Steve Ballmer: Gartner Symposium and ITxpo 2004
Oct. 20, 2004
Remarks by Steve Ballmer, CEO, Microsoft Corporation Gartner Symposium and ITxpo 2004 Orlando, Florida

ANNOUNCER: Ladies and gentlemen, please welcome Gartner Vice President and distinguished analyst Tom Bittman. (Applause.)

THOMAS J. BITTMAN: Good morning and welcome to Day Three of Symposium and the third of our Mastermind Keynote interviews. Joining me today is my colleague Darryl Plummer. (Applause.) And please join me in welcoming the CEO of Microsoft, Steve Ballmer. (Applause.)

Good morning.

STEVE BALLMER: Good morning.

THOMAS J. BITTMAN: Steve, as we like to do as we start these interviews, we've asked our audience to give us some questions, so if we could cue the tape.

(Begin video segment.)

QUESTION: Yeah, I'd like to ask Steve if there's any intent on Microsoft's part to make developing between the Java community and .NET community a little easier?

QUESTION: I'm wondering if the Linux OS is causing you any problems and whether you're considering opening up the Windows OS?

QUESTION: What about the cost of security we have to pay for keeping Microsoft with us?

QUESTION: I want to know how Bill Gates can think that security is not going to be an issue in two years when we're all adding money to the security budget?

QUESTION: What is Microsoft doing to ensure that the Windows operating system is secure?

QUESTION: I want to ask Steve Ballmer, is Microsoft going to acquire SAP?

QUESTION: I'd like to ask Steve Ballmer what he's going to do to reduce the cost of his software on virtual systems.

QUESTION: I'm just wondering when you're going to change your licensing model to something that makes sense.

QUESTION: When are we going to move to a pricing model that's not based on feature and feature bloat, but rather sustainable, long-term usage of the products?

QUESTION: How are you going to end up collaborating everything in your communications together with your IM, your back-end messaging and even the IP telephony?

QUESTION: I'd really like to know what happened to BizTalk.

QUESTION: I'd like to know when "Longhorn" is coming out?

QUESTION: Steven Ballmer, this is a question for Halo 2. I need to know a little more about the maps, the weapons and maybe stuff like that -- so, yeah.

(End of video segment.)

(Laughter, applause.)

THOMAS J. BITTMAN: I love that.

STEVE BALLMER: November 9th, Halo 2. We'll answer all questions on that day. (Laughter.)

THOMAS J. BITTMAN: So we can't do that now. Okay, all right.

DARYL C. PLUMMER: I'm a Halo 2 fan, so we can stay on that for the rest of the afternoon.

I want to ask you first what do you think of that A-Rod play?

STEVE BALLMER: Well, I think he clearly made an illegal play but it was a heads-up illegal play. (Laughter.)

DARYL C. PLUMMER: You had to be a fan I guess to like that one.

All right, let's get into it. One of the things that we get asked a lot about is "Longhorn," as one of the customers did ask there, but "WinFS," less demand. Who should care about "WinFS" and is the fact that there is not as much noise about "WinFS" the reason why you are separating it and delaying it a little bit?

STEVE BALLMER: No. No, we made a decision to pin various aspects of the "Longhorn" deliverable because we thought we needed and our customers needed clarity. The massive technologies which we're trying to bring together in an integrated form in "Longhorn" is large. The vision is a complete one for really a next-generation platform designed for this day and age, the day and age of the Internet and XML Web services and the massive information that people manage today, but we thought we needed to get clarity internally and externally in terms of what we'd really be able to ship when.

And we made a fairly conscious choice to go from being completely vision-driven to saying now let's decide what we're going to ship in this date, and then what we'll ship beyond that, and that was the reason we moved the "WinFS" component out. It's making actually quite good progress, but not good enough progress to make an '06 shipment.

The truth of the matter is it's hard for many customers to really relate viscerally to the next-generation application development platform unless they are every day somehow building client applications. And certainly the ISVs we talk to who build client-side applications, people like Adobe and Daso Systems and Autodesk, they understand why their applications would benefit from the kind of richer storage capability that we're building with "WinFS." Why is an upgraded file system important? How does it really help support a whole new level of application interoperability? Because in some senses, it's the foundation for taking cut-and-paste and service-oriented architectures and everything else to the next level in terms of persistent information that can be shared in a smart way amongst applications.

DARYL C. PLUMMER: But let's examine that for a second. You're talking a bit about some of the advantages that developers will gain from a "Longhorn" release. "Longhorn" seems to be more even about "WinFX" and "Avalon" and "Indigo" and those advantages. Are you saying that "WinFS" gives those other aspects of "Longhorn" even more advantages?

STEVE BALLMER: "WinFS" is yet another developer advantage in the "Longhorn," I guess I'll call it wave now, because it's not going to be in the "Longhorn" first release. There is a lot in the "Longhorn" wave for software developers. The problem we always have with software developers is they only rewrite their applications a few times, where that may be once, and so the question is how do you talk about the whole wave, which does include new presentation, new messaging approach, new API and extensibility approach through .NET, in addition to a new file system and storage approach in "WinFS." That's sort of, let me call it the complete set of capabilities we think about for the next generation app dev platform.

I was the development manager for Windows 1, the last time anybody let me really line-manage anything -- (laughter) -- but Windows 1 had a kernel, a messaging system, a presentation system and through DOS it had a file system. Essentially every one of those components has a next generation let's call it XML generation advance in it that really is from a developer's perspective what's in the "Longhorn" wave.

THOMAS J. BITTMAN: Itanium really didn't give you a volume play on 64-bit computing but Opteron and now the coming -- I have to get this right -- EM64T 64-bit extension in Intel coming next year might create a different opportunity, a more seamless opportunity for users to move to 64-bit. In the "Longhorn" timeframe, do you see a lot of these folks using 64-bit Windows?

STEVE BALLMER: Yeah, I actually think we'll start to see a lot of pickup on 64-bit in the next 12 months at the server level and at the high-end desktop level. I mean, I will tell you we have customers who have high-end engineering applications who are really pushing on it for the release that we'll make the beginning part of next year that gets 64-bit Intel -- or X86 compatible Windows into the marketplace for engineering workstation applications. Our own engineers love to use it for compile, test, build because it improves productivity, and there are a number of server applications that I think will become red hot for 64-bit X86 next year.

And certainly by the time of "Longhorn" in '06, I'm not going to say every machine everybody buys will necessarily be 64-bit, but I actually think that will happen sometime in '06 or '07, and it will be driven as much by what Intel and AMD do as what these customers do. If Intel and AMD price these things at some point roughly comparable to where chips are priced today, I think they will flow everywhere.

THOMAS J. BITTMAN: So you don't see 64-bit as necessarily an exception choice, it might end up being in some cases a default choice?

STEVE BALLMER: I think it will be the default choice and the question is when and that depends on yields in prices. Probably there's nobody here who will say I don't want 64-bit, that would be my honest belief, as long as 64-bit is roughly the same price as today's 32-bit.

THOMAS J. BITTMAN: All right, we really need to dive into security. Another one of the things that people are looking for are improvements in the area of vulnerability and improving vulnerability, reducing vulnerability and security enhancements. What can they expect between now and "Longhorn?"

STEVE BALLMER: Well, the way you have to think about the way we have learned, and I think at this stage if I can say nothing else I can say I think we've learned a lot more about security than basically anybody else in the world -- that's kind of the good news and the bad news, being in the position we've been in with the kind of market share -- is we really need to focus in on a few things.

One, we do need to engineer in fewer vulnerabilities going forward. We have changed our development process. We have a whole new set of development tools, which we will productize also for our customers to help spot potential security vulnerabilities. We have trained our engineers in a different way. That training is available for third parties. And that does make a dramatic difference; we see it in Windows XP SP 2, we saw it in Windows Server 2003.

I can't say today all vulnerabilities will be eliminated, plan on it. I think it would be nave for anybody to say that, partly because the hackers get smarter, too, and the threat models get more sophisticated, but we've made big progress.

We'll issue our next big release of Windows Server 2003 SP 1, another set of vulnerability reduction. Of course, we have now our consistent patch release program in place that can help reduce vulnerability. But I think what we have to drive for is just improvement in each release. You shouldn't expect to see another release of at least the client other than a service provider potentially between now and "Longhorn."

But even if we were perfect in what we release today, most of our customers are telling us we're not going to upgrade all of our systems, even if you had something you could tell us is perfect, so you have to plan on an approach that lets people protect imperfect as well as perfect, if such a thing existed, which drives us to this question of resiliency. How do you help protect systems? Firewalls help protect, behavior-blocking can help protect, quarantining can help protect.

John Chambers is coming on next. We made an announcement yesterday that their network admissions control and our Active Protection Technologies and quarantining technologies on the network, we've got to make sure that those things integrate so that you can deny access to somebody on the network until you validate that they are in a shape that you want to accept into your network; are they running the most recent patches, do they have AV installed, where have they been, what have they done. There should be a whole set of policies that you can apply before you allow a laptop that's been roaming in your company to come back onto your network.

THOMAS J. BITTMAN: That's in the "Longhorn" wave, though.

STEVE BALLMER: The technology, we have some things that we'll have shipped beforehand and other things which I think the complete division, I have to say, yes, are in the "Longhorn" timeframe. But we're going to stage some of this stuff out with our Release 2 next year.

The key is vulnerabilities must be reduced, resiliency in the technologies to protect a potentially a potentially vulnerable system. You are vulnerable to germs, you take certain protections to protect yourself from getting sick. We need to also help our customers protect their networks, we need to help reduce vulnerabilities, also a very, very important step, and we need to improve the level of education on how to implement these technologies while, of course, also getting the patches out and better tools to help them be deployed.

DARYL C. PLUMMER: But that's a lengthy process, it's always going on, getting the patches out and getting the security vulnerabilities covered. Back in our Spring Symposium, one of the questions on the video, we thought Bill had taken a leave of his senses when he said security will no longer be a top-three issue in two years. The clock is ticking. Can you explain what he meant by that and whether or not that is something that you're actually shooting for?

STEVE BALLMER: If it was something we weren't shooting for, nobody should come to this keynote. Will we achieve that? I think it's permissible for people to have varying points of view about whether we'll achieve it. (Laughter.) But Bill comes back every day and says to me, "Come on, man, what's the problem? This engineering group is not getting it done. Come on, come on, come on, come on, come on. We've got to get this stuff to be in shape, baby, so that it is not a top-three issue in the next two years."

So if you're Bill -- if you're me, I'm not allowed to say that -- if you're Bill, you're allowed to say that which goads the engineers into brilliant performance. (Laughter.)

DARYL C. PLUMMER: But he has actually taken leave of his senses?

STEVE BALLMER: No, no. Some people would say it's a management technique internally. (Laughter.)

DARYL C. PLUMMER: Security protection through personality.

STEVE BALLMER: This group of people and others want us to invest in the R & D that could make that statement come true. We have a lot of stuff we're doing. We think we're going to make a boatload of progress in two years. Whether that statement will be true or not, we can say remains to be seen, but the fact of the matter is it expresses Bill's fundamental optimism about the good work he thinks we're doing in the area. And if we don't get there, we're not going to put it below top-three, we'll keep it top priority until it doesn't need to be anymore, but thank goodness we have it at the right level that he can even be as excited as he is about the stuff that we have in the pipe.

THOMAS J. BITTMAN: So we need to have Bill here in two years is what it means, so we can discuss that, OK, discuss his management techniques. (Laughter.)

STEVE BALLMER: You may have to have him here anyway to discuss his management technique then. (Laughter.)

THOMAS J. BITTMAN: In the line of security, the market perception -- this is a market perception -- is that Linux is more secure than Windows. Now, you have an issue of benefit because there are so many Windows desktops, so many Windows devices in consumer homes, but the perception is Linux is more secure. Is that just not true or is this something where you should be spending more money on marketing perhaps?

STEVE BALLMER: Well, I would say it's just not true but it's sort of a funny thing. You get up and say, yes, yes, we're a lot more secure than the other guy, not nearly as secure as you want us to be but we're a lot more secure than the other guy. Nobody cared. What people really want to know is, do you meet the bar, are you providing us what we need with your products on the security front? The answer to that question is most companies, most customers tell us they want us to do more.

Now, if you actually look at the data today, there are more vulnerabilities in Linux, it takes longer for the Linux community to get a fix out, so the days of exposure to vulnerabilities are longer. You can go to our Web site, our "Get the Facts" part of our Web site, and all the data is there from third parties, blah, blah, blah, but that's truthful, it's still not reassuring to people. But if people are making a decision, Windows versus Linux, I think it's a good decision to bet on Windows, a very good decision from a security front.

Some people say, well, nobody attacks Linux because it's just not popular enough to attack. Believe me, hackers hack for fame, and if there's fame in hacking Linux, they'll hack it too, so there's no safety in picking the unpopular product because if you pick it, it will get enough popularity, it too will get hacked. The fact of the matter is more vulnerabilities and more days of exposure to vulnerability and there's third party evidence on that, which I encourage people to look at.

THOMAS J. BITTMAN: Do you feel you're having any success going after the hackers?

STEVE BALLMER: I think we have actually done a reasonable job in partnership, of course, with law enforcement, who has the ultimate job here. I think we've done a reasonable job of rustling up some of the people who have caused some of the greatest damage and cost and hardship amongst our customers.

With that said, the hackers can, in a sense, pop up anywhere at all times, so I don't think we can expect to ever kind of wipe out the phenomenon just from a law-enforcement perspective. I think it's important that it become a lot scarier for people though.

DARYL C. PLUMMER: So when you start talking about Linux, the open-source question hovers in the air. You guys have not been seen as a fan of Linux nor of the open-source movement. Do you think that -- and we talked a bit at dinner last night about free Linux and what the hidden costs are. Are there any open-source models that you think actually work and provide value?

STEVE BALLMER: The real question is, we compete, these are guys we compete with, right? And I'm happy to tell you at all times why I think what we do is better than what any one of our competitors does, commercial or noncommercial, open-source style competitors. And I think we do deliver more value, I think we can stand behind our products in a way that the open-source community can't stand behind its products. I think we have security advantages, I think we have TCO advantages. I believe those things. That doesn't mean the other thing shouldn't exist, I'll just tell you every time. And if there's ever an area where I don't think we're better, I'll tell our people to go make us better.

We'll bring to market this next 12 months a version of Windows Server for high-performance computing. That's a very important product to us. Why? Because we have had an engineering value proposition deficiency versus Linux in some high-performance computing environments.

So it's a competitor, and I think we compete very well. When it comes time to ask somebody for a new feature in Linux, who do you talk to? When it comes time to get indemnification for intellectual property on Linux, nobody will give it to you. When it comes time to say I want an SLA guarantee on support fixes on Linux, who will do it? The TCO story is very much on our side and that's before you ever get into intellectual property costs. We indemnify for all intellectual property. I mean, heck, we got a $500 million judgment out right now to pay Eolas for browser patents, which we think we'll prevail on, but it's still $500 million. Who would pay that money in the Linux world? All of the companies represented in the room here would owe that money; there's no intermediary to take that intellectual property risk. I think our products got more features, et cetera, et cetera; I mean, I could go on and on.

DARYL C. PLUMMER: If we look a little bit beyond that, we start thinking about, yes, there's a place for Linux in the world, but from your perspective the client-side Linux story in other countries in particular is beginning to build momentum. They're getting better features and capabilities as well. How do you fight that off over the long term? Is it just about value and integrated innovation?

STEVE BALLMER: Today let's talk about the client, because that's where you went. There is no appreciable amount of Linux on the client anywhere in the world. People can sit here and try to read the drama stories from other parts of the world and you can either assume they're true or not. The city of Paris, people said the city of Paris was going to adopt Linux. Well, the studies come back, it would be dramatically more expensive to move to Linux, there's no ROI case for the next seven or eight years to even consider a movement from Windows to Linux for the city of Paris. That's the real data.

The state of Pernambuco in Brazil, same thing, it was reported [as] going Linux. No, in fact, the case is made.

Now, Munich is Munich. There is the city of Munich. Yes, we lost the city of Munich. But the fact that the same story gets told 65,000 times and there's still only one customer and they're still -- how do I use a good, polite word here? -- they're still diddling around to some degree -- (laughter) -- to try to decide when they're really going to do the migration. I mean, come on, where's the evidence?

The evidence is clear. People say China. Our products have higher market share in China than they do in this country, most of it, of course, not paid for. (Laughter, applause.) We didn't adopt the conscious pricing strategy in China to match Linux prices; it's just that's what most people happen to pay us. (Laughter.)

THOMAS J. BITTMAN: So TCO is important on the operational side and it seems like integrated innovation has become a key theme -- Windows Server System, the Office System, Visual Studio, and then there are various initiatives in the industry, on-demand, adaptive enterprise, adaptive this, dynamic that. What is Microsoft's view on self-management, on automation, on virtualization? Where are you going with [DSI, Dynamic Systems Initiative]?

STEVE BALLMER: Well, certainly TCO is a big deal in the broadest sense, I mean, from the desktop to the datacenter, and really I would say the No. 1 place I get pushed today by our customers is as much on desktop management and desktop deployment TCO where there's a set of technologies that are maybe a little bit different than the ones -- I mean, DSI is part of it, but DSI gets even more and more important as you get into the datacenter.

If you take a look at it, we made a decision about three years ago that management and TCO had to be one of our key strong points relative to competition, that if we were really ever going to win long-term in the datacenter, we were going to get serious, so we invested in our Dynamic Systems Initiative, which is a framework for management, including investment in the development tools that let you write more manageable secure apps, the operating system infrastructure that lets those things run correctly, and then the tools like Microsoft Operations Manager, Systems Management Server that let you manage these things with lower and lower cost.

And you should expect to see us drive hard with a complete management approach from datacenter to desktop, from virtual machine technology on up to systems-management tools.

THOMAS J. BITTMAN: It's going to take some time though. This is not "Longhorn," this is beyond "Longhorn," bits and pieces?

STEVE BALLMER: Well, every year something comes and every year, the next year there will be something else that comes. I mean, in a sense can we say -- if we were sitting here five years from now would we say the management initiative is over, management and total cost aren't important anymore? I'll bet five years from now at Gartner, people will be talking about the importance of TCO and management. So every year we're going to be continuing to invest in new capabilities, new innovations that really push this dimension.

DARYL C. PLUMMER: Integrated innovation spans the office-worker, knowledge-worker space as well. Can you give us some insights into the Office System and what your vision is for it? I know you've done a little more specialized work in Office, but what's your vision for Office System?

STEVE BALLMER: Yeah, let me just pick a little bit of a theme you set there and then I'll go to Office specifically. Why do we talk about a concept we call "Integrated Innovation?" Sometimes even our own people get confused about it, because we think it's one of the top requests we're hearing from our customers. Our customers don't want seven ways, different ways to manage all of our products. Our customers would like a coherent development platform that includes Office as a rich client talking to our server back-end.

And so we think most of the integrated innovation points are about reducing cost and complexity and that's how Office thinks of itself and how Office has to think about itself relative to Windows Server, et cetera.

Office itself, people say, what is Office? To a lot of people it's a spreadsheet, a word processor and an e-mail product. That's now how I define Office. Office should be the definitive product that helps every information worker with any broad task that anybody would do: it's workflow, it's portal, it's business intelligence, it's speech recognition, it's search, it's natural language. Office should be a place that helps people with all of the broad horizontal functions that a knowledge worker engages in.

I think it's just this morning we announced a new technology code-named "Istanbul." It is, in some senses, you could say like Outlook but it provides real-time Voice Over IP-type communications. It's just another part of Office because Office has to help the knowledge worker with all of the broad things that knowledge worker or information worker does.

So that's my vision, and as we come out with our next version of Office, you'll see a number of the technologies I talk about show up in the Microsoft Office System.

THOMAS J. BITTMAN: It seems like Integrated Innovation is somewhat the opposite of openness. It's like circling the wagons on the Microsoft environment and focusing on simplicity. Can Integrated Innovation extend to partners like, let's say, Sun?

STEVE BALLMER: Yeah. Well, let me say Integrated Innovation does not mean closed. If a customer says to me, "We really want management to work consistently across all your server products," there's nothing contradictory in that with saying, "And we also want you open to XML Web services, we want identity to federate well with identities that come from Sun or whoever else in the world people may have an interests in." So openness and interoperability and integration to me are not at all inconsistent. Customers want both, they want good interoperability and they want our stuff to behave in a consistent form.

With Sun, since you brought it up, we're running down the path of really trying to get some important levels of interoperability. The first one we focused in on is identity, a lot of good technical work going on, and it's been years since Microsoft engineers and Sun engineers really spoke at all until after we resolved our issues back in April. And I think we've definitely made with Sun six months of progress in six months, and I think within a reasonable timeframe we'll have something interesting to talk about.

DARYL C. PLUMMER: These things all tie together though. It's very interesting; you were talking about Office System being sort of a central place for the knowledge worker to operate. In partnership with companies like Siebel or SAP, that would mean that Office System would have a larger role in their application architectures and you also have the efforts around XML Web services and "Indigo" and "Longhorn" to build the infrastructure application architecture. So all of that seems to be the strategy going on here.

STEVE BALLMER: Yeah, I agree with that.

DARYL C. PLUMMER: I want you to go into it for a minute because when you start thinking about this and bringing "Longhorn" and bringing Avalon and bringing Indigo out there, people start saying things like, well, "Indigo" is your enterprise server spot, but that seems a little shortsighted of the strategy that may be underlying all of this. Can you give us your insight?

STEVE BALLMER: Well, let me start on the rich client, because I think that's where you started. Wouldn't it be nice if, in fact, we could get the level of integration people want between enterprise applications and local intelligence, specifically in Office on the client? We've got a thing we've done with Siebel where you can entirely now front-end the Siebel business logic with an add-in to Microsoft Outlook, which for most sales reps is the way people actually want work. Outlook is a base station for many sales reps, it's got a reasonable UI, and if you can populate and participate in the sales-force automation workflows in the business from within Outlook, that's a good thing. We've done that work with Siebel or we are in the process of completing that work with Siebel.

Salesforce.com, you can do an entire front-end of the salesforce.com hosted CRM from within Microsoft Office.

Take workflows in SAP. Many users in companies in the room don't actually have SAP software installed on their desktop. If somebody wants to take, say, an invoice, e-mail it to somebody, get a comment and an annotation and then put it back into SAP, we need that kind of level of, let me call it interoperability between the Microsoft Office world and the SAP business logic world, and XML Web services technologies, technologies like "Indigo" are part of facilitating that level of interoperability. You could talk about at the server level, but it's just as important down at the smart client level.

DARYL C. PLUMMER: So you go from the smart client to the server, service-oriented architecture, it's on everybody's lips. "Indigo" supports that metaphor. Does it become more important -- Microsoft is not then as open with service-oriented architecture in the distant past and the recent past you're coming up. How important do you see service-oriented architecture as being to the vision that you just laid out?

STEVE BALLMER: Well, there are two independent thoughts: One, do we love open XML Web services, and two, do we love service-oriented architectures? The truth is we happen to love both, but somebody could decide to be service-oriented and proprietary, or they could be XML and not take a service-oriented architecture approach.

When we designed our "Indigo" technology, "Indigo" is a next-generation, runtime-system building-block for writing XML Web service-oriented architecture applications using .NET, that's what it does. It will let you talk to non-.NET applications, XML Web services running on UNIX, Linux, you take your pick, but it was designed precisely to bring the benefits of XML Web services and service-oriented architectures together, at least in the Windows .NET world.

THOMAS J. BITTMAN: Microsoft development has been about productivity, and yet you also want to extend more into the enterprise. Can you do both? Can you extend .NET both for improved productivity and reduced complexity as you move into enterprise capabilities, or is this a dichotomy?

STEVE BALLMER: No, I don't think it's a dichotomy. I think people want us to say we'll do A or we'll do B; most of our customers actually want us to do A and B. And that's --

THOMAS J. BITTMAN: And Halo 2.

STEVE BALLMER: And Halo 2. But come on, that's what innovation is about. Innovative people are supposed to figure out how you can do something complicated but do it in a simple way. That is what -- I mean, a lot of what the innovation in our industry is about is exactly on that line. We used to joke around in marketing meetings that every product add had to either be powerful or easy or powerful and easy or it had to be advanced and simple or advanced and simple. Even our most recent Windows Server ad, "Do More with Less," the truth is our customers push us to innovate on both fronts, the value functionality feature front and on the simplicity and cost front, and I think our engineers have risen to that test many times and will rise to the test in terms of enterprise software development.

THOMAS J. BITTMAN: But it's a challenge to, I suppose, be all things to all people, and one thing that came up recently that was really interesting, Oracle is in the middle of a hostile takeover bid for PeopleSoft, and Microsoft passed on the word that they had also talked to SAP. So can you explain a little bit more about your strategy around Microsoft Business Solutions, where you want to take that? Is that small, medium? Are you looking to extend much more into the enterprise?

STEVE BALLMER: We sell today business applications, Great Plains, Navision, Axapta, Solomon, we sell those products [to] small companies -- I met with a lady the other day, six employees, she runs a very small not-for-profit in D.C. -- on up to very large companies, billion dollar, $2 billion manufacturing companies, divisions of large companies, we sell those products to that entire range of business today.

When we say with Business Solutions we're not targeting -- what we really say is we're not targeting the largest enterprises, we're not going to bid on re-engineering the supply chain at General Motors, that's not part of the design point, because we think it stretches too far in terms of the complexity it adds that would simply put the product out of the simplicity bin needed for the current customers that we target.

But when I say we target companies that are a billion, $2 billion, $3 billion in revenue, those are enterprises. They're not all mom-and-pop grocery stands, they are real enterprises. But we're still not going to violate the basic design point which says we've got to be able to keep this thing simple enough to install, so I guess we are recognizing the boundaries there to some degree on the power-versus-simplicity spectrum.

THOMAS J. BITTMAN: Let me add on that. So can stockholders expect to continue to receive larger and larger dividends, or is Microsoft going to become a venture capital fund in the future? You've got so much cash.

STEVE BALLMER: I thought that had something to do with Business Solutions, and I was trying to make the bridge and I was thinking --

THOMAS J. BITTMAN: SAP, SAP.

STEVE BALLMER: Oh, SAP, OK, that's the bridge.

No, look, we're not a venture capital fund today; we'll never be a venture capital fund. I may have gone through a period in my head where I wasn't quite sure about this but I really understand it now. If our shareholders want to invest in venture capital, we should just pay them back their money in dividends, some money, and let them go buy venture capital funds, unless we want to go build a real core competence in venture capital where the answer is no.

Had we acquired SAP, which we chose not to pursue after some discussion, and the only reason we talked about it publicly is we knew Oracle was going to drag it out in court based upon deposition, but had we acquired SAP, it wasn't a venture capital strategy, it was an Integrated Innovation strategy. Some of those scenarios that I talked about relative to Microsoft Office might have been easier to solve had we been part of one company. It would have created new value for our customers and that would have been a positive thing.

THOMAS J. BITTMAN: But that's still true, so we might expect something in the next year or two?

STEVE BALLMER: There are no ongoing discussions, at least not on that front. We have a very good relationship with SAP, even while we compete in some dimensions, but certainly acquisition is no longer part of that dialogue.

DARYL C. PLUMMER: Stay on the relationship with other vendors for a moment. You guys have been friends with IBM and then not friends with IBM, and you've been not friends with Sun and now you're sort of friends with Sun. It's kind of a promiscuous antagonism that you've got going on there. (Laughter.)

STEVE BALLMER: You can't win. Last year, two years ago, it's, "Why can't you be friends with these guys?" and now it's promiscuous antagonism. (Laughter.) But please proceed, Darryl. (Laughter.)

DARYL C. PLUMMER: Let's talk about Halo 2 now. (Laughter.)

Actually, what I was looking for is just a quick answer. You've touched a little bit on the Sun thing and the interoperability you guys are working on. Can you give the audience a sense of what they should expect from the relationship, wherever it goes from here?

STEVE BALLMER: Yeah, look, we decided that there were some things that we and Sun know our customers want us to do. People absolutely want us to get identity to work well across Sun's systems, Microsoft's systems and the respective standards efforts that we're both involved with, the Web services standard, the Liberty stuff, et cetera, and we're working hard on that in conjunction with Sun and a number of other parties.

We've had a couple of discussions with customers to try to get them to prioritize what are the other things that they would most like to see us bridge, not really the Sun world, Sun really represents the broader Java world, because when all is said and done, IBM, nobody owns Java but Sun. And so the question is what are the other sets of issues people want to see us bridge.

One of the people in the video, one of the customers here talked about what are you going to do to make it easier to somehow bridge the Java and .NET world. There are a lot of good questions there, we think there's a lot of need to work on, and we're trying to get our first concrete deliverable so that we can show progress, customers can see that there is progress and we can kind of build the positive cycle, if you will.

THOMAS J. BITTMAN: And a part of that probably is you both have common enemies, you have common competition. Is that competition growing? Is it Linux, is it IBM?

STEVE BALLMER: Well, we certainly compete with Linux and in a sense -- you guys had Scott on stage here, but I think he would say he certainly competes with Linux. We compete with IBM, Sun competes with IBM, but, heck, we compete with Sun. If anybody wants to know, I'll take you outside and I'll be glad to tell you why a Windows Server System beats the pants off of a Sun system; despite our newfound partnership, I'd be happy to have that discussion -- (laughter) -- and so would Scott.

So let's be a little bit careful on this thing. We want to do the things that our customers are pushing us to do with respect to interoperability. We think Sun is an important player, in a sense maybe as important a player as IBM, who you could say, I think, people are a little bit surprised by the Web services work that we have managed to do cooperatively with IBM and others, but I think people see Sun in almost that same level of importance in terms of some of the architectures that they've driven in the industry.

THOMAS J. BITTMAN: And in terms of growth you also compete with piracy, so what are you doing, what can you do to grow your real share, not pirated share in emerging markets? It's been a real problem. Pricing and functionality of Windows and your other products, what can you do to really make those products applicable to new emerging markets?

STEVE BALLMER: Well, one thing I think is important for people to understand about emerging markets, today PCs are not selling to the lower end of the population in countries like China and India. The people who are buying these machines are relatively affluent people, for those countries very affluent. They don't buy low-end PCs mostly, they mostly buy the same kind of PCs every one of us wants to have in our house.

And so in a sense should the prices be lower? Well, not really. This is a market where until the government and situational factors conspire to help reduce the amount of software piracy, I don't think those affluent people -- it's not a price issue. They can not pay, so they don't pay.

Let's say there's a manufacturing company in China that exports around the world. Should it have lower prices on its software than its equivalent competitor in the United States or in France? I think probably not.

The area in which we think we have to be innovative is, as PCs do start to get and need to get a lot cheaper to get outside of business and outside of the most affluent parts of the populations in emerging markets, it's going to take innovation in the products to engineer them to be lighter, smaller, cheaper, something. And not just the software; the bigger issue probably is the software-hardware combo. There needs to be the equivalent of a $100 computer, not just a $400 computer, if this stuff is really going to go down market in some of these countries.

In all of these countries there is a least PC concept; it's called the Internet caf. If you stop and think about it, an Internet caf in a less affluent market is a pay-by-the-drink computer use, and it is an important thing to keep in mind because it has a very important place in the market. We actually have five times as many Hotmail users in countries like India and China as they have PCs in the country because of the popularity of shared PCs.

So there are a number of experiments and thinking like that we do need to do to go down market, but the problem we have right now is people who should be paying for software, the biggest problem we have right now is people who should be paying for software aren't.

DARYL C. PLUMMER: Well, what about other software-pricing models, subscription pricing? We haven't mentioned that one. That's a big deal for some people and I want to know how you feel about it.

STEVE BALLMER: Today you could say we offer subscription pricing at least to business customers. You can do through our Enterprise Agreements, if you want a non-perpetual license, we finance essentially the purchase or use of our software for any large account who wants to in China or in the United States or in any other part of the world.

DARYL C. PLUMMER: Is it by employee pricing, number of employees, number of resources in that shop?

STEVE BALLMER: Or by PC. You could say, look, for the number of PCs you have, here's the price and we'll finance it. You pay us this much per year for three years. If you don't want to own the software at the end, you can pay us a little bit less. We offer the equivalent of a financing option for business customers, not for consumer customers. But consumers who want to finance software don't want to just finance the software, they also want to finance the PC and PC financing is available in many countries today.

THOMAS J. BITTMAN: We've talked about today's issues, we've talked about "Longhorn." Let's talk a little bit about the future. What do you see as the grand challenge, the next grand challenge in computing? And probably more importantly, what is Microsoft going to do to lead change in computing?

STEVE BALLMER: Well, I think there will be many things that there will be many important breakthroughs. I don't think there's going to be one thing we say that is the only next big thing. I think natural user interface is a big, big, big deal, not just speaking to your computer but having your computer actually understand your intent. When I tell my secretary I want to go to the meeting at Gartner, I could say just get me ready for meeting at Gartner. She knows to go get files and do this. To my computer I say, "File, Open, Search." I mean, it's a very low-level dialogue I have with my computer. (Laughter.) So natural language -- and people laugh and say that's futuristic, but natural language is a big thing.

I think we're underestimating today the breakthrough that will be represented by service-oriented architectures and XML Web services. The truth is, the stuff went through kind of a hype boom and then kind of people stopped talking about it quite as much, but what that means for interoperability and app integration, that's a huge, huge, huge deal when you couple it with the development tools and everything else that goes around it.

I think the notion of information availability and how really good is the access that people have to the information in their businesses today, it's still woeful. When I talk to CEOs, not CIOs but CEOs, the No. 1 thing people complain about is for all the investment they make in IT, when they really want to make a decision they don't have real-time access to the information that they think they need to make the decision.

So for all of everything that matters, so giving better integration of systems, better data availability, these are all areas that I think they'll have significant innovation but two of the most important will be I think the change in the user interface over the next five to 10 years and the change in the way applications get built because of XML and SOA.

THOMAS J. BITTMAN: And perhaps change in operations.

STEVE BALLMER: Change in operations will be a huge thing from a cost and efficacy perspective. I was just choosing to focus on the ones that are a little bit more businessperson facing I would have said.

THOMAS J. BITTMAN: OK, OK.

Well, Steve, thank you very much. Now, we will have you back next year and we're also going to try to get that young man to give us some feedback on Halo 2 next year. So whoever brought him, please bring him back next year and we'll see where we are with Halo 2.

STEVE BALLMER: Because he's a cool kid, have him send me some mail [ ], I'll make sure he gets a copy of Halo 2 on opening day. (Applause.) Thank you.

THOMAS J. BITTMAN: Thanks a lot, Steve, good job. (Applause.)

STEVE BALLMER: Thank you.

Read More: