founder for 15%, and then left the rest to hire great employees.

Now, it is dependent on the area you're working in. He was working in AI. He wanted people who were all making million-dollar salaries at Google, Uber, and other places as engineers. So, you had to give them 3%, 4%, 5% each, and you'd normally not think about it. But if you're competing in an AI startup, you're not going to get the best talent without it.

Especially if this issue we talked about earlier, if somebody can do their own startup, they will. So they're not comparing them to your job; they're comparing them to them starting their own company. And if you don't do that, you're including only the people who couldn't start their own company, who won't help you evolve.

We apply a thing that I hear a lot: "You know, I could hire an engineer for a few expert basis points! So why would I ever pay 5x or 10x?" I always say, "Wait, and in two years ask me again the right decision."

But the quality of people you can get if you're super generous with equity, I think this is the shape of things to come. This is my single biggest beef with YC: not enough option pools.

Right? Not enough option pools, not enough focus on recruiting the right co-founders. I was fortunate; Neil trusted me so I could shepherd him into sort of a very different approach. I think he may have had the highest option pool in his batch.

Yeah, look, I think it's great. I have found that it's harder to convince founders of this than I would have expected. Yeah, because it's not intuitive.

Well, also, most investors say make the option pool as small as you can. Yeah, because neither investors nor founders want to— they'd like to own as much as possible. It sounds really good, and until you feel it, it's hard to convince them.

But I'm—or they don't like to own the higher percentage, but the pie is much smaller. If you say, "I'm maximizing the size of the pie," then it doesn't matter what percentage you're on. Look, I think this is the most important piece of advice that we've talked about among many important things today.

I think being super generous with early employee equity and getting founder-quality people in the first 10 employees – I think all the evidence is on the side of doing this, and yet almost no one does. So there's a huge edge if you're willing to do it, and I absolutely, a hundred percent agree that this ends up being the single most important thing in the first six months of a company. It’s incredibly important.

In fact, I've written two pieces on one. Once you're doing a startup, how do you engineer the gene pool of a company? And there's an important part—you sort of have to have a process. You know, I find it silly to advise people to hire the best team because everybody says that. It's not actionable, but when you process to follow this process I call gene pool engineering, you are trying to maximize your chances of success and minimizing the risks you're taking on by virtue of the team you're building.

That's one part of it, and that's sort of more mechanical. The other part is instead of functional hiring: hire a VP of Engineering, hire a VP of Marketing, hire a CFO, hire a VP of Customer Support. You do hiring for this nonlinear stuff: how does your VP of Marketing improve the quality of your VP of Engineering? Make them think harder.

This kind of nonlinear thinking is something I've written a different piece called, "I think the labor of love, the art of hiring," something like that on our website. That's a long piece about this, and both these were a couple of years ago in TechCrunch.

Also, I think the essentials reading fonts know in my view, even if they don't follow the advice, it'll change how they think about the problem. We will add the links to our website.

One final question: one area to wrap up on, we've talked a lot about how to pick your employees, your team members. I think it's almost as important to pick your investors as well. So how do you advise founders if they're trying to build a significant company that's going to do something very disruptive and be around for decades?

I think every piece of advice you get is the smaller part of what you get. The money you get is the smaller part of what you get. Advice and the right approach is the much more important part.

There are investors who are happy with 3x their money; in fact, they sell as soon as they can and make 3x or 4x. There are people who care about your vision or people who understand the technological approach you're taking and will be much more tolerant as things go wrong. Right?

So, those are the factors I suggest people optimize for. How do you tell? How do you know if an investor truly cares about your vision? Talk to other founders, especially founders who've gone through a large promise, a large vision, and had hiccups along the way.

When things go wrong around an ambitious path is when you can actually judge an investor: how they think about hiring. Look back in retrospect a couple of years later or talk to founders who can and say, "What worked and what didn't?" Like, those aren't the key questions.

I think an investor is an employee who you can't fire, and that's how you should think about it. Otherwise, all the same principles apply. I get very frustrated with investors because they mostly detract from value.

Like, most investors are negative value add to a company that's trying to be ambitious if they're just trying to get to liquidity as soon as possible. Then there's plenty of investors who do that well.

Have you had founders who come pitch in and say, "That's too ambitious?"

I've never had that happen. I have had the following conversation: "That's ambitious; that's awesome. Build a team for it. Now what is step one, two, and three?" Let’s be thoughtful about how you discover the risks on the path to the vision.

Because frankly, achieving a large vision is first about discovering what's hard, what the problems are, what else influences success along that path. Once you've found the problem – which you can only discover by doing things – then you can...

The only recipe I've ever seen work for making really impactful companies is both a giant vision and a good step one, two, and three. Yes, you have to have both, and neither without the other.

World War I: the related thing I often tweet is I hate pontificators. It's so easy to do studies; it's so easy to opine on things. I love doers versus pontificators. Each and every Nassim Taleb has written "skin in the game."

But also, not so much Black Swan, but his second book after that, so skin in the game is about doers versus pontificators. Then he has another book about the right kinds of asymmetric risks to take that he wrote after the Black Swan that I feel is more important than one.

One thing that has just worked for me again and again in my career is I happen to love doers, and I don't love talkers. I just got lucky because I biased the people I surround myself with so far in that direction. It's been great.

Final question: What are you most excited about doing now? What do you hope to get down in the next 10 years?

So, I'm 63. I actually, at 60, defined what I want to work on for 20 years. I wrote a piece called "Reinventing Societal Infrastructure with Technology." It's about a 50-page document, and it was the following exercise. I won't go into the details because we don't have enough time. I said, "If I took the U.S. GDP, what parts of the non-governmental GDP concern me personally with technology?"

I expected I'd come up with a small part of the GDP that was open to working on. Turns out there's no part of non-governmental huge U.S. GDP that one can innovate on in ways that are not 5%, 10% improvements, but rather a hundred to a thousand percent improvements in resources.

So, I sort of decided if 7 billion people on the planet want the lifestyle that 700 million people, the top 10%, have – and that 10% has an energy-rich lifestyle, a healthcare-rich lifestyle, an education-rich lifestyle, a transportation-rich life – you get the idea.

The very rich lifestyles without destroying the planet: could you get seven billion people to the lifestyle of the 700 million people without destroying the planet and without needing 10x of everything? Could you do it?

I could think of ways to do it in every major part personally. The only thing I'm sure of is entrepreneurs. So I subtitled this document "A Call to Entrepreneurs."

It's amazing because no part of the GDP is immune to innovation. Five or six years ago, when Pat Brown said he wanted to eliminate animal husbandry, it became very clear we could change how most of this planet's usable land area is used. Most of the land on this planet.

That's Pat's mission, and I think I said yes to him in our very first meeting. We didn’t last. I didn’t even need to know the details; it was just a vision too big to not attempt. So hamburgers to rocket labs doing rockets, which we did about the same time, to my new passion: could you build buildings with 80% less material?

That's why I'm so excited about printed 3D buildings! There's nothing from food to building to construction to rockets to everything computation enables, from AI to... it's really exciting. And I feel like 20 years is not going to be enough.

I hope I stay healthy.

"I hope you do too."

I'm sure you will. That's a great place to leave it. People can read that document and call you up.

"Yeah, thank you very much."

Thanks!

"Thanks, Sam. This was fun."