The guys who gave Linux its very own 800 number look to serve the Fortune 1000, without losing perspective.
This San Francisco start-up aims to prove that a Linux company can grow up without growing stale.
|ALL PHOTOS © GARY WAGNER |
Who says support isn’t sexy? Certainly not the three founders of San Francisco’s Linuxcare, Inc. Though a recent advertisement featuring a cheeky parody of 3Com’s Palm V ad campaign may have been a touch too sexy for some, Linuxcare’s support business has plenty of appeal. Big league venture capitalists Kleiner Perkins Caufield & Byers are investors in the company. And what started as an idea for an Open Source VPN (Virtual Private Network) software package between buddies at a Linux user group meeting, has now become one of the major players in the Linux community. Linuxcare’s three founders, Chief Technical Officer David Sifry, Executive Vice President Arthur Tyde III, and Vice President of Marketing David LaDuke recently sat down to dinner with Linux Magazine’s Executive Editor Robert McMillan and Publisher Adam Goodman following a tour of Linuxcare’s futuristic headquarters in San Francisco’s South of Market area.
LINUX MAGAZINE: How did Linuxcare start?
ARTHUR TYDE III:We were in the bagel business. We had this idea for refillable propane cylinders.
DAVID LaDUKE: That’s enough wine for you Art. No more wine for Art. We were all working together on other things and these guys were steeped in the Linux movement because they had been involved with it for five years. My background is that I’ve been involved with a bunch of technologies that have been beautiful and that I’ve fallen in love with. In fact, I’d fallen in love with four technologies. Linux is my fifth.
DAVID SIFRY: Statistically, you’re our ace in the hole.
AT: Every technology Dave has ever fallen in love with has died.
DL: When I looked at Linux, it seemed completely devoid of any real brands, other than what Red Hat had started to do. So we knew this was a great opportunity.
LM: But you guys had been in Linux for five years. At what point did you start thinking that you could base a business on it?
DS: I knew this was a fundamentally strong business model for a long, long time. The hardest part was actually finding the right people to do it with.
LM: So you could have launched Linuxcare two years ago?
DS: Well, no. We had this moment of epiphany so to speak.
AT: I mean, Dave Sifry’s idea of marketing is a colored flyer.
DS: Yeah. I couldn’t market myself out of a paper bag.
DL: The question was, when was the market really ready for a pure service company?
DS:If we had done this two years ago, we also would have had to release a distribution.
AT: To be a pure service play, you have to have someone out there in front of you.
DS: Because we depended on the fact that there were already people using Linux distributions. People already had gotten Linux and started to feel some pain before we arrived.
DL: I think it was really just sort of the three of us looking honestly at ourselves and asking “What do we have that could really be of value?”
DS: Yeah, you said it yourself, I remember…
DL: Go back to basics. Let’s go back to our basics. Before Linuxcare, we had been trying to get funding for an Open Source VPN [Virtual Private Network]-based business plan. This realization happened as we were standing around after a meeting with the Band of Angels [A group of Silicon Valley investors] guys where we pitched our VPN plan. They were very, very nice to us, but basically…
DS: They handed us our teeth.
AT: They handed us our teeth and we were standing outside and we had been pushing this boulder up hill trying to get our business going and Dave LaDuke said, “We need to do what we do best.”
DL: It was one of those things that looks so obvious in retrospect that it seems amazing we didn’t think of it earlier, but I think it took a confluence of things. The three of us coming together; seeing where the market was.
DS: We had to learn what each of our strengths were. We had to learn how to fail in order to succeed.
AT: By the time that we had the team together, and learned how to write business plans, and learned how to do the financials… By the time we found out about all that stuff, we were too late with the VPN plan.
DL: And we had this come-to-Jesus meeting where we asked ourselves, “The enterprise wants to use Linux, but what’s missing?” And the answer was support.
AT: There needed to be a toll-free number.
DS: There needed to be an 800 number. There was no toll-free number for Linux; and we all sort of looked at each other and we just collectively went, “Oh!” and slapped our heads.
AT: Duh! It was so obvious. This was around August of last year, and none of the Linux vendors was really professing any interest in support. If they did, it was for their particular flavor of Linux. So we said to ourselves, “This is so obvious, and we’ve got the background to do it.” If you look at the background of the three founders, I mean we’re incredibly complimentary. David [Sifry] is the technical guru who has worked on Linux forever and lives in the world of source code. My background was with extreme high availability call center type things like emergency road service. And Dave [LaDuke] was extremely good at marketing and branding.
|ARTHUR TYDE III |
|DAVID LaDUKE |
|DAVID SIFRY |
DS: When we turned our business plan around we actually saw that this was the right idea and that we were here at the right time. And we were able to write a business plan, put the financials together and talk to the investors. Because we knew each other so well, turning it around into something that was timely was a lot easier to do.
LM: You launched the company in February. What happened between August and February?
AT: We had to sign this contract in blood. It was scary.
DS: There was this clause about Art’s first born and all of that kind of thing. Sacrifices and stuff like that. Actually, we went to people and we said, “Here’s our plan.”
LM: Who did you pitch it to?
AT: We talked to a lot of different people. Every month, Dave LaDuke and I would very dutifully go down to the Silicon Valley Software Entrepreneurs Forum — which is an amazing venue — where they get entrepreneurs and venture capitalists to speak to people that have good ideas. So we would go and we would listen and sometimes it was useful and sometimes it wasn’t. Some people were further down the road than we were, some were not quite as far.
But we learned a lot. And there we heard this guy from an outfit called The Sand Hill Group speak. A gentleman named Rangaswami who was written up in the Wall Street Journal a number of times for being a very entrepreneurial VC angel investor kind of a guy. And we heard what he had to say and we were both very impressed. We thought “We have to talk to this guy.” And when we heard that he lived in Sausalito, we thought “We absolutely must talk to this guy,” because, frankly, we were really sick of driving all the way down to Palo Alto for meetings.
So we met with the Sand Hill Group in Mill Valley — at a diner — and had a great meeting with them, where we pitched the new plan. Their reaction was “Wow!” It was just a whole different feeling. All of a sudden we had this snowball that was getting bigger and bigger and bigger. Whereas before it was like…
DS: Pushing a boulder.
AT: It was like Sisyphus. I mean after about 500 tries of pushing that rock…
DL: We learned really quickly that money wasn’t going to be a problem for us anymore and we made a decision that money was money, but brains are what counts, so we went with the smartest investors we could get. That has really made all the difference in the world. Ultimately it led to the Kleiner Perkins round.
They’re the smartest, most successful VC company in the world, bar none. And it isn’t just money; it’s brains, it’s the connections, it’s also just the ability to take an opportunity and make it real. And I think that’s probably the most exciting thing for us. You look at something that was just this ephemeral idea and now it’s real. It’s amazing to see that happen.
Here are these people who have accomplished huge things. They’ve got more money in their pocket than you’ve made in the last year. Our first meeting was really kind of cool. We went in and we did our presentation. We took a break halfway through — like a ten-minute break — and when we started back up, John Doerr came in, Vinod Khosla came in, William Hearst, III came in. All these guys came in and we thought, “You guys aren’t supposed to be here.” It was very interesting to watch the process.
At one point, we put up a slide with our financials, because they all want to look at your financials. And John Doerr kind of looked at them, and Ted Schlein and the KP Partners were looking at the numbers and they started to ask a bunch of questions. So we said “Guys, look, these numbers are bullshit. We have absolutely no idea what the numbers are going to be. But what we can tell you is over the last five years, we’ve seen this kind of attendance at the user groups, and we’ve seen the metrics of who shows up change in these ways. We’ve seen major companies announce Linux offerings. Recently, many of these same companies are saying, not only do we have Linux product offerings, but they’re tier one offerings. Right now Linux is a tier one offering for Oracle. When they come out with their NT version, their Linux version comes out on the same day.” So it was interesting.
AT: Right at the end of the meeting I just wanted to lean my head down on the table. Because, you say to yourself, “My God I can almost feel the heartbeat.” It was as if you could hear the heartbeat for the entire Valley just by sitting there. When you start to think about it, these people can easily buy small governments.
“Whoa!” I thought “What’s going on here?” This was one of the reasons I came out to Silicon Valley in the first place — because there’s no place in the world like it. And there’s just no place where you have this kind of energy and entrepreneurial spirit and capital.
Look at where Linus Torvalds has chosen to live. He could live in Finland; he could live anywhere he wants. He lives here in Silicon Valley. It’s a special time and a special place and it’s gone through all these waves of innovation. The PC revolution, the Internet revolution, Open Source is now right at the brink, and it’s amazing to be here right now. It really is. These are the good old days.
LM: It is interesting that you actually brought up user group statistics when you were making your pitch to the venture capitalists.
DS: Listen, people going to user group meetings is a forward indicator. I mean that means you actually have to take the time, get off your ass, get dressed, go out and get in your car and go drive to the meeting and sit yourself down — basically just to hang out with a bunch of other enthusiasts.
LM: How many people work at Linuxcare right now?
AT: Well, let’s see…what day is it? Over a hundred?
DL: I don’t think we’re over a hundred yet, but pretty darn close.
LM: You’ve talked about keeping the fun and keeping the original spirit alive. Doesn’t it get harder and harder as you grow and as money comes in?
DS: That’s a good question. It all comes down to focus. I mean, when we started this thing, we said to ourselves — we have to really plan to scale the business the whole way, whether it’s by getting a big telephone system, or getting CareTracker[Linuxcare's Web-based tech support system] right so that it scales with no problems. But what that also means is that the best way to help a company continue to grow and keep good values is by building a really, really strong corporate culture.
And the thing that we were really lucky about is that we all come out of the Linux culture. I mean Art founded the Bay Area Linux users group. He and I met at its first meeting. And so the people that we brought into the organization from the very beginning, really embodied the kind of culture that we were looking to build.
If you look at some of the great companies that have been built throughout the technology industry, the ones that have succeeded have been able to take the culture they came from and really grow with it. And sometimes they fail when they lose that culture.
LM: What companies did you look to as role models — companies that successfully maintained their unique culture?
AT: I think one of them was certainly Apple in the early days. Linux has that sort of religious feeling. Another one was Netscape in 1993.
DL: Companies get boring when they scale. We do spend a lot of time on culture. I mean we have a very, very clear vision of the culture that we want and humor is part of it. We have to answer to the customer. That’s the whole focus, but when customers call, they want to feel that they’re talking to a human being who can relate to them. If that person is in a relaxed frame of mind they’re going to do a better job.
DS: Every new employee is issued Nerf weaponry on their first day at Linuxcare. It’s standard issue.
LM: How many rounds do you get?
DS: One clip of ammo.
AT: But it just so happens that we’re strategically located a block and a half from Toys R Us.
LM: Do you have a corporate account there?
DL: We should.
LM: What do you think of Microsoft’s corporate culture?
DL: I think there’s a certain part of the Microsoft culture that’s very driven, but I think Microsoft is driven by…I don’t want to say greed, but it’s not driven by the same egalitarian values that push Linux forward.
DS: I think that’s sick competition.
DL: That’s it. They are competitors.
AT: They’re good at it.
DS: And it’s a fun thing. I think if you’re working at Microsoft it’s fun to compete. It’s fun to win.
LM: Do you think the Open Source culture can scale beyond your company, to your customers in the Fortune 1,000? Are these companies ever going to be active contributors to the Linux community in the style of HP’s sourceXchange?
DS: Yeah, actually I do believe that they will.
LM: What’s it going to take to make that happen?
AT: And a certain amount of education. Once people start thinking about open source and realizing its benefits, you will begin to see more and more internal projects. Software that does not provide a competitive advantage will become open source. The reason is purely risk management.
I’ll give you an example. At Cisco they developed this entire printer sharing system. And why did Cisco decide to open source it? Well, because the software’s developers basically went to management and said, “Listen, we may get hit by a bus. This is an insurance policy for you now. What happens if we move on or something else happens? How do you keep adding new drivers? If you open source this, you essentially distribute the risk and you get back a tremendous amount of additional development from the open source community that you won’t have to pay for.”
LM: Is there a role Linuxcare can play in sort of encouraging that kind of education?
DL: Well you know a lot of it is just communicating with a customer. Communicating with people in companies. It’s interesting because usually Linux is installed somewhere in the middle, right? A lot of people in the past have complained that the tops “don’t know about it.” But these days I’ve noticed that top management now seems to be very aware of this.
LM:Won’t having clients with diverse open source projects to support make your job harder?
DS: That’s why we hire the best people. I mean, Linuxcare isn’t focused on level one and level two support. We really start at level three.
LM: What does that mean?
DL: Level one is your most basic level of tech support. Anyone can do it, so you might as well hire people in Poughkeepsie or Pescadero, where you’re not paying the kind of salaries that we need to pay here in San Francisco. Level two requires synthesis, which means putting together information from various different places and coming up with a specific answer. And level three requires depth. Almost all of our people are level three. We have some level two, and we’re investing in people so we can grow people from level two to level three.
LM: You guys clearly understand the power of open source. How does upper management learn about this in other companies?
DS: I think that larger organizations, by and large, are risk averse. In the Linux community there’s this tremendous evangelism, and a tremendous amount of energy. You’re seeing this outflow in growth and creativity. And this incredible grass roots influx of Linux into organizations. And it’s even more grass roots than the Mac was, because you don’t even have to pay for the hardware; it’s already there.
And so a very classic sales example for us is — and I’m not talking about a proactive company — a CIO reading CIO Magazine, or InformationWeek, or something like that. And he will see an article about Linux and begin talking with his networking manager and he’ll say, “Hey, what do you think of this Linux thing? Do you think it’s got legs or what?” And the networking guy will turn around and say, “Well yeah, we’ve been running Linux on our internal network for the last year and a half. That’s what our Web servers are using, our mail servers and DNS servers are using it too.”
LM: Describe the typical Linuxcare customer.
DL: Our focus is the corporate enterprise. Fortune 1,000 or Global 1,000, whatever you want to call it. One of the luxuries of being completely focused on support is that we’re not in the retail channels so we don’t have mom and pops calling. Occasionally they do and if they are willing to pay for incident packs, we will support them.
LM: How many of the Fortune 1,000 do you have as customers?
AT: It’s well into the double digits. It’s the big hardware vendors, the e-commerce players, the .coms. The idea here is to grow the Linux pie; to make Linux really big. So we not only support every Linux box that Dell sells, but we also go in and certify their machines and we try to make sure that they know how to design a machine that’s going to be optimized for Linux.
Or we’ll go into ISPs [Internet Service Providers] and say, “Listen, you have an application for Unix. We can help you port it to Linux.” Or suppose a hardware vendor has a product that runs on Solaris and HP-UX and NT, and they want to get it running on Linux, we’ll write the driver for them.
LM: Aren’t Fortune 1,000 companies freaked out at the thought of an operating system that you can update weekly, or daily if you want, that has no official standard? Just as a matter of principle, doesn’t that just frighten the hell out of them?
AT: No, not at all. You’ve got to remember that with Linux you’re not boxed into this Microsoft NT Service Pack No. 38, or whatever. We have customers who have servers that have been up and running now for over two years without a reboot.
DS: And they’re running things like Linux 1.2.13. Some distributors have very creative ways of dealing with the need for standards. For example, SuSE sells something called their Business Edition over in Germany. And the Business Edition is basically the prior distribution.
AT: No, no. Two back.
DS: It’s two revs back. It’s exactly what they were selling for 40 bucks two revs back in a box priced at $500.
AT: No, no, no. You have to understand: the box is bigger.
DS: Okay, and it includes two additional books. But they sell it for $500.
AT: One of the reasons we got the support contract with Dell was that they originally thought they could standardize with one distribution worldwide, but (at least at this point) they are not that centralized. When they went to their European offices and said, “This is our plan,” the Europeans said, “No it isn’t. That’s not what we’re going to do here in Europe. We want to run SuSE.” And they got a similar reaction from Japan. So they asked themselves, “Do we want to have multiple support contracts, or do we want to have just one?”
LM: Is Red Hat a competitor?
AT: Well, it’s kind of funny. Who does Red Hat list on their S-1 as competitors? I think they list 23 other distributors of Linux. I don’t think they mention them by name; it’s very interesting. I don’t believe that we’re mentioned.
DL: No, we’re not.
AT: Quite honestly, I do consider Red Hat competition. You know, actually, it’s cooptition. It really is. They’ve got their core competency, which apparently has to do with the retail channel. We’re not in that channel. We’ll never be in that channel. Our focus is just different. We’re a pure services, pure enterprise play. And they have a different market, a different channel, a different user base, and a different philosophy and that’s good. I mean, I think it’s good for a market to have a lot of different kinds of layers.
DL: Otherwise it’s like Microsoft. Sure we compete with Red Hat, but we also like to work with Red Hat. The idea is that a healthy market has lots of competitors. It’s sort of like computers back in the early 1980s, where you had Ataris and Commodores, and everybody was leap-frogging each other with features. That’s when it was interesting. I mean, the last ten years have been boring. These days it’s more interesting to walk around Home Depot than a computer store. And now the Amiga is coming back. What does that say? Linux is breathing new life into hardware.
LM: Are all the distribution vendors participating in your Linux certification programs?
AT: Everybody except Red Hat. And we’ve invited them and talked with them, so the door is open. We really want them involved.
LM: Why aren’t they on board?
AT: We think that it’s pretty obvious that one of the risks for a leader in this market is the perception that it’s a winner-take-all market. I think that perception is corrosive to the health of the market. When there’s an early leader, sometimes there’s this view that standards are a bad thing unless they’re set by the market leader.
I think there is a kind of cultural disconnect between the companies basically. They’re in North Carolina. We’re here. We’re part of this community. They’re part of some other community.
LM: Some people have said that Red Hat is trying to distance themselves from the other Linux companies.
AT: Keep in mind that the Research Triangle Park ethos, if you will, is dominated by companies like IBM and the old guard computer companies. So I think that might have something to do with it. I don’t really know.
But the fact that they’re opening up a Silicon Valley office is a good thing. They’re kind of coming out of seclusion in North Carolina and starting to get a little exposure to Silicon Valley. Hopefully they’ll start to realize that there is significant value in cooperation.
LM: Why aren’t people talking about Caldera in this light? They’re based in Utah.
DS: Well the thing about Caldera is this: We talk to Caldera frequently. They’re very communicative, very open. We talk to the Caldera guys all the time, just like we go have beer with the SuSE guys. There’s good two-way communication. We know what they’re about. They know what we’re about. We’re all trying to build the market.
LM: Let’s talk about a more obvious competitor: Microsoft. Has Microsoft succeeded in choosing the battlefield in competing against Linux?
AT: Yeah, absolutely. Honestly they have. They’re attacking Linux. You don’t see Microsoft going after Red Hat. Why? You don’t see them coming after us or VA or SuSE or TurboLinux. They’re going after open source software in general.
DL: To a certain extent open source Software is going after Microsoft too.
AT: The battlefield is a perception thing.
DL: I agree with that. I think “Microscared” [Linux Magazine's, Spring 1999 cover story] makes great cover copy, and I’m sure Microsoft is worried about this phenomenon, but it’s such a sea change. It’s like what do they fight? What can they fight?
AT: It’s a great opportunity. I guarantee you, if Microsoft came out with Office for Linux, I’d buy a copy. Absolutely. PowerPoint is a great application. We would love it if Microsoft Office came out for Linux. We’d definitely support it.
DS: If Microsoft came out with a Linux distribution, we’d support it. We’ve got no problem with that.
LM: Well, what if they came out with a Windows interface for Linux? Microsoft has always been good at controlling the layer of abstraction at which programmers write. What if they released a windowing system for Linux?
DS: So let’s suppose that Microsoft decides; “What we’re going to do is we’re going to port the Win32 APIs to Linux”. Maybe they’ll have it running on top of X, or maybe they’ll just basically have a complete new implementation. You know, a Windows library that sits on top of a Linux operating system.
DL: You know, something that crashes every now and then.
AT: Hey, it keeps things interesting.
DS: If they do that, then we’ve won! If Microsoft decides to release a Windows interface for Linux and keep it closed, I guarantee that you will see the largest reverse engineering effort in the history of mankind. Because we have control of the operating system at the system call level. We would be able to see exactly what their software is doing, and we would be able to easily tell how it works. I kind of hope it happens. I have no problem with it.
AT: Actually, we would support the re-engineering.
DL: When you can finally go down to the very bottom of the machine and see everything that’s going on, it’s game over.
LM: What is Linuxcare doing to support the open source community?
AT: We wrote it into our business plan. We encourage our employees to work on open source projects, even while they’re at work. And there are a bunch of very good reasons for this, not the least of which is that the more open source software there is, the more people are going to need support contracts.
Second, it’s great because it helps our guys to continue to build up their expertise. So when we’re supporting a particular open source application, we can know it more deeply than anyone else.
We’re using this stuff. We eat our own dog food. Let’s face it, it’s good PR too. But most importantly it’s all about the community. If the open source community didn’t exist, Linuxcare wouldn’t be here.
Robert McMillan is Executive Editor for Linux Magazine. He can be reached at email@example.com.