With Forrester Research Inc. predicting that the emerging application service provider (ASP) market will be worth US$6.4 billion by 2001, few companies are more ideally positioned to capitalize on that explosive growth than Citrix Systems Inc., the Fort Lauderdale, Florida-based outfit whose world revolves around application-sever technology.
Best known for its WinFrame and MetaFrame server-based application deployment and management offerings, Citrix is already drawing considerable attention. Fortune magazine this year ranked Citrix No. 5 on its list of 100 fastest-growing companies, and Citrix technology has been adopted by 98 of the Fortune 100 firms.
At the helm is President and Chief Executive Officer Mark Templeton, who joined Citrix in 1995 as vice president of marketing following management stints with UB Networks, Keyfile Corp. and LANSystems. Templeton recently sat down with Computerworld Hong Kong Editor Don Tennant to engage in a freewheeling discussion that provided an intriguing view from the top of a company that appears to be in the right place at the right time.
Computerworld Hong Kong (CW HK): Sun Microsystems just unveiled its Sun Ray 1 network appliance, which, unlike the JavaStation, requires no processing on the client side. Is that a positive development for Citrix?
Mark Templeton (MT): We were a part of that announcement. Our role in that particular device is to deliver the Windows content to it, just as we were very much involved in the original JavaStation launch, providing the Windows content to the JavaStation, as well as in their recent iPlanet announcement, which is their ASP architecture. So Sun has been a partner, and our role is to bring Microsoft Windows applications and environments to their idea of a device working under an ASP model.
CW: When Sun acquired Star Division, it got the StarOffice cross-platform suite of office productivity apps. What's your assessment of that move?
MT: To the degree that they build a device that can run StarOffice -- that has the JVM and has whatever OS, I guess Solaris -- they can do that. But what a waste of time. I mean, buy a Macintosh. If you don't want to run Windows, buy a Macintosh. Why would you want to do this? Same thing with Linux -- if you don't want to run Windows, buy a Macintosh -- there are thousands of applications for it. The need for an operating system is only created by the existence of applications.
CW: What do you make of Sun's adoption of a thin-client strategy that supports Windows, given its perpetual anti-Windows rhetoric?
MT: They can't be credible with customers that are outside the space that they've already carved for themselves without a Windows story. Turn the tables around with your favorite Sun salesman -- put his or her shoes on and think of yourself going in to pitch this great set of ideas, and how do you answer this question: "Wait a minute, I've got all of this Windows stuff -- it's the center of computing gravity. What do I do now? Let me get this straight -- you want me to get rid of all that?" You better be a highly-paid salesperson on that kind of a story.
CW: Network Computing Devices Inc. (NCD) appears to me to be the closest thing Citrix has to direct competition. Do you agree?
MT: Yes and no. We don't really have any significant direct competition. There are other ways of doing what we do that are just alternative means. And there are some similar things like NCD that pick up a piece of what we do.
We welcome this kind of thing, because what we're competing against is not other technology companies. What we're competing against is the preconditioned mind set around how you distribute computing. We're saying distribute computing in a different way, where you move complexity to the server and distribute simplicity.
So if companies like NCD and Microsoft talk about this idea of running applications on the server and doing application serving, that grows the primary market -- it makes the pie bigger. Our job, always, is to get our fair slice of the pie.
CW: Has the uncertainty about when Windows 2000 will ship been disruptive for Citrix?
MT: Not at all. First of all, we're very much a partner with Microsoft in the development of Windows 2000, so it's not a big mystery to us. Secondly, Windows 2000 is probably the most powerful operating system that's ever been created, and it takes time for these things to be completed and for customers to test and evaluate them, and to have a plan to install and implement them. We have a product plan that I think we've done well to coordinate with Microsoft that says that our MetaFrame product will be compatible with Microsoft. So we're not going to add a bunch of new features when it first ships. You'll buy a MetaFrame box, and whether you have Windows 2000 or NT 4 Terminal Server, as a customer it won't matter -- you won't be making a decision or a choice. We want customers to choose the operating system platform from Microsoft that they want to choose. We don't want to encourage them one way or the other -- we'll have compatibility with both.
Then later next year we will have another release of MetaFrame with an additional set of features.
CW: Can you elaborate on what those features will be?
MT: We're not announcing anything yet regarding MetaFrame 2.0, but let me give you the kinds of things to look for. We will always do more on the user experience side, and that's through Program Neighborhood (a suite of applications that allow system administrators to publish applications on servers and push the applications down to selected users according to predetermined access rights). On the back end is the management side, and we're doing some new things around scaling the management platform to be able to handle much larger environments and much higher performance. We're moving from a load-balancing kind of model to a full load-management system, where the load-management system will be dynamic and predictive. For example, when you or I ask for a particular kind of application to run, it will understand things like whether you are a power user of this application or not, and if you are, it will calculate your load requirements. Some other enhancements will involve authentication and directory services. The idea is to make the management platform much more robust and much more scalable.
CW: If you had to place a wager on it, when would you bet Windows 2000 will ship?
MT: I'm not much of a betting man. I go to Las Vegas for conferences and feel bad about 25 cent slot machines. Microsoft wants Windows 2000 to be right, so I would not be surprised to see it ship late this year or very early in Q1 2000. It's going to be a great product for them and for us, too.
CW: You've always contended that Java is a development language that can't also be an operating system. Do you see last month's announcement by IBM and Sun that they were abandoning their joint project to develop a Java-based OS as a vindication of your stand?
MT: Well, first of all, we weren't the only guys that thought that. Either this is an industry with very astute, diabolical tricks to make this all so confusing that it's hard to choose, or it's an industry that throws a lot of hype on the wall and just tries to see what sticks. I think it's the latter, not the former. In order to navigate through as a company our size, the trick for Citrix is to keep the friction to a minimum, and one of the ways to reduce the friction is to recognize what's irrelevant. You never spend any time on it.
There are guys out there that didn't make that decision and did a whole bunch of development work on that and missed the real thing around Windows terminals. Where are all those Java devices that were supposed to be everywhere? Where are they? I'll tell you where they are: dead ends. You have to make those kinds of calls and decisions; sometimes you're right, and sometimes you're wrong. We happened to be right on that one.
CW: Do you buy the concept of write once, run anywhere with Java?
MT: No. It's not reality. I'm not taking a shot at Java -- it's a challenge for Java in that they want to be write once and run everywhere. But it's been write once, test and debug everywhere. There are a lot of incompatibilities between JVMs and so forth. So if that's ever going to be something serious and significant, it's going to take more time.
Sun has its own agenda. It's funny to me that people criticize Microsoft for working their own agenda and don't criticize Sun, Oracle, IBM. Hey, when you're that big, you have to have an agenda. Customers want companies like that to have an agenda; even we're getting to a size where customers want that.
In one of my former lives working for a networking hardware vendor, I got to host a lot of very large customers, and I'd always ask them, "Who's your favorite vendor, and why?" Nine times out of 10 they'd say Cisco. Cisco's a tough company to do business with, but boy do they know where they're going. I'd ask them why, and they'd say, "Because Cisco will show us the roadmap and the way to be successful. If we stay on that road with them, it's proven we'll be successful." And to do that, you have to have an agenda. It's good for customers.
CW: What you said earlier about Linux seems to imply that you see that as one of the dead ends you're not going to invest in.
MT: Not necessarily. Everyone's keeping an eye on Linux. It's just like the NT story or the Unix story -- you have to sort out what you're talking about -- whether it's the client or the server. If you're talking about the server, what kind of server?
So when I said buy a Mac instead, I'm talking about the client side of the story, which is where all of the hype and drama is. For hobbyists, great. On the server side it's a different story, and the jury's still out on the server side. On the server side Linux holds the promise of being basically a free Unix that can do a decent job as an application-specific server like a Web server or a database server.
CW: What sort of demand for Linux are you seeing from the large enterprise customers that you're targeting?
CW: When you talk to your customers, do they express any concerns about anything that Citrix hasn't gotten right yet that it needs to get right?
MT: It's not around technology -- that's not what they're asking us for They're asking us for a better view into best practices. What that means is this stuff is not plug-and-play; there's a little art and a little science in an implementation. The science you can study; the art is learned.
So what they'd like is for us to beef up our ability to teach the science, which we're doing; but then to also capture the art and teach that, without sending everyone to the school of hard knocks, which is what the world is today. That's one of the big things they want from us that we're not doing right now, and I committed to them that we are going to do something about it. What we're going to do is very straightforward: We're going to build the capability to capture the art, turn it into best-practices education, training and consulting, and then feed that back into our distribution channel as well as provide it direct where appropriate so that we can leverage the learning of a few into the knowledge of many.
CW: It occurred to me that Citrix might well be a company that Computer Associates CEO Charles Wang would want to get his hands on. Do you have any poison pills in place to prevent that?
MT: We have appropriate poison pills for the wrong kinds of takeover situations. But we're not in any way interested in that -- it's not the company strategy. We think we have a long and interesting life as an independent company with multiple channels, multiple products, and doing business on a global basis.
CW: Given that Citrix has no direct presence in Greater China, have you at least come up with a strategy for China?
MT: It's a Windows 2000 strategy around Unicode, to start with. We have a plan to provide a Chinese implementation of MetaFrame, but it's not real soon. Right now the roadmap is about 18 months out, and there's a possibility that we could be pulling this in by partnering with a company in China to help us do it.
We've opened up in Japan, which gave us all the schooling we needed to do double-byte software development and localization. We're just going through the business planning process for 2000, where we're making our Asia-Pacific organization, based in Sydney, think about their world in two dimensions: the world they know today, Australia/New Zealand; and the emerging markets of Asia. The emerging markets we're very interested in are India, China and Korea; and all the ASEAN countries because the use of English there makes it easy. My direction to the Asia-Pacific managing director is to tell us what we need to do to be more aggressive. That's going to potentially involve doing some joint ventures or some acquisitions in these emerging markets to go faster.
CW: What's the most difficult decision you've had to make since you became CEO of Citrix?
MT: The most difficult decisions that you ever have to make are related to people, and making change there. So the most difficult decision I've had to make since becoming CEO is the decision around reorganizing the company in April of this year. When you do that, it creates some disruption and some confusion. It's the price you pay for progress, but it creates situations where you have to take jobs that have gotten too big and you have to cell divide them. Those are hard things to get done and communicate to people and have them feel really good about it when you come out on the other side.