See all Y Combinator transcripts on Youtube

Speaker 1

00:02

Well, do you think we'd be here? This is a...

Speaker 2

00:05

No. So, 5 years ago, Matthew and I were sitting in your seats. We started Cloudflare 5 years ago. And I thought I'd just tell you a little bit of what's happened those 5 years, and then we can get into some of the things we've learned along the way, so you guys can build really big, successful companies too.

Speaker 2

00:24

So today, so what is CloudFlare doing? We're building a better internet. We have 2000000 customers who have signed up for our service all around the world.

Speaker 1

00:31

Including Y Combinator.

Speaker 2

00:32

Including Y Combinator. And why people sign up is that we make their web properties load fast, we protect them from online cyber attacks, including DDoS attacks. We just make sure we load balance, make it available.

Speaker 2

00:46

We just make the internet run better.

Speaker 1

00:48

We're Cisco as a service.

Speaker 2

00:49

Cisco as a service.

Speaker 1

00:52

It's only, you know, $250 billion market cap.

Speaker 2

00:55

There you go. And so over these 5 years, today we have 2000000 customers, We have about 120 employees in San Francisco and London, 2 offices, and every day about 3,000 new people sign up for our service. It's been definitely a rocket ship.

Speaker 2

01:10

If you look at our growth numbers, it is the perfect opt-in to the right on all metrics, Just grow, grow, grow, grow, growth, growth, growth, growth, growth.

Speaker 1

01:18

And yet last night I was at your house having dinner, freaking out about what in the world were we going to talk to these people about. And Andrew backstage is saying, is it basically just going to be the, here's why you should sign up for Cloudflare show, which we definitely don't want to do. So what, I don't know, what do you want to talk about?

Speaker 1

01:41

I have no idea.

Speaker 2

01:43

Truth be told, Matthew's very nervous. He's like, how are we going to make this without it seeming lame? And I was like, no, no, I feel like we actually, I think seeing co-founders together in Dynamics, you, many of you probably have co-founders.

Speaker 2

01:54

And I feel like there's a lot of stories online where things don't necessarily work out, where they end up no longer talking or getting pushed out. I think seeing a model where 5 years in, we still talk to 1 another.

Speaker 1

02:06

We're at a distinct disadvantage because neither of us is very funny and neither of us has been fired, so Andrew definitely has been. That's a nice 1. He was the amount.

Speaker 1

02:20

Okay. He was the other 1.

Speaker 2

02:24

Okay. So, When we started, lots of people come to us, we talk on Frizz all the time, they're like, how did you do it? Like, how did you get to where you are today? And what I always say is there's no silver bullet.

Speaker 2

02:41

There's really no silver bullet. And you've heard lots of successful and incredible lineup of speakers today tell their version of what you need to do to be successful. And, you know, Matthew has this analogy about sharks and mosquitoes, and I think that's a really good place to start.

Speaker 1

02:56

So, you know, I think that, so there are 3 of us that started Cloudflare. And Michelle is the person that makes the trains run on time. She's essentially our COO, although we'll talk about why actually that's not, we don't have titles really.

Speaker 1

03:13

Lee Holloway, we don't let him leave the office. He's currently chained to his desk writing code right now. And so he's the 1 actually building the stuff. And my role is basically to assemble the IKEA furniture and tell the story of what we're doing.

Speaker 1

03:30

Which is actually strangely a really important role. But you need those 3 things in order to do it. And so, you know, I think a lot of times, and early on, you know, we freaked out about how are we going to raise money and how are we going to, you know, how are we going to get media and how are we going to get press? And what, you know, I really have always valued in Michelle's role has been that what Michelle worries about more than anything is, you know, how are we going to hire people?

Speaker 1

03:58

How are we going to make sure people get along when they're in the office? How are we going to make sure we're planning, making sure things get done? And again, as Michelle references, it's a little bit like worrying about sharks versus mosquitoes. I tend to worry about the sharks that are out there.

Speaker 1

04:12

But it turns out that if you look at it around the world, sharks only kill about 6 people a year or 10 people a year. Mosquitoes kill hundreds of thousands, millions of people a year. 750,000. There you go.

Speaker 1

04:27

See? See? And so sweating those little details ends up being a significantly bigger deal in terms of being successful. We, early on, I remember when we first interviewed our first PR firm that we hired, And we were like, we need to be on stage at every event talking about how we're going to change the Internet and we're building a better Internet.

Speaker 1

04:51

And everyone just said, whoa, whoa, whoa, whoa, slow, slow, slow, slow, slow down. And again, Michelle, to her credit, was always focused on how do we hire great people, how do we listen to our customers, how do we solve those little problems that aren't particularly sexy that don't get written up in the news, but that are absolutely critical. And so, you know, I think that I think that that's part of why we've made a made a pretty good team.

Speaker 2

05:14

Definitely. And so what does that mean early on? So again, 5 years ago we started. We literally started 5 years ago.

Speaker 2

05:21

And for the first year, we just built the product behind the scenes and then opened up a private beta. And so what does that mean? That means the first version of our product that we released, I was literally embarrassed about. I remember cringing when I said, oh God, asking someone to sign up for it, because it wasn't ready yet.

Speaker 2

05:39

And I knew we had all this long list of features, but eventually you have to ship product. And so we invited, you know, 10 people signed up, 10 friends, all 10 signed up, and we took all 10 websites offline. So that's the exact opposite of our value proposition of making things better.

Speaker 1

05:58

And we fixed those problems and had them sign up again, and then we took all 10 of them offline again.

Speaker 2

06:04

But then we fixed those problems, and the next 10, maybe 8 went offline, and we kept getting better and better and better and better.

Speaker 1

06:11

It's amazing that they tolerated us. It was, and actually, It's something like 90% of those initial customers are still users. The 10% that dropped out just went out of business.

Speaker 1

06:21

So it's pretty remarkable.

Speaker 2

06:25

And that's true. And the way we were able to get these initial customers, because often people say, how do you get your first customers? And it's really hard.

Speaker 2

06:31

When we got to 100 customers, we took the whole company to Vegas. There were 6 of us, so it wasn't that many people, but it was like a big celebration. And so for months, we tried to get to 100 customers. It was really, really hard.

Speaker 1

06:42

My favorite story. So when we, there was no, like Amazon Web Services hadn't really taken off. We, Lee, the guy chained to the desk and I had started an open source project a long time ago and we had about 80,000 users and we needed servers to develop CloudFlare on.

Speaker 1

06:58

And I was like how are we going to get servers? We don't have any money and what are we going to do? And Michelle said, you know, you always talk about how loyal this project Honeycutt community is. What if we just asked them if they can donate servers to us?

Speaker 1

07:12

And that seemed absurd at the time, but we emailed anyone who was within about a 50-mile radius of Palo Alto, which is where our first office was, and said, hey, if you've got an extra server, and it ended up emailing about 200 people. We got a 70% response rate on people who said either I've got some server or I know someone who does, And then Michelle and her Jetta, which is the same car we actually drove down here in, drove around from person to person to person picking up all of these servers, none of which worked, putting them in the back of her car. And then we assembled this to, we got essentially to what turned into these Franken servers that was what we originally built Cloudflare on. But the most important part was Michelle actually showed up, talked to these people, said, hey, here's what we're thinking of building, what do you think?

Speaker 1

08:00

And they were the first people that were giving us feedback on, hey, I'd use it if you did this, but I wouldn't use it if you did that. And that was really what... That driving around in your Jetta, I think, was what kind of got us our first people using the service.

Speaker 2

08:14

Yep, definitely, although I also couldn't code, so I had nothing else to do. So that was a good use of time.

Speaker 1

08:20

That's important, again, it was, like, early on, we had a conversation about, you know, Michelle's background was in biochemistry and chemistry and had worked at, you know, for Toshiba and Google and we had met in business school and, you know, from my perspective, like, I knew instantly that she was the sort of person that really filled in the blanks of the things that I wasn't good at. I had known Lee for 10 years. And what Michelle did was she made sure we got things done.

Speaker 1

08:57

And, you know, what we always talk about is early on people ask us a couple of things. People would ask us if we were dating, which was strange because we weren't.

Speaker 2

09:08

And we're not.

Speaker 1

09:08

And we're not, still. Don't say that so... Kidding.

Speaker 2

09:14

I'm going

Speaker 1

09:17

to go Hang out with Andrew. So they would ask us that. And then the other would say, how do you split up issues?

Speaker 1

09:28

And if you're sitting in the audience and you're a co-founder team and you're fighting about who does what, I hate to tell you this, but you probably have the wrong co-founder, because, like, we've, I don't think any of the 3 of us have ever, it's been so clear that Lee builds the stuff, Michelle makes sure it gets done, and I assemble IKEA furniture, that that's, you know.

Speaker 2

09:51

Think of it as a Venn diagram, back to my science roots. So we, and I've used this so many times, where Matthew Lee and I, we're very different. We cover so much surface area, but we have a little bit of overlap, so we share the same vision, and we trust 1 another.

Speaker 2

10:05

And that makes an amazing founding team. And so, if you've already picked your co-founders, I mean, better to have the conversation now than in 3 years, seriously, because again, 5 years in, I now understand why co-founders get into fights and get pushed out of companies. Because your role changes and all these different sorts of dynamics happen. So today you should think, okay, my co-founders, do we cover a lot of surface area?

Speaker 2

10:28

Do we have different skill sets? And is it somebody that I trust fundamentally? And those are really, really, really important questions to ask yourself, because if so, then you have a really strong foundation to go forward. And so I think for us, we covered a lot of surface area.

Speaker 2

10:41

Whereas at first, it wasn't obvious like that we were good co-founders. I mean, we didn't know each other that well.

Speaker 1

10:48

We weren't that, I mean, we knew each other, but we weren't like best friends. We were

Speaker 2

10:51

not best friends. Like, so sure, we went to business school together, but we were not, we just didn't know each other. And I did not know Lee at all.

Speaker 2

10:57

I knew him through Matthew.

Speaker 1

10:58

At first, actually, Lee, when we were starting this, Lee said, I understand why we're going to work well, but why do we need Michelle? About 3 months in, he said, I now understand why we need Michelle, but we're not sure why we need you.

Speaker 2

11:14

Thankfully, we kept all of us. But these are the sorts of things where, you know, again, those decisions early on helped us attract other people to fill in the gaps where we didn't already cover a lot of area and make progress early on. And again, I think, you know, If you take anything away from today, it's as a startup your greatest asset is momentum You have to make progress you have to make progress faster than you're the incumbents in the marketplace and that can mean anything from building a product or or or building a community of people who are following your blog, or it can mean just assembling a team that can build things scrappily that people sign up for and use.

Speaker 2

11:57

And that progress, you'll start to gain a lot of momentum around you, and that's how really big companies get built. But it takes a long time.

Speaker 1

12:03

And it's interesting, like the decisions that you make early on end up affecting much later. I remember our first board meeting, we raised money, we went into our first board meeting, and we were like, okay, so here's the team, and like Michelle was going to be vice president of user experience, and Lee was vice president of engineering, and we were hiring this guy who was a guy named Girish Patenge, who was at Facebook, and brilliant on their operations team at Facebook, and we were recruiting him out to come work for us, and we were like, we want to hire this guy, he's going to be great, he's going to be vice president of technical operations. And 1 of our board members said, how many people has this guy hired?

Speaker 1

12:45

How many people has this guy fired? Like Is this person really someone who deserves that title? And there are really 2 schools of thought on titles. There's sort of the Mark Andreessen school of thought, which is that when you hire someone, there are only So many different variables that you can play with.

Speaker 1

13:02

You can increase their salary, you can give them more equity, you can increase their span of control, what it is, who it is that they supervise, what it is they supervise. You can give them a bigger title. And of all of those things, the cheapest is title. So make everyone, you know, executive, senior, vice president of, like, the earth.

Speaker 1

13:24

And that's 1 school. The other school of thought is the, you know, Mark Zuckerberg school of thought, which is when you've joined Facebook, everyone has to take a step down. So if you're a vice president, you become a director. If you're a director, you become an associate.

Speaker 1

13:36

If you're an associate, you become a like peon or whatever. I don't know. And so and we've, you know, if I think if we had to do it over again, we would just tell call everyone engineer, like just I don't I don't care if you're doing customer support, you're an engineer. I don't care if you're doing finance, you're an engineer.

Speaker 1

13:50

I don't care if you're doing customer support, you're an engineer, I don't care if you're doing finance, you're an engineer, I don't care if you're in sales, you're an engineer. What we did instead, though, at this first board meeting, we were like, everyone's a vice president. And the board feedback was, honestly, none of you deserve to be vice presidents. And I remember we were driving back in Michelle's Jetta again from Palo Alto, and she said, you know, honestly, I don't deserve that title.

Speaker 1

14:13

Lee doesn't deserve that title. We haven't fired or hired anyone. So maybe we should just get rid of titles. And you have to appreciate that for a second, how hard that is as a founder of a company to say that I'm going to I'm actually going to say I don't want that.

Speaker 1

14:30

But that then set a precedent where everyone we've hired since, when they say, oh, I want to be, you know, executive senior vice president of whatever, I go, you know, Michelle doesn't have a title, why do you get 1? And that has made sure that we're selecting people who really want to be there, but fundamentally it also meant that Michelle really had to trust me and trust the rest of the organization so that when it was the right time and we have now thankfully not fired that many people but hired a ton and still talk to every single candidate that we hire, that now we've earned that title. And so, that early decision that really required a humility was, and it was easy for me, because they were like, someone needs to be the CEO, so you gotta be the CEO. But was really hard for Michelle and Lee But it was the right thing to do and that's I mean 1 of the things I admire the most about Michelle is Michelle has Michelle has ego, but no vanity and that's like if you can find people that are like that, you want them to be around you.

Speaker 1

15:39

Because like, think of how we're building an infrastructure technology Startup. Think about how many reporters call us every single day saying, woman, tech, infrastructure, technology startup. I want to write it. Let's put that on the cover.

Speaker 1

15:55

And Michelle has always said, it's the company first. It's not about me. It's not about any of that. Let's make sure we're telling the right story about the company.

Speaker 1

16:04

And again, if you can find people like that to join your founding team, those are absolutely the people you want.

Speaker 2

16:10

They're out there. You can find them.

Speaker 1

16:18

You know, the other thing that I think has been really key to our success is that we have, you know, while I think we have sort of eschewed the women in tech story, I think that having a woman on our board and having that for the sake of diversity is really, I mean, we're very different people. My last startup, there were 3 of us that started it. We went to junior high together.

Speaker 1

16:43

We were essentially 3 white guys that all had some combination of technology and law, and we fought like just cats and dogs. I mean, it was, it's a miracle we even talk to this day. Whereas you grew up in Canada, Lee grew up right here in Cupertino, go Canada. High, kind of Low vanity, high ego, great, great, I like Canadians.

Speaker 1

17:09

We hire lots of Canadians. But I think having that set of diversity, and I'm still really proud of the fact that we walk around the office and the number of different languages that are being spoken are just incredibly diverse and that has so much more of kind of an effect on how rich the product is and how we sort of see the rest of the world in a much more unique way.

Speaker 2

17:35

We do have a really diverse team. And so diversity means lots of things. Gender is 1, but where they used to work, where they grew up, all those sorts of things really matter.

Speaker 2

17:45

And so we're a web infrastructure company. Again, that is not something you just kind of learn overnight. But the number of people on our team that come from a web infrastructure background, it's very, very small. In fact, for the first 25 hires, no 1 did.

Speaker 2

18:02

And in a way, so some people early on really gave us a lot of, oh my god, you need to go hire all these Sun Microsystem people and Juniper people and all these people who really understand how they get, or Akamai people. And they're really, they're really, there's investors who really pressured us to do that. And we kind of said, we're happy. Let us keep doing our thing.

Speaker 2

18:23

And what I hear now a lot of people is if you are trying to do something where you know a lot about the industry, sometimes you don't check your assumptions enough, and it's really hard to really find that idea that really works. And so we had a lot of adjacent experience related to the industry that we're doing. So we had people who knew web security, and we really understood developers, and all these different sorts of things, and that together has made something magical. But it was a very adjacent industry.

Speaker 2

18:52

And so if you're in our seats where you're like, oh, I'm trying to start something, but I'm actually not an expert in it, I would say you are at an advantage.

Speaker 1

19:00

As long as you're interested in it. You have to actually, you have to deeply be like, I'm fascinated by this problem. I don't feel like I necessarily understand all the different corners of it.

Speaker 1

19:10

But sometimes by not understanding all of that, like everyone who competed with us always charged on bandwidth. We had no idea we were supposed to do that, so we just don't. And as a result, you know, we've that's been a real key to us growing as quickly as we have.

Speaker 2

19:24

We were a bit naive of how hard it would be. We've had to we've come up against some very hard technical problems. But you know what?

Speaker 2

19:31

If you're looking to hire engineers, good engineers want to work on hard technical problems. And we have found really great solutions to those technical problems. And that's been corridor success. What we always say is, okay, if we can solve this, that just increases the hurdle for the people that are behind us.

Speaker 1

19:45

It's, and it's like 1 of the things I'm struck by when, because we're in a fortune position now where a lot of entrepreneurs come to us for advice and there are a lot of people that gave us advice over the years, so we try and make time for that. But oftentimes the ideas that come in, people try to understand all sort of 4 corners of the idea from the beginning. If you can see all of the problems, it probably isn't a big enough idea.

Speaker 1

20:10

The reality is it takes about as much time and about as much effort to build an iPhone app that tells people when you're running late, which is just a really ridiculously stupid idea, as it does to build CloudFlare, right? And that's, you know, so which do you want to work on? And if you're an employee, which do you want to work on? And if you're an employee, which do you want to work on?

Speaker 1

20:34

You want to work on something where you actually can really make a dent in what the universe is doing. And, you know, today, 1 out of every 20 web requests, 5% of web requests flow through our network. You know, it's huge, huge, huge organizations rely on us to be available and online, and it is incredibly stressful. But at least we matter.

Speaker 1

20:57

Like the curse is, if you're sitting around and you see people in your organization who are bored, if that's true, start firing people, because you really do, like, the somewhat, the critique sometimes our team will level on us is that, you know, we always talk about how we're running hot, where we want to run incredibly hot based on what it is. There's never been a company in history that's done a billion page views per employee. We do 5000000000. And that's really, really hard.

Speaker 1

21:33

And we as a result, though, we end up attracting some of the most talented people in the world. Just this week, this guy, Oliver, walks into our office, starts, turns out He's 1 of 7 people in the world that have done the cryptographic signing of the DNS route, and he holds one-seventh of the DNS key, and he says, you guys are doing interesting things, I want to come and work there. Work on hard things, work on big challenges, work on things that make you uncomfortable, make it bigger, and it's going to be hard no matter what you do. But if it works, in 1 case, you've made an iPhone app that tells people when you're running late.

Speaker 1

22:13

In the other case, you've helped protect democracy in Hong Kong and and you know have have the you know government of the UK as a customer and and Really do things that are meaningful and impactful

Speaker 2

22:26

and get to work with really incredible people. Yeah, so again the the when you think about the idea you're working on, you want to make sure it's big and matters. Because when we started, someone backstage asked, did you really think that it was gonna, because we've been doing it for 5 years now, did you realize you were signing up for a five-year term?

Speaker 2

22:47

And so if you look at the data, the average time to exit, whether it's an acquisition or IPO, even though I know Andrew said don't do that, whatever

Speaker 1

22:54

it is. I'm less a negative on IPOs.

Speaker 2

22:57

Is 8 years. Average time is 8 years. But I did not think of that when I started Cloudflare.

Speaker 2

23:03

And you know what, it doesn't matter. Because I was like, this is interesting, I'm signing up for this. I think if it's going to work, it's going to be amazing. If it doesn't, well, at least I can say I tried.

Speaker 2

23:13

And so, you know, don't get too hung up on that. But do pick something that you're going to put your blood, sweat and tears into, because there's definitely a lot of blood, sweat and tears.

Speaker 1

23:20

The other thing that's crazy, I think, is that early on, it's a roller coaster. Yes. And it's up and down and up and down and up and down.

Speaker 2

23:27

Like by the hour.

Speaker 1

23:30

It's no different now. I mean, we had a phone call this morning that didn't go particularly well. And we're driving down here going, wow, that wasn't our finest hour.

Speaker 1

23:42

And then the phone rang again, and it was another phone call, and it just completely switched the day. And that's, and this is Saturday, right? And, and that's, that's just, it's amazing how, how much going up and down matters. And that's why, you know, again, coming back to having other people around you who you really trust and you really value, whether they're cofounders or they're investors, and I think actually that's 1 of the things we've been really good at is picking really great investors, But having someone that you can trust and that you, that when you do have the worst of all possible things happen, you're willing to call them up and say, here's this really horrible thing that happened.

Speaker 1

24:30

What we talk about is our litmus test for who we invite to be on our board, which is essentially who your investor is, is if you imagine the worst of all possible things that could happen, and in our case, that's our network gets hacked and sites on our network get redirected to something else. That's like our disaster scenario. Would we hesitate for a second calling someone who is on our board? And if the answer is yes, they probably don't belong there, right?

Speaker 1

24:59

And If you have any sort of scooby sense about someone who you're about to have as an investor that like, I'm not sure I really want this person, but you know, they're going to give me a really high valuation, run as fast as you possibly can. Like in our last round, we actually took the lowest possible, the lowest valuation that we received, which was half of what the highest was, because we wanted 1 particular person to be in the boardroom, although we didn't actually even technically give him a board seat. But still, he's there, and he's incredibly valuable. And I mean, I think that that was choosing based on surrounding yourself with great people, whether again, it's co-founders, employees, investors, ends up being so much more important over the long term than sort of saying, we're gonna dial, I'm gonna maximize on how much the stock is worth.

Speaker 2

26:01

Yeah, so if people tell you, oh, maximize on valuation, the person doesn't matter, that's not true. All investors are not created equal. And 5 years in, we've done 3 rounds of fundraising, we've raised over $70 million, we've talked to a lot of investors.

Speaker 2

26:19

People who have huge brand names, top notch, they're not all created equal. And so you need to find what we've maximized on and what, again, I really strongly encourage you to do, is talk to them and find out what the vision for the company is. And you wanna make sure the visions align. And I'll tell you the ones, the investors that we have, our visions align.

Speaker 2

26:40

They wanna make the internet better. And that might mean making decisions today that delay revenue and all these different other sorts of things and growing at a responsible rate. So, you know, we and we've as a board have said that's what we want to do. We've had other investors who, again, are really great investors who said, oh, you have to go faster, you have to grow faster, you have to do it, you have to hire 4 times the number of people you're hiring and we're just like we don't want to do that because we don't think that that's responsible

Speaker 1

27:07

and we get another company might

Speaker 2

27:09

like I choose the other and that's fine But you want to align if you but once you make a decision on who to take money from You can it is very hard to get out of that decision. So it's okay to have those conversations up front. And we, again, well-known investors, we've gone very far down the path with, and imagine you bring the board along with you, and then you've said, you know what, we're actually not gonna go that way because we disagree with their vision.

Speaker 2

27:32

And you can do that as management of your company. And so these are just the sorts of things that if you pick the right people along the way, the other people around the board say, okay, well I trust you guys. So if that's it, then we back you. And let's find the right partner for us to build the company into everything it can.

Speaker 2

27:51

And so those are examples of where people really do matter.

Speaker 1

27:55

So we're out of time. We're out of time. Thank you so much for tolerating our sort of rambling discussion.

Speaker 1

28:03

And good luck. Yeah.