CRN Interview: Scott McNealy, Sun

Sun Microsystems Chairman, President and CEO Scott McNealy has led Sun for more than 15 years and is currently navigating the company through a transition from a high-end server provider to an end-to-end software and systems company. McNealy spoke with Editor in Chief Michael Vizard and Senior Editors Elizabeth Montalbano and Joseph F. Kovar about Sun's strategy at this week's SunNetwork show in San Francisco.

CRN: What role will the channel play in Sun's future strategy?

MCNEALY: Fundamentally, we are the most partnered company in the world. We are the most channel-dependent, and that includes the integrators, the hosting and outsourcing companies; that includes the resellers, the master resellers. We are leveraging that. And the human nature I always have to fight in this company is that if we start to get a little more successful, the field wants to hire more folks. I'm getting smarter and smarter about these things. I've learned some lessons here. I hope I can remember them, although we don't always. But I don't want be signing more real estate when we grow; I don't want more head count when we grow. R and D, yes, because that's what the channel wants. The channel wants technology. They don't want a lot of people bumping into them. They don't want nobody from Sun out in the field. They want people out there to help them and help them navigate inside the company and get them access and get them where they can get their training and all the rest of it. They just don't want to be getting kicked out of the customer's room because there's no room because it's filled with Sun sales and support people. We got rockin' and rollin' there and we just overhired in the field, and I think that was a big part of why we had to do some cutbacks when the slowdown came. As we grow next time, I want to see us drive our revenue by employee much higher by leveraging the channel much more aggressively. I think we get it this time.

CRN: What are Sun's numbers in the channel?

MCNEALY: I can never figure them out. It's the ones that matter that matter. It's not the numbers that matter.

CRN: When you look at a company like HP, there's this perception that they have huge numbers of partners and they get the channel.

MCNEALY: They're going after the dollars our partners want. I'm trying not to go after those dollars. I'm trying to go after the product dollars.

CRN: Are you saying they're going more direct?

MCNEALY: I'm saying they're going after more of the systems integration and hosting and all those other dollars. Absolutely. That's not going to be news to your readership. That's not news to anybody. That's their stated strategy. Our stated strategy is to partner.

CRN: So is there anything coming down the pike in the way of partnering and programs?

MCNEALY: You guys are talking to the wrong guy in terms of going into all those details. . . . [Sun Channel Chief Gary [Grimes is the guy. I will say I spend a lot of time with our channel partners. In fact, I'm going to a customer roundtable with one of our big ones from the East Coast. When I was in Boston last time, I did a whole big Boston-area reseller roundtable. We have our regular reseller conferences where we have a Sun network for just the channel partner. We check in with them. I exchange e-mails on a regular basis with a bunch of them and we try to take their input and get their input. But I'll tell you they've been happier over the last three months with the direction we are going. I'd be surprised if your research doesn't show that. Because the anecdotal [evidence, which really isn't anecdotal because I get a lot of it, they are very excited about where we're headed, and there's a simple reason: I ain't giving the field head count. I'm just saying, 'you go partner, because you're not going to get the coverage by going direct.' They like that.

CRN: As Sun moves toward becoming more of a solutions company rather than selling just hardware, how do the partners fit into that?

MCNEALY: The message is we get the partners in a room like this with the customer. We tell the customer, 'Scribble on the white board your nirvana.' And the partners and Sun all run back into the lab for a couple of hours and fix it up and show it to the customer. We go through this whole routine. When it's all done the customer says, 'Yeah, I like that. It screams, 'we've got a proof of concept, and we get 90 percent plus hit rate when we get a customer in with our customers.' There's no such thing as going into an iForce center and just talking to Sun people. It is by definition a multivendor partner lab with hardware, software and integration partners, and if the customer says, 'Yeah, that's what I want,' we write up a bill of material and say, 'Here's what to order from each one of us.' Then we have the Customer Ready Systems manufacturing organization [to which then we ship it all to Sun into our factory. We assemble it, test it, configure it, tune it in the production configuration, not in the lab proof of concept. If that's all a go, then we ship it to the customer either to do their own provisioning or we ship it to a SunTone certified service provider who may have been part of the program to begin with, the iForce program. The whole concept is we do all of the integration, all of the stuff in the lab and manufacturing environment, not in the customer's shop. And we do it all with partners. Literally, we have thousands of partners who are all helping us on these things, and I think we have 15,000 executives through the iForce centers doing these proofs of concept. It's very, very powerful. We have 75 of these iForce centers worldwide. 50-plus of these centers are partner managed. They're the lead. There's always got to be a lead who owns the space and invites the other partners in. We let our partners take the lead. This isn't Sun taking the lead. I don't care who the partner is, whoever wants to be the lead. If it wants to be EDS . . . i2. I don't care who it is.

CRN: How does the pulling together of the open-source stuff and licensing model around the application server on top of the Linux hardware change the economics of the game for the people in the channel, compared to ,say, what they would deal with from an HP or an IBM?

MCNEALY: I'm not sure I understand the question.

CRN: You seem to be loading more value in terms of software in with the hardware from the get-go. Does that mean the ideal channel partner for you is somebody who has intellectual property in terms of software and develops applications on top of that, vs. just being your basic integrator?

MCNEALY: That's always been our message from the time we started. Go back and look at my speeches 15, 18 years ago. We said, 'Gang, you can't be box shifters, and you can't be just straight integrators of the hardware.' There's no value longer term plugging third-party DRAM into the box. We have always said you've got to have business-process reengineering expertise, you've got to have some unique hardware or packaging expertise, you've got to have some unique applications expertise. You've got to have something vertical, market customer understanding and expertise. We've always said that and again. That's not news to these readers.

CRN: But at the same time, it leads to the perception that you have a small channel vs. a highly selected channel.

MCNEALY: I don't understand that. I don't have the perception that we have a small channel. I think we have one of the most far-reaching and broad and technically competent channels on the planet.

CRN: But in terms of sheer numbers compared with an HP-Compaq.

MCNEALY: With HP-Compaq how much value can you add to a printer? That's an HP reseller. There's a different between a VAR and a broker. Why is Dell winning? There is not one other computer maker in the world that's making money on selling Wintel machines except Dell. Dell announces that they're a value-add, big R and D, computer designer and manufacturer. That's not true. They understand that the only value you can add to a banana is a bruise, and that the longer you have that banana in your possession, the more likely you are to bruise it, and certainly the more expensive that banana is going to become. Intel and Microsoft have awarded Dell for not adding any value so that Intel and Microsoft own the architecture. So Dell announces this fantasy of value-add, and then adds zero value. They understand they are totally a broker. They are Safeway. All the other ones are trying to add systems integration and support and the rest of it, and then they have to sell it out through the broker. So you get Dell plus [that and the value add that they offer is just bruising the banana anyhow because the customer wants the standard, off-the-shelf white box, the standard, latest Microsoft release and all the rest of it. You get this whole convoluted system, all of which is operating under water big time. And Dell's cleaning up, just cleaning up.

CRN: Why do you have to buy 100 PCs to get the new Sun Linux desktop?

MCNEALY: There are no economics to buying a Wintel box for your home vs. the enterprise desktop from Sun for your home. We don't have the channel, we don't have the economics and the value of the enterprise client is that it's a client to an end-to-end systems architecture of network-based services. The PC is more of a stand-alone environment to run your games and surf the Web. We're talking about a mail, calendar, app, Net-based client, more dedicated use kinds of environments where you have a clear, dedicated application that you put on top of the server and provision through the portal to the desktop. That's just a different model. And the economies clearly set in at something other than one, and probably less than 100. But we just picked the number of 100 because that's what a call center might be, a health-care outpatient office might be, or some small workgroup that has a dedicated task, a retail bank outlet or something like that where you have to have a server environment.

CRN: There are a lot of custom apps built around office using VB. Are there places you're going to target StarOffice on the client to more so than others?

MCNEALY: Yeah, dedicated, limited-function kinds of things where people aren't power [Microsoft Power Office script users. There are a lot of people out there who couldn't tell you what an Office script was if it hit 'em in the head, and they couldn't even tell you what anything inside of [Microsoft Word besides backspace, delete, cut, paste and print. There's a whole bunch of Office users that don't need anymore than that because they're spending their whole day doing the application. If you're an order-entry clerk, you're not messing around with PowerPoint presentations. You're goaled on getting these orders in and getting them in fast and getting them in right. You don't even know that Office exists on your computer. There are a lot of people that actually use their computer to get their job done.

CRN: Is there any thought to putting a scripting language into StarOffice so people can build applications?

MCNEALY: It has one. It has all that. And I'm not a guru on that and I would never touch it because I don't need StarOffice to get my life done. I figure I'm a way more effective CEO because the only time I click on StarOffice is when I click on a Microsoft document. When I click on a Microsoft document and I hate 'em, up pops StarOffice and up pops the document fully rendered for me. It's been a year and a half since I've done a Microsoft document and it hasn't rendered. When I'm grumpy, I just send back Microsoft documents and say, 'Send me something I can read.'

CRN: Has there been any movement around the WS-I, and how you feel about that whole effort?

MCNEALY: We've got people who worry about that. I don't spend any time worrying about it. If they don't bring us in, that's kind of a problem for them. I don't consider it a problem for us. Sun will adopt anything that the planet adopts that is royalty-free interfaces. The planet adopted Windows, but we won't adopt that because it's not royalty-free, it's not an open interface. If anybody wants to know how this whole antiterrorism, diplomacy strategy works in standards bodies, we could write the book on it. We've been doing this. I'm kind of bored with them now. We've kind of trained the world, there's a whole bunch of people that know how to go do it. And there's a whole bunch of people who are not interested because of their monopoly or post-monopoly positions. We invited IBM and Microsoft into the Liberty Alliance. In fact, we weren't even in charge. We didn't put ourselves in front of it. If they want to be that way, that's their deal, that's their game. They're probably doing the right thing for their shareholders. I fail to understand it.

CRN: Do you feel that during the dot-com heyday, when everybody was selling as much as they could, that everybody got a little bit greedy?

MCNEALY: Define greedy.

CRN: Well, we decided that IS stood for 'implement something.'

MCNEALY: Sloppy, maybe. Greedy, no.

CRN: OK, sloppy's fine. When people got sloppy, Sun itself got a little more focused on large data center and 64-bit systems and took its eye off the 32-bit ball?

MCNEALY: I'm a little more defensive of my team than that because nobody monetized the Internet bubble more than we did. Nobody gained more share, nobody put $6 billion in the bank like we did and protected it, nobody managed the downsize of the bubble more responsibly. As I said this morning, there aren't any of my executives sitting there on post office walls with serial numbers hanging from their neck. I think if you look across the board nobody got defocused, nobody went off and did huge, crazy mergers and that sort of thing. I think the team has done an excellent job. No question that market opportunity for selling servers, the biggest whiff we ever had was most routing on the planet ran on Sun OS in the mid-'80s and we missed the routing business. Well, we were busy! Yeah, so fire me, right. Yeah, in the early '90s, we were busy and we didn't really notice the storage thing, we got to it eventually. So in the late '90s we said, 'hey, this low-end server space is pretty cool, and it's been the fastest-growing part of our server business for the past five years.' And we're still gaining share in that thing. So, am I upset? No, we can only do so much. I look back and I go, 'I didn't sleep a lot in the last 20 years.' So you can say we got sloppy, you can't say that we got greedy. It's just, we were busy. If you'd told me 20 years ago we'd be struggling as a $12 [billion, $13 billion computer company with absolute No. 1 market share basically across the board and crushing the competition in market share, and then for instance, Hewlett-Packard would have lost $622 million last quarter alone in the computer business as a competitor before the $2.5 billion write-off, and that IBM has never made a penny basically in their Unix business. If you'd told me that 20 years ago, I'd have taken it hands down. Yeah, I suppose if I could absolutely predict the future, we could be big compared to where we are now. But I think given the day that we've made some pretty reasonable decisions not perfectly. I always say we make more mistakes in a week than most companies do in the year. But that's because we're doing more than them. What have we done in the last 150 days? I just ripped off some of the big things that we've done. We're just movin and crankin and taking some chances and trying some new things. Yeah, the Java station didn't work, but Sun Ray is. We think we've learned some things and we know how to come back at it. StarOffice 4.x wasn't so hot, 6.0 is a pretty gosh-darn good product.

CRN: Why do you think Sun got hit so hard?

MCNEALY: Because we were doing the best and fundamentally we had the worst year-over-year comparison by far. The other one is I let the headcount [grow. When you're in the bubble, it's hard, and when people with green-spiked hair and tattooed and pierced body parts are throwing greenbacks at you and saying 'Where's my server?' and your lead times are growing and growing like this [holds his hands about six inches apart, and the people coming in saying, 'I'm working my butt off, I need more hires,' You have a hard time saying no. You have a hard time saying no to the guy who wants 400 Netra servers. Enron came in and said they were going to buy 15,000. How did I know? Some of them did. AOL was buying them by the crates, and still are. Exodus was filling up data centers faster than we could believe and all of a sudden, boom--they're gone. Excite@home [was going like crazy, boom--they're gone. WorldCom was buying $100 million a quarter. Who would've guessed? It's easy now, but back then it was really hard to be totally responsible. I still think we were more responsible than most. How irresponsible were we?

CRN: Are you looking for more visibility these days post-boom for the business as a whole, and saying we need to see what's going to happen better than we did in the Internet bubble days?

MCNEALY: I've totally given up on forecasting. Forecasting is not a process I'm trying to improve. We're trying to get more instantaneous feedback on purchase orders received. To me, that is way more interesting data than anybody's forecast because nobody in the world has any credibility with me in the world on forecasting. What I want to do is try to straight line the last four minutes.

CRN: That's a message to the guys in the channel. They can help you do that.

MCNEALY: Absolutely and we're bringing them in. One of the big new processes we put in is the channel portal. This is a key community of ours. We want them indirect, we want them to be able to enter orders, get access to bug fix information, get everything they need instantaneously. If we can shrink that communications time, it's a big, big win.

CRN: It's been a few months since you flattened the organizational management structure. Still having as much fun?

MCNEALY: It's a blast. We are not doing it just at my level. When we went back and looked at it post-bubble, when you come to a screeching halt. We were about 16, 17, 18 percent managers. By the end of this calendar year, I'll have this down to 10 percent managers inside the company. You don't have so many chiefs all trying to make decisions. There are really clear charters. There are no more charter issues inside Sun. Everybody knows what to do, who's doing it, who's in charge. All the processes have clear owners, not committees. [We know who's making decisions, who's the dictator on the decisions. They're all getting graded, evaluated and ranked.

CRN: So the goal is to make Sun an easier company to do business with?

MCNEALY: That was the other message I sent this morning. One Sun, one sales organization, one face. Now that doesn't mean we don't have areas of expertise within Robert YoungJohns [the new head of sales. But if you have a sales title, you report to YoungJohns. If you have a marketing title, you report to [Mark Tolliver. If you have an engineering title, you report to one of the product groups. If you have a finance title, you report to the CFO. If you're in charge of making something, you're in manufacturing. It's real simple. If you do IT projects, you're in the IT organization. It's real simple.

CRN: Talk about how important acquisition is to Sun down the road.

MCNEALY: We will be mostly organic growth, but we will add pieces of technology. The biggest acquisition we've ever done is the Netscape software suite. It was kind of a three-year, screwy acquisition. The net cost of it was near zero for us by the time you got all the AOL purchases, payment, revenue and all of it. It turned out to be a real sweetheart deal.

CRN: You sat on GE's board. What do you think of the whole furor over former CEO Jack Welch's compensation?

MCNEALY: You know, I think Jack negotiated a package and made a mistake a long time ago. He should've taken all the cash up-front, and he could've gotten more cash up-front if he had known that he would be shamed into not taking something that was legitimately and reasonably and responsibly offered to him by the board of directors to keep him from leaving GE. If he had left GE in 1996 and gone over to Honeywell or AlliedSignal and taken [current GE Chairman and CEO Jeff Immelt, Bob Nardelli . . . and 15 other senior execs, the GE shareholders would've paid a lot more than airplane rides for the rest of his life to keep him there. And nobody's written that story, have they?

CRN: How focused will you be on selling Sun ONE as software vs. selling Sun ONE bundled on servers?

MCNEALY: The problem with selling Sun ONE on other people's servers is there aren't many other servers to go sell to. HP is fundamentally going away, is going to be going to Wintel. And the app server and the directory is bundled into Windows. So that's hard to sell into that market. It's free there. What's the other server? IBM, and they're bundling WebSphere in, so it's kind of hard to sell an app server when they've bundled in WebSphere for free. After that, what am I going to do, sell to the Mac server market? There's no 'there' there.

CRN: So you're saying your software sales will be tied mainly to servers?

MCNEALY: It's kind of like asking Lucent [Technologies, what kind of market is there for your call-forwarding software on your telephone switch? There isn't. This is what I mean by it's a feature, not an industry. This is why BEA [Systems and Veritas [Software and even IBM WebSphere have a challenge.

CRN: So PCs are the only place you buy software separately.

MCNEALY: In the world. When was the last time you bought software for this [holds up a tape recorder? You have 100 microprocessors in your automobile. Did you buy the left-turn signal software separately? You've got a thermostat. Did you buy the software separately for that microprocessor? You've got an alarm clock; it's got a microprocessor in there. You don't go out and go shopping for the OS and the middleware separately in anything anywhere except in the computer business and mostly in the PC space. And that model is so old, so dead, so last millennium. I don't think we're ahead of our time, I think we're just ahead. We've done two things. We've opened the interfaces, and two, we allow you to take most of it out and put other things in and it will still work. Integratable and open interfaces and you solve the customer frustration you had when DEC didn't move fast enough. So if we didn't move fast enough on app servers early on, they used BEA, no problem. But once we caught up, now BEA has a problem.

CRN: How are BEA and Oracle reacting to your software push around the app server and bundling MySQL open-source database on servers?

MCNEALY: They should be happier. Well, Oracle should be happier because now they have more money to spend on Oracle apps and databases. They won't have to spend it on app servers, because you know Oracle didn't win all the app-server business. They were losing revenue dollars to BEA and IBM. They should be thrilled to death with this. BEA's got a different problem. I don't know where they go. There's no 'there' there for them. When we put a file system and clustering in Solaris, what does Veritas do?