Company Building with Vinod Khosla of Khosla Ventures

May 14, 2014.

Company Building with Vinod Khosla of Khosla Ventures.

People seldom realize there's a massive differenc between a 0 million dollar company with 3 people and a 0 billion dollar company with 3 people.

There is a differenceIt's in the culture. It's in the mission. It's in the people.

It's in the way they talk about this and so I'm going to talk a little bit about how to build a billion dollar company even with 0 billion in revenue.

Startups are about passion for a vision. They feel like Cat in the Hat. All of you recognize the slide. By the way this was the slide, the exact slide I used my first startup at Daisy system in my business plan.

In my slide presentation I won't tell you what technology we used back then but it feels like that.

But there is a difference and I can usually recognize this almost right away.

many of you have seen me argue with you about these enrollment sounding points. It's a difference. I see Mark sitting there arm wrestling over this issue every time we meet.

So what's the startup process about?

Strategy to leverage your assets and minimize your liabilities.

By the way most people in the business Plan forget the goal is to minimize liabilities in addition to leveraging your assets.

You start with a short term advantage to break in the back door.

It's usually a technology advantage. You have to turn it into an asset.

You want to experiment a lot early.

I call it organized chaos phase of a company and you want to turn it at some point as you grow as you have more to lose, into an execution as a process.

But the main thing you're trying to do when you start up Is optimized risk management.

Your building teams and I'll talk about that and you sell.

So sell sell sell was going to be my emotional storytelling pitch

I'm not going to go into it.

I'm only going to cover two optimizing risk management in people in gene pool engineering.

So those are the two topic I'll try and talk about and someday I'll do the sell sell sell talk when Carmen is not here so I have nothing to compare against how poorly I do it.

Risk management is fundamentally what you're doing. There are many kinds of risks If you're building a billion dollar company.

You have to think big, but you can't act big. Because your burn rate will get to high.

So think big act small In the real cash of a company, whether you're trying to build a 0 million dollar company or a 0 billion dollar company.

The trade-off you make and how much conviction you have and courage you have to pursue the billion-dollar vision

while doing the million-dollar tactics.

I think last year I said be obstinate about your vision be flexible about your tactics.

It's an art form, the big hairy audacious goal. (BHAG)

But risk management is getting to a base camp that is much smaller than your ultimate vision.

As I explained many times if climbing Mount Everest.

Getting to base camp is important. But way too many people in this industry, take a lot of investors, get their companies to a base camp that isn't with in sight of Mount Everest.

This is where the tactic you select Is very important your base camp has to help scope the path to Mount Everest.

Validate each element.

Just because you believe what you are doing. Doesn't mean you are right.

In fact most of you are wrong. A lot of words. Validate my own rule.

I seen many of company build the wrong thing on budget and on schedule.

Test what you are trying to build is right past every assumption and then worry about execution.

So it's okay to be inefficient. It's ok to do it with flexie planning.

Theres no point in planning if you don't know what you're doing. I think a speaker talked about this.

And set up a culture of experimentation early.

I said I would talk mostly about risk there's many many risks in a startup phase.

You have financial risk, people risk, technology risk, market risk, investor risk, groupthink risk, risk trade offs abound.

I'm only going to talk about two of them. If I had enough time I would talk about all of them and you're always making trade offs.

If you don't raise more money you're taking a financial risk.

If you put more features in the product you're taking a financial risk but reducing your market risk.

If you're putting fewer features you're taking less of an investor risk.

So let's talk about financial risk.

The big mistake I see, it's because people believe their assumptions they don't validate them and this is weird minimum.

Viable product is a really good book.

I prefer to plan for the unplanned then to build a plan.

Most things people know, If you look at your past two years of history. You didn't expect what you know today.

But don't forget what you know today.

Their will be equal amount of new things you'll discover by this time next year.

Plan for the unplanned.

That's why flexi planning is so important in managing major risk.

When you plan, the only plan that really matters is the worst case plan.

The contingency plan.

The one thing you can't afford to do is run out of money.

One of the things I say at board meetings and it really irks the board members, I seldom go to board meetings, But when I do

I say to waste money.

But wasting money and when you get this slide deck

um theyl be um. Their will be a deck on it.

They walk down the risk curv. Not up expense or time curve.

Don't build a time versus burn rate plan. Build the risk vs burn rate plan.

What really matters are the risk reduction points.

That's when your burn rate to go up.

That's how you do good risk management financially. By basing it on risk triggers not time.

What are risks worth taking?

If you can afford money and there's a reference on that. I recommend you read that reference, It was a blog post by Brad proud.

If you do a 1x down turn you loose the money spent on the experiment. But 10 X upside.

The key to failure is to have them small and not catastrophic.

People risk. People hired functionally is the wrong thing to do.

If you think about the billion dollar company you want to build, the people you hire will dictate what your company will become.

Not the plan you have.

So let me move on to the people race because that's the second most important thing.

If you survive long enough which is what financial risk determines then the main thing is people and people risk.

A couple of topics I'll try to cover very quickly.

Let me start with your job as founder and CEO.

I talked about the difference between a 0 million dollar company and a 0 billion dollar company and I talked about thinking big and being consistent with the vision.

An obstinate about your vision, but flexible about your tactics and that's what takes courage.

So many practical shorts that show up

that you can lose your vision if you don't have courage.

I love your convictions.

A couple of tactical things I like to suggest, be proactive about deciding your priorities.

I would venture to guess the majority of you are reacting to things that come in and not being proactive.

If 50% of your time is not free Sunday night because you're thinking about your week you're almost certainly not going to be proactive.

I suspect if you have 50% of your time free 50% of that will go into emergency that pop up today. So 75% of your time will be reactive and if you can't save 25% Of the time as measured retroactivly, you are not leading your reacting.

Can every phone call, every email message, every crisis be proactive?

It's the way you follow a vision. The specific process I want to talk to you about.

What is the way of quantifying this issue of what kind of people you want to hire?

I've never heard somebody say to me that they want to hire B quality people.

Everybody says we want the best people, we want a great team.

I say what are you doing about it? The answer is almost certainly very big and not definable.

So let me talk about a couple of things.

Engineering diversity is important.

Going to talk about collision correlations are about a merger of ideas and random interaction.

Diversity is more important to a company success in gene pool and the hardest thing to engineer. Because everybody likes people like themselvesand and so they hire people like themselves. People they know.

Be diverse about age. Young people think differently than old.

Stumpy from experiences come from fresh ideas.

A soup of diversity is really good.

By age, by the company they came from, by the industry that they came from, from the country they came from, whatever type of diversity you can engineer.

I find it's a major indicator of success long term.

There's a difference between a manager and a visionary.

There's a reason most reasonable people don't start unreasonable companies.

It's not reasonable to say you're going to build a billion dollar company

Only unreasonable people do it, only people with vision.

They tend not to be pragmatic. If they were they wouldn't do unreasonable things.

By definition visionaries are not pragmatic and again you need both.

But frankly if I could only have one,

I'd rather have instinct and vision over processing management.

I almost always prefer the founder as CEO and the President as the manager.

But sometimes you can do it the other way around.

Larry Page and Sergey Brin talked a little bit about this.

It worked out great when they actually got a good manager for a short time.

It really helps them I believe.

People focus on your strength.

I haven't really seen many companies fail because of their strength.

But they always have weaknesses or liabilities in their business plan that they fail to focus on.

The same is true of teams.

Many people fail not because of what they know or what they know they don't know.

Mostly failures come from what they don't know they don't know.

And diversity helps tremendously with that so gene pool engineering.

How do you engineer diversity?

Going to go through a very specific examples because I hate people saying hire good people.

Well nobody tries to hire bad people.

How do you build a great team.

I tried to quantify this process.

How do you engineer a diverse gene pool for a company

But also taylor to a company's needs?

Let's start by identifying your risks.

Remember the investors are worried about your risk.

Well you should be worried about your risks to. It makes sense

If you don't know what your risks are you can't engineer a team to address your risks.

So let me go through each of these steps a little bit and illustrate it in a quantifiable way.

Most people have functional needs or chart. They start their

slot they need to fill. I hate these charts because its function of thinking

I need a VP of Engineering slot.

How about you identify your key risk. You don't need to read the risks.

I should have made the font smaller so you don't try to read them.

By the way I use that technique in presentations a lot.

I make the font small enough so that people won't read it

because it goes from something worth reading to a picture

that explains the vision but identifies the risk.

In your company,

What will cause success or failure?

Or sometimes upside?

With each risk identify the target companies where you could hire from to solve those risks.

Find the names of the people in each of those companies and a sign a priority to those names.

So for your risk number one you now have 15 names.

That have identified people who come from the places that have sold that risk before and you know the people that help them solve that risk.

That's a great way to address that risk to make sure you have multiple companies. Not just the last company you worked at.

Your engineering diversity into your gene pool.

Try and look outside the segment.

One of the classic examples I gave is Jennifer sitting here.

Lines attached, looking at steel mill, the shortage of engineers in New Zealand.

Where the company is based. They hired a guy that had worked on disk drives.

You would think there's no connection to be had.

This guy new surface morphology.

And a year later the founder Sean came to me and said

he figured out all new ways in which bacteria can grow on surfaces.

Because he looked at surfaces when doing disk drive surfaces.

What else are you trying to do?

Show you have a plan.

But if you build a team that's gone through all the possible ways you can make mistakes when developing your product you're way ahead.

If you figure it out the biases people had and you have a diversity of biases you eliminated another source of errors.

If you don't do this you're going to make many of these mistakes all over again for yourself with your dollars.

When you could have bought the people that could have helped you with your mistakes.

I sometimes feel with many of us as founders, know and feel this way.

If you would like in a startup you do things in every wrong way possible until you find the one way that works and it almost seems to be the last thing you discover.

I'm sorry to say my biggest advantage and start up I screwed up.

I screwed up more times than anybody.

But I keep finding new ways to screw up. But collecting the mistake.

The mistakes you can make

by hiring the people who already made them

Is a really important part of engineering your gene pool.

So you get a gene pool engineered QP risks.

That's going to increase the probability of success very very significantly and save you a lot of time and money.

So how do you hire in some of these areas that you don't know how to hire in?

You are a great engineer at Google and you go and start a company.

You've never hired a CFO.

You've never heard of VP of Marketing.

You probably not worked with one.

You probably have a very naive view of what a VP of marketing does.

This is a paper that is on our website about how you might approach a problem In hiring in the domain you know nothing about.

Many of you have run into this.

I'd suggest the majority of you have hired the wrong person

and a year later said I didn't know what I was looking for.

Please read this paper.

I was talking to somebody and they said their last hiring mistake probably cost him 20 million dollars because they hired the wrong person.

The gene pool engineering paper in more details is also on the website.

These are things on our website that are resources for you

that I've spent a lot of time on.

Let me talk about this 4th thing.

I think I got in a lot of trouble because the last tech crunch last September.

I said a lot of VPs don't help a company and 70% of them hurt a company.

I truly believe most of your board members will hurt you.

Its sad, I may be wrong.

But it will take a lot convince me I'm wrong.

It's nice to have diversity of opinion on your bored. So don't get me wrong.

Diversity always helps.

But this paper which was published when we hired Keith

Talks about what you should look for and how you should engineer the board.

Because you're the ones with high influence.

You'll have an emotional reaction to one little comment from them will change something you do or change your trajectory.

Also starting with the wrong people is catastrophic.

Some people say I just need to hire three engineers, let me just hire.

That's not what you're doing.

You are hiring the three people to do task.

But you're doing something much more important.

You're hiring the people that will hire the people that will build a billion dollar company.

If you just hired the people to do the task you will build a zero million dollar company. Not the 0 billion dollar company.

Because they're going to be the people that hire all the other people.

Recently we had a situation where a CEO Needed some critical talent. He needed to upgrade.

Andy Costello talked about how your task Is to keep improving your team and that was a really compelling pitch.

So I called the CO and they had a great candidate and I referred him to them.

And I said please meet him and he said great and he sent me the list of people.

And this candidate was going to come in for an interview and I called him and I said, If I were a good candidate and I interviewed with these three people I would not come back for a second interview.

And I said if they interview him. Just interview him

and he came back In my book I would think he's a bad candidate because he doesn't have good judgment.

This is the problem once you start in the wrong place.

This is the difference between a 0 million dollar company and a 0 billion dollar company.

I want to keep connecting.

There is a massive difference between these two companies.

So in that particular case I said okay I'll do the first interview.

I want to sell him before I start evaluating him.

Don't get caught up in titles.

Let me go back to the old chart.

It's not about functional hiring, It's about collecting talent.

When I started sinai I hired is great database guy from xerox parc and people were arguing with me.We're on a limited budget.

We'd only raised two or three million dollars why are you hiring a database guy If you're not doing databases?

I said no I'm hiring a really smart guy and although we never did databases and no plans to do it he affected the file system we did.

And sun ended up with NFS which is still the standard today

because we happened to hire a database guy.

But really we didn't hire a database guy we collected talent.

So don't get caught up in titles.

Create fancy titles if you want to hire somebody.

Create a chief science officer.

Create a title any title if you have talent.

There should be no budget.

There should be no wall hiring.

Just make something up. Make them general manager of something.

You never don't need that great person.

You never don't have a budget for that great person.

You always have a budget.

You always need that person.

At that level the people that will help you sort out the direction of your company. Your brain trust. Absolutely key.

So one of the points at making this paper about hiring.

"The science and art of hiring"

A VP of engineering Shouldn't be evaluated on what he does in engineering.

If he doesn't make your VP of marketing much much better,

or if your VP of Marketing doesn't make your sales guy and your vfo and your VP of engineering much much better, they are the wrong top level brain trust for the company.

Because you need a diverse team to help its rate division

as you move towards Mount Everest.

It's not about functional hiring so let me stop there.

I covered really only two or three when I was doing this talk.

Just like I told you, you should decide what you want to talk about.

I started in my presentation slide by reasons to invest so I did the same exercise.

These lights may be on here or not.

So I said what does company building involve?

Came up with a list.

Just like the first slide I gave on presentation and the list went on and I ended up with 48 bullets I want to talk about

and I said that might take 6 hours and so I decided to focus on three. Following the rule of 3.

But someday I hope will have time to talk about all 48 of them.