Back to agenda

Financial Analyst Meeting 2006
July 27, 2006

COLLEEN HEALY: While they're setting up for Q&A, let me quickly announce the winners from the eval drawing that you've been filling out.

The Xbox 360 bundle, which includes the gaming console and some cool software games, that goes to Jeff Jacobe .. Congratulations.

The Motorola Q running Windows Mobile goes to Rob Breza. Congratulations.

For the Xbox Live 12 -- Xbox Live membership, good for 12 months, which includes some games as well, that goes to Ted Chou.

And, lastly, the Microsoft Streets and Trips with GPS Locator: The winner is Chris Kwak.

Great. So, with that, let's turn to our last Q&A session. I'm delighted to facilitate it. We are going to run this session the way we did our first session, which is that the investor relations team has paddles. If you have a question, please raise your hand and a member from the team will come to you with a microphone. And, with that, let's start with number two, please.

NEIL HERMAN (Stanford Group): Neil Herman, Stanford Group. Microsoft now has approximately 70,000 employees, obviously a very large company. What do you do to fight the bureaucracy at Microsoft? And clearly the type of people and your hiring processes have changed over the years, really emphasizing teamwork. I guess the other question is: Would the Microsoft of today have hired young turks like Bill Gates and Steve Ballmer?

STEVE BALLMER: Maybe I'll start and if Kevin wants to weigh in.

I would say that, first of all, when we say nurture all kinds of innovation, we have to nurture all kinds of people. The notion, as much as I know teamwork is more important today, you give me the right kind of person who is incredibly bright, incredibly productive, and perhaps less able to work productively with others, we'll find a role for them. And we do that all the time, every day.

So nurturing all forms of innovation isn't about nurturing various forms and kinds of people.

I think size of a company is only one aspect you need to factor in when you think about agility and bureaucracy. A lot of it is kind of attention to detail and policy. I claim Microsoft was more bureaucratic when I started in 1980 than we are today. We have gone through phases where we were less bureaucratic than we are today, and phases where we are more bureaucratic. It's an everyday thing that leadership has to sit on top of and push for greater ability. I may let Kevin supplement that.

On the R&D side, this focus on innovation and integration as opposed to integrated innovation where we really tried to incubate and then integrate as opposed to integrate things while they're being incubated, that's a major focus on improving the agility of our R&D machine, and I think we're going to get a lot out of that, and there's a lot that a number of folks are driving.

But maybe I'll let Kevin talk a little bit, because I know he talked about it at the sales conference last week.

KEVIN TURNER: You know, one of the things that we focus on is making sure that we define boundaries for people and allow for flexibility in the middle. And it gets back to really absorbing and prescribing to the model that discipline and accountability enable empowerment, and then when people get the fact that the discipline and that the accountability pieces that go along with that, once we are great at those, then we allow for flexibility in the middle and that they're able to really bust through a lot of the bureaucracy and the decision-making and the flexibility and the speed and all those things that go along with it.

So we're really about defining the boundaries and the accountability, making sure the discipline and the consistency of that discipline is tight and razor sharp, and then allowing people, turning them loose, so to speak, and allowing them that flexibility to really get what they need to do to be able to take care of our customers and our partners.

COLLEEN HEALY: Thank you. Number two, please.

QUESTION: I think Neil actually kind of got to my issue of how do you assure innovation, but I think we just heard that.

Maybe if I could just ask on Vista, can you give us a sense, the premium versions, it sounds like if I were buying Vista I'd want the premium version so I can run the Glass interface. Shouldn't we expect that we'll see some benefit from the new SKUs, the increase in price?

I'm also curious of your sense, having seen the feedback on Vista thus far, how big the upgrade opportunity is for Vista to run the premium, to run the Glass interface, if it takes a lot of graphics processor, and I'm just not sure that we've got a good handle on what percentage of the market out there for consumers that are likely to upgrade have a machine that's capable of running that.

STEVE BALLMER: I think you've got to look at Windows units in aggregate. And I think there's a lot of reasons to want premium versions but hopefully if we're successful we're going to be successful in some markets where we may also have some lower price points. I mean we don't sell quite at the U.S. price point, for example, in China where we're trying to make the big push so in a sense you get a weighted average of a lot of phenomenon that would go into that. Now it can't be less than 0 but if you add 1 unit X and one at 5X or 4X or 3X, the first X can weight you down because you're asking now a revenue per unit question.

In terms of the upgrade I think people are a little confused. Maybe I'll just spend a minute on upgrades because I have felt like investors are a little confused about the upgrade opportunity. Most upgrades to Vista will happen when people get a new machine. That's been the truth on every Windows upgrade of all time. Corporations many of them will choose to buy a maintenance plan or an enterprise agreement plan and to some degree we've been selling those that included Vista now for a number of years.

There's no spike associated with either of those two phenomena. The third phenomenon is somebody walking into a store, as an example, and buying an upgrade to Vista. That's an interesting phenomenon. It's real strong first quarter pop with a tail but a strong first quarter pop but I'm not giving out forecasts. Don't take the numbers and run them to the bank. Don't put them in your models because we're giving you all the appropriate guidance. But let's say it's someplace between 0.5 million and 3 million units on a product that we might get $89 or 4 million units on a product that we might get $89 on.

It's an interesting amount of money, but it's an amount of money that is a few hundreds of millions in a one quarter timeframe with a tail that goes down from that as opposed to being a multibillion dollar cycle of retail upgrades. And I know Rick and I have talked about this but I do sense that that's part of where people -- we let you get a little ahead of yourself in terms of thinking through the models was that there was a much bigger pop in Vista itself immediately.

Now Vista is an enabler of a lot of the new businesses that we talked about. I think it helps PC growth. It's got a lot to help us in terms of piracy and Windows Genuine Advantage, upscale SKUs but the actual upgrade and so when you ask me how good is the upgrade cycle I think you're primarily asking me how good is the upgrade cycle in stores, and I think we have about the same percentage of machines that will be eligible to run Vista in the install base that we had essentially with XP and some of the earlier releases. There's still a lot of execution. I think there's a lot of reasons people may buy the high-end version besides the glass UI. But obviously one of the key things that both Kevin Turner and Robbie Bach have to get done is to try to get all the units that are out there sold. Because the more heat we create in stores, the more pressure there's going to be for new machines in business et cetera for the Vista upgrade.

COLLEEN HEALY: Thank you, number 3 please.

QUESTION: Can you talk about what kind of changes you expect in the boxes from Vista? I understand it's going to require more memory and more computing power. Could you discuss what you think the standard configuration is going to be in a year out and then two years out?

STEVE BALLMER: Yeah, just tell me where you're coming from -- where's the mike, I lost it. Yeah, tell me where the question is coming from so I can put it in the context of some Microsoft financial issue. I'm just trying to understand.

QUESTION: Well I'm just trying to get a sense of what this means for the overall PC business. It seems like it's kind of slowing down. A lot of people are expecting us to be a fill-up, what's your take on this?

STEVE BALLMER: Like I said, I think most people will choose to buy new PCs. I think most people -- what do you think, most people get a gig, of memory and --

CRAIG MUNDIE: A gigabyte of memory, you're looking at many tens of megabytes of disc storage as the standard configuration.

STEVE BALLMER: Tens of gigabytes.

CRAIG MUNDIE: Tens of gigabytes, that's right, of storage. And so many, many of these machines will all be built on dual-core processors. But I don't think that those things are going to drive the price points up. I think you just see the continuing effect of you get more for your money. And so I don't think it's going to create some fundamental shift up in the price point of the hardware.

STEVE BALLMER: But I do think a lot of people are going to want to buy PCs. Does that buy us another 1 point, 2 points, 3 points of PC growth? It's hard -- now we're talking about a huge industry and it's hard to predict that we'd see, you know, 10 points of PC growth. But I think that will show up in the PC growth rate and therefore in our Windows unit growth rate.

COLLEEN HEALY: A question over here, number one, please.

Trip CHADRI (Global Equities Research): Yes, Trip with Global Equities Research.

I have two questions. One, it's on industry structure, one for enterprise. We are seeing some hardware vendors, EMC, HP and to some extent even Sun, acquiring software companies. The history says this kind of strategy hasn't worked for the past 30 years. Do you think something has changed that makes us feel that this strategy of hardware Windows buying software Windows will work?

The second question I have is on consumer Web, since, as you mentioned, as Steve mentioned that he is spending more time on advertisers. Somehow as analysts we are a little frustrated with the pace of the acquisition Microsoft is doing. Think about a company called Nebo, three-people company, three months time, 4.5 registered users, having phenomenal traction having, only $100,000 in investment. BitTorrent is another example.

Why are we not very aggressive in acquiring these companies, including you, too, and when they become so big then we spend billions of dollars on it? Why don't we nip them when they are so small and so aggressive? Thank you. (Laughter.)

STEVE BALLMER: Do you have an opening in the Corporate Development Group? (Laughter.)

We are buying more companies than we have ever bought, and if the challenge is could we do more still, I'm not going to say no. I will not say no.

Now there are all kinds of reasons, even on small companies, to buy or not buy. Some things get expensive very quickly. Surprisingly quickly. We looked at a deal recently, the company has been around a year, 1.5 years. They wanted over $500,000,000. They didn't -- they had less than $20 million in revenue. We had a hard time closing so we it -- you have to be real --. We thought about it, just so it's clear. We didn't stop from thinking about it though.

You have to get in pretty quick and you have to decide which of these things really have monitizable user bases because there are plenty of ways to buy activity without necessarily augmenting, in a substantive way, your position. And part of what we need to do is crank through that.

CHRIS LIDDELL: If I can add something. We bought 23 companies last year. So that's two a month. So that's $700 million. We are buying them not when they're $100,000 admittedly but if you take the average price that we are paying we are, generally speaking, buying them early in their history.

The other thing you have to remember is, you can look at it from a financial perspective, you have to look at it from an integration perspective. Every time we buy a company we have to integrate it into our company and actually make it work. And part of the pace at which we can buy is our ability to actually bring those people in because there is, generally speaking, large teams from Microsoft, and actually integrate them in and deliver on the promises that they had.

So the pace that we are doing, three times what we were doing the previous year, just bringing those people in and scaling is a challenge. And I think we're about the right pace at the moment.

The other thing to note is we spend let's say upwards of a billion dollars on acquisitions, we spend $6 billion to $7 billion on R&D. From our point of view we are still very much a build a company, not a buy company. We'll buy where it makes sense, where we can accelerate growth, but our heavy focus is on investing in our own people and investing in our own ideas.

CRAIG MUNDIE: I mean, in my talk today you saw two acquisitions that I would say are much closer to what you're advocating than not. The C-Dragon stuff that we bought was a tiny little company that had very powerful stuff and we thought was going to be compelling, and we bought it early in its lifetime. We just announced the Azyxxi acquisition yesterday, which if you look at it is very similar to what you're advocating. So I think we're attuned to that.

STEVE BALLMER: On the first question about enterprise structure, I don't think that there is any fundamental shift going on in the way the kind of hardware/software services mix plays, except for us. I think the fact that we as a pure software play, and SAP perhaps as a pure software player getting relatively stronger, is perhaps a structure change.

But there has always been a lot of interesting or at least important from a volume perspective software that was essentially in the hands of hardware companies. EMC is -- it's hard to know what EMC is, are they a hardware company, are they a software company -- I'm not --. We have a great relationship, I think a lot of EMC. But I think they would bristle if you called them a hardware company despite the fact that they sell some hardware.

The HP acquisition is an interesting one I think, and I just haven't had a chance to talk to Mark Hurd about it yet, but I don't see a fundamental change in the industry structure coming.

COLLEEN HEALY: Question number two.

QUESTION: Just a question on Zune. Last year we talked about or you talked about the competition with Apple and iPod. You talked about the flexibility, having multi-platform vendors. We get Zune here, it's integrated hardware/software/service. What other areas could we expect you to do that, to integrate that offering? That's the first question.

The second question on Live, Office Live and other products. That transition from a license model to that subscription model, that seems really difficult to manage, just from somebody watching on the outside, especially in the SMB market. I mean, how does sales and marketing manage that? How do you not have some sort of revenue dip within a timeframe?

CRAIG MUNDIE: You know, again I gave an example where in our history a lot of times we started looking at the different markets, and our first instinct to some extent is to use the ecosystem to address them, and we did that with game machine, and then ended up deciding for a whole variety of reasons we had to have our own branded presence there.

We I think have come to the same conclusion that in order to get to critical mass and the brand awareness, it required a level of investment that we couldn't do using the traditional ecosystem.

I mean, another example is much smaller, but may ultimately have a similar effect, is the SPOT watches. We started doing those things only with partners, and just a couple weeks ago we also introduced one that Microsoft has actually designed — and, again, in order to be able to change the price point and deal with the branding and merchandising aspect.

So I think in general we do believe in the power of the ecosystem, but sometimes you just can't get to critical mass in either branding or accelerate development if you have to bring the whole ecosystem along, and then we make these strategic shifts.

STEVE BALLMER: We don't have a significant new thing to say. I would say philosophically we used to say we'll never do hardware, with one little exception, and then we did Xbox with the one big exception, and now I think what we're saying is where it's smart and where we need to -- don't expect us to do PCs, I think that ain't gonna happen, but where it's smart in some of these new device types we will consider it.

In terms of the question about essentially subscription transition: those of you who have been analyzing or looking at or investing in Microsoft, we've already gone through one of these huge transitions. Basically a big part of our revenue used to be licenses, and we moved it to Enterprise Agreement, which in accounting sense looks a lot like a subscription, and we wind up talking about this funny unearned revenue; if it's a real subscription you don't have unearned revenue, of course, but that's been a successful transition.

And taking a long term approach, if your customer is in a better shape, then you can make more money from that customer long-term, we'll make that transition, even if it winds up creating some short-term holes.

Now, my opinion is Office Live is much more likely to be accretive rather than diminutive or diminishing, in part because you want to have Microsoft Office use Office Live, and even if we integrate from a pricing perspective Office Live and Office some day, I see that as a much more gradual transition than some sort of major blow in the revenue. But the truth is I'd love to have all of our productivity customers having an ongoing relationship where we provide value and they provide money back to us, that would be a great thing. That would be a good long-term position to be in. And if there's some short-term pain associated with that, we would take it.

KEVIN TURNER: I think it's also important to re-emphasize what Ray talked about was, it really helps with some of the piracy and some of the other stuff as well, so there's a whole segment there that we're not capturing today that that platform allows us to get into.

COLLEEN HEALY: Thank you. Number three, please.

QUAN TRANH: Quan Tranh from Morning Star.

As you build and maintain large datacenters to support your services initiative, services platform, can you talk about what that does to your cap-ex spending? Over the long term does Microsoft become a more capital-intensive company?

CHRIS LIDDELL: You want me to take that? Yeah, we talked on the call last week about our cap-ex going forward. We did increase our cap-ex from fiscal year '05 to '06, and we expect to increase it again in fiscal year '07. A large part of that is driven around the datacenter expansion. So you can expect us to continue to spend there. If we're successful in the marketplace, which we tend to be, then we're going to feed that with the cap-ex.

So it will impact capital spending. There is also an additional capital spending associated with other facilities in fiscal year '07. That will probably drop off from where it is. But it is an area where we're going to spend.

In the overall scheme of it, though, of Microsoft I'm talking of hundreds of millions of dollars, so it is going to be significant on a proportional basis. in terms of absolute dollars it won’t be significant. In terms of how it impacts our accounting, generally speaking, on the servers we're looking at a three-year depreciation rate, so that will impact our income relatively quickly. On the buildings it's 15 years, so it will be a relatively modest accounting impact.

COLLEEN HEALY: Thank you. Number one, please.

QUESTION: I'm just wondering if you could talk about the threats and opportunities represented by virtualization.

CRAIG MUNDIE: I think Bob Muglia in his comments talked about the fact that using these different types of virtualization technologies, whether it's the virtual machine capability in hardware/software, we do see those things as playing a long-term important role, particularly as we move to some of these more advanced microprocessor architectures. And I think that will be true both in the server environment and the desktop.

Today, most of the use of those technologies has been targeted at server consolidation in the management space, and I think we intend to be fully competitive in the near term in those areas.

I think as it begins to move beyond that to the client and then in support of the advanced architectures, again we're investing in order to get technology we think we need, and while there certainly are other companies out there similarly investing, the fundamental underlying hardware platform has been established in a uniform way within the microprocessor chipset, and therefore anybody can build upon it. So I think we think it's as much an opportunity as a threat at this point.

STEVE BALLMER: There's really three competitors, us plus two competitors I'd probably highlight for your radar screen: EMC, where EMC is kind of a high -- they think high price, lower volume, so in a sense I think we'll play in a different way than they play. The other competitor we're actually working hard to also sort of cooperate with and have the right interoperability, and that would be XenSource out of the open source world. And I think as long as we do the right job here, not too -- in the not-too-distant future of getting virtualization well integrated in with Windows, do the right stuff with virtualization in Windows on the client, and as Craig talked about, our next Virtual Server product, we should be in pretty good shape there.

CRAIG MUNDIE: I'll follow what Steve said. It is interesting when you look at the thing we did with XenSource, it wasn't about putting our stuff on their of the traditional virtualization environment, it's to build shims so that the Linux stuff would be hosted on our world, and I think that's just an indication of where we intend to go.

COLLEEN HEALY: Question number two, please.

QUESTION: Thanks. A question for Steve on the online services market. If you look at the -- you made the comment about best-in-class services that you're going to be going after. How long do you give yourself a runway to kind of close that gap and continue to make that incremental spend? And if the gap doesn't sufficiently close that you can kind of first or second in a reasonable way, what other options do you consider in terms of maybe adjusting strategy to go down an acquisition path or try to find ways to get scale to compete with some of the online properties?

STEVE BALLMER: Well, I mean, the truth is I don't -- plan B is plan A, which is to just keep pumping and working and coming and coming and coming and coming. There's no acquisition path out. Look, take search; you know, how many guys are there? There's nobody who's got better relevance than we do, except the market leader, and I don't really think that's the acquisition you had in mind. (Laughter.) And I'm not sure that would create any kind of shareholder value for anybody. So, you know, and we've got -- we'll just keep pumping at it. And we've had to do this in some other businesses, we think we've got clever enough people, we've got the ability to embrace some things they can't embrace easily, because it's inconsistent with where they are from a business approach.

Take e-mail; the truth of the matter is there are ways to improve search overall, e-mail overall; none of these experiences is perfect, and there's nobody I look around and say they have better, they just have different e-mail experiences.

This is not a place where the hundred-thousand dollar guy is going to solve the issue; this is a place where we need to have some R&D investment, coupled with some cap-ex done at scale. And it may not be wise for me -- I'm not going to say we're not going to get smarter every day, we're going to get smarter every day, but the notion that there's either a path to fold our hands or just say we can't do it, let's go plan B, buy somebody to get this problem solved, those don't exist. There is no fold-the-hand strategy, and there is no let's-just-acquire-our-way-out-of-this. I don't actually think either one of those works. Folding our hand is folding our hand on one of the major new opportunities around. It's inconsistent with long-term growth. We will continue to invest. And in the grand scheme of things, if that's, you know, whatever it is, we don't really invest in the grand scheme of things that much, even in our online business. It runs close to break-even, as Chris had a chance to talk about. I keep investment in that for a long time to get it right and really build a business that can make billions of dollars. I wouldn't fold it quickly at all. I wouldn't even dream about folding it anytime, anytime, in any kind of sort of foreseeable or imaginable future.

Now, okay, this is part of my theory: pick good businesses, invest in them, and do well. We picked, we're working hard, and darn it, we'll work hard till we do well. And whether it's me or the guy who has to replace me because we're not doing well enough -- (laughter) -- you know, we will do well, darn it, we will do well, because we don't give ourselves the option of not doing well.

And you could say that sounds goofy; I wrote down some of the things where I think we're differentiated. We are smarter. We're building every day. We're looking for outside ideas to help enhance in acquisition. We've brought in some amazing talent from other companies like IBM, like Gary Flake, who joined us from Yahoo! There are great folks coming in. We're moving some of our own internal talent into the area. We've taken a few of the key guys in our core Windows group to move them into the infrastructure. Obviously Ray Ozzie has come on board, and Ray is spending a disproportionate percentage of his time in this area. We moved one of our R&D leaders, one of our top guys, Steven Sinofsky over to really take a look at the Live user experience in addition to Windows.

What's the old expression from someplace, failure is not an option. Failure is not an option. The players may change, but failure is not an option. And I don't think it should be an option for our investors.

COLLEEN HEALY: Thank you. Paddle number three.

QUESTION: There's a sense out there, sort of a vague sense that all applications -- this may be directed to Craig or Steve -- are going to be delivered over the Web, and you're not going to need to ever install any software on your PC. And I'm just kind of curious as to whether A) you think that's ever going to happen, and if so, when is it going to happen? And if it is going to happen, how does your business evolve to that?

CRAIG MUNDIE: I think you have to separate that into two separate questions. One is, is the software delivered through a service but it actually is being installed on your PC, but is the model of installation different than the historical one of, you know, insert CD-ROM, wait 20 minutes, and see what you get. We clearly are moving in the direction where we can deliver software even if its value is to run locally over the network as opposed to only through the traditional means. There are other people who say, no, no, there isn't really anything of interest that runs on your computer. All that that thing is a portal through which you can observe an application that's running somewhere else. And what we've been trying to do is create all the mechanisms in Microsoft so that the consumer, or Microsoft, can essentially slide the slider from completely local execution to completely remote execution.

And as a function of the business model, the problem you're trying to solve, and the relative computational requirements, and data storage requirements, you can set that slider in exactly the right place. And so, when Ray talks and he says, you know what, for us it's about the ability to have applications, servers and services, or clients, servers, and services, it really is about having the ability to have that completely smooth capability.

In the enterprise, you know, years ago when we started the NT strategy, we said, we're going to make the architecture the same between the client and the server, and that turned out to pay big dividends over time. In a way, with Live, what we're trying to do is say, every one of our historical products wants to have a service component. That thing could range from almost de minimis, part of buying or installing it, could be part of the maintenance things where we're adding OneCare and other stuff, or it could be Office Live for a small business where they really choose to buy the thing as a hosted service, and have it delivered to their office. And I think we're well prepared now to be able to operate across that spectrum.

COLLEEN HEALY: Question here, panel member two.

BILL WYMAN: Thank you. Bill Wyman, ISI. Can you talk about the software and product development process from both a business and a technology point of view? You said we're not going to have five-year product cycles anymore, but you've got 50 million lines of code, not an easy process, and it kind of undermines confidence in the ability to deliver against all the things you talked about. Can you share with us some of the things you're doing on a practical basis to address that going forward?

STEVE BALLMER: Yes. I would like to opine. Let me get after this. If you take a look and analyze sort of where we shipped XP, we're going to ship Vista, and what happened in-between, there was one thing that cost us time in 20/20 hindsight that was probably wrong. We can manage that code base. We can extend it. We can add to it. It's not as modular as we would like it to be, but we did a lot of work, actually, during the Vista cycle to make it more componentized, more modular, that's all great stuff. But we made an upfront decision that was, I'll say, incredibly strategic, and brilliant and wise, and was not implementable. And I'll take full responsibility for the decision, because it was a decision that was made by me, by Bill, by Craig, by Jim Allchin and the senior guys in the company. We wanted to, and there were very good strategic reasons, to try to create a whole new, let's say, set of services in the operating system that would present a whole new paradigm for applications to do that at once and to have the things integrated together. We tried to incubate too many new innovations and integrate them simultaneously, as opposed to letting them bake, and then integrating them, which is essentially where we wound up.

That approach had worked before, it's not like the approach had never worked before, but it can't work now just given where everything is. There's too much complexity that that implies. It doesn't mean that we can't do major releases. So, we worked down that path for a while, and then we said, no, this isn't going to work, and we essentially, I'll say, rebooted where we were. By that time it was clear security was a priority, and we did an amazing job, if you want to talk about an engineering tour de force, the one year we spent doing Windows XP SP2, it's amazing that we were able to touch and change and modify that large a piece of software, and make that many improvements in one year. Not a marketing tour de force, Windows XP SP2, we picked it for a reason, we were trying to reassure customers, but we probably did more positive change to the code base in that 12 months than at any other time in the company's history, and we took that opportunity to really do some work that would increase our overall engineering agility in Windows, led by a guy named Amitabh Srivastava, who works here, and Dave Cutler, some of our core OS architects, really did a phenomenal job.

We got XP SP2 shipped, and now we're going to have a two-and-a-half year cycle for Vista, and get a very, very exciting, impressive release, and that's about right for an OS. I mean, the truth of the matter is, customers don't really want OS releases every year, and they don't really want OS releases every five-and-a-half years. So, I don't think there is anything fundamental, other than we learned our lesson, and we commit ourselves to a path. And, frankly, for a variety of reasons, with Jim Allchin retiring, we chose to move this, as I mentioned earlier, Steven Sinofsky over to the business. Steven had an unparalleled track record of regular shipments with our Office code base, which is also very large. And so, I think I know I'm very confident that we have learned the one key lesson we needed to learn from that time frame. We still have security issues that we are going to have to work on for the rest of time because the bad guys aren't going away, but on the other hand we've sort of stemmed the tide in terms of really making significant improvements on our core security approach.

CRAIG MUNDIE: I'll just add a couple of thoughts. In doing this Windows XP SP2, we actually developed a completely new engineering methodology we call the software design life cycle. And we apply that across all the companies. And, in fact, Windows Vista is the first product to really go through that cycle in its entirety. And then, ultimately, I think relative to these issues of testing and security-related matters, that will ultimately lower our total time and cost to do that. Steve also alluded to the initial work that we've been doing in componentization, which allows us to build derivative products, or to make changes in a more controlled way, and while there's still more work to do, that also puts us on a good trajectory.

And then finally, although it's a bit longer-term, there will be, I think, some fundamental changes, more in probably the five-to-15-year time horizon, where we will bring a new level of formal composition to the methods by which we construct, or anybody constructs, very large-scale software systems. And a lot of our research work in how to build software systems and test them, is now tilting in the direction of investigating those approaches. If that really comes to bear, and I think it will, then many of these questions of just thinking that complexity grows either linearly or exponentially with the code base size will essentially be dealt with.

If that was true in other disciplines, you'd have a tough time building skyscrapers and bridges and other things, because they get bigger, but we still manage to get them to be built and stand up. And I think software as a whole industry is really only getting to that stage that we've reached in other engineering disciplines where we have this formal composition and division of labor that gives us that kind of scalability but maintaining reliability and predictability.

COLLEEN HEALY: Thank you. Next question, number one.

QUESTION: My question has to do with Vista and piracy. You mentioned a little bit earlier today about genuine Windows, but with Vista you're going to be adding in Defender and updates for IE 7 clearly designed to reduce piracy. Will that be enough? Do you think that will have effect out there, and is there anything in our guidance with an assumption that piracy will be diminished in the numbers that you gave?

CHRIS LIDDELL: There is some small element in our guidance associated with piracy, but it's a continuation or an extrapolation of the sort of improvements we've seen in the last couple of years. So, we're not looking for a step change in the next year. We're seeing some benefit from our anti-piracy efforts that's helping us in terms of the differential between PC growth rate and our revenue rate that is built in, but more our linear extrapolation that we would be coming from.

COLLEEN HEALY: We likely have time for one or two more questions. The next question from number two, please.

QUESTION: Hi. Just a couple of quick questions. The European regulators seem to be making a crusade out of picking your pocket and creating challenges for you. To the extent that you can address the topic, I was hoping that you could talk about some of the actions you're taking to try to resolve the issue, or if you think the issue can never be resolved, in light of the positions that they've taken. That was the first question.

STEVE BALLMER: Well, our job obviously is twofold. Number one is to comply with the orders that we've received from the European Commission. We've worked very hard on compliance, we've both shipped a version of Windows without Media Player, and wrote thousands of pages of documentation, albeit, an effort that was not viewed as sufficient. So now we're writing thousands of more pages of documentation. But, job one is compliance. And despite the recent, let me call it, setback, we remain committed to complying with the orders that we received from proper government authority.

Number two, our job is to appeal. We have gone through the appellate process, we have argued and presented our evidence, we've argued our case in front of the Court of First Instance, and we will get a ruling. The Court of First Instance will give us the ruling when it's ready, that's the way appellate courts work everywhere in the world. And I hope when we get guidance back from the Court of First Instance it will do a lot to help tune up the situation, shall we say, overall.

I also hope that, in a sense, between all the learning both we and the commission have gotten from this first phase of compliance, that we can now put that together in the rearview mirror, because there are still questions of compliance in front of us. Obviously, we want to ship a compliant Vista, and right now we understand what that means in the U.S., and under the consent decree there's work to be done to understand that under the directives of the first order from the European Commission. So we'll continue to work on compliance in the context of Vista, and I'm putting a big premium on compliance, it's an area where not only I but of course Brad Smith our general counsel and others spend quite a bit of time.

COLLEEN HEALY: Great.

STEVE BALLMER: I'll just add just let me. Nobody likes one-time legal charges, not you, and certainly not me.

QUESTION: A second question, which you started to address, is it's very difficult for large companies to sustain innovation. Maybe you could share with us some of the insights as to how you prioritize the hundreds of thousands of ideas that people have how to improve the products, so that we can get some confidence in the things that you choose have that five-fold of the cost of capital returns that you talked about. Can you give us some more visceral sense to how the process works to surfacing and developing the best ideas?

STEVE BALLMER: You sort of asked two different questions, unless I mistook you. One is, how do we make sure we nurture innovation, and two is, how do we make sure we get a return on innovation. Right?

QUESTION: Yes, I did ask both those together.

STEVE BALLMER: In a sense, partly what you step up to, if you want to nurture innovation, you're going to nurture some things that may actually never pay back all that well. Hopefully, you catch the ones that aren't going to pay back all that well before you spend too much money on them. We're always nurturing one of the areas we've been willing to grow is our investment in research. And our research team lives under the context that the stuff they work on doesn't necessarily have to work. It's not a shame if it doesn't work. We're going to invest in it. At some point if they want to give up on it and try something else, that's OK, because we want to be more speculative in some of the things that we do, and that's part of what it means to nurture innovation. That's not $5 billion a year, but we do sustain in research alone investment of, what, about $200 million, something like that, which is certainly significant.

So we need to nurture innovation in a lot of forms, so that in the areas where it is important, we've got the technology to get after it. I'm very glad, for example, that there are folks in Microsoft Research that have been working in a lot of relevant areas to search, essentially, who can help us with that campaign.

I don't think the big issue is the big issue is not can the small innovations have a good ROI. The big issue is not are there too many small innovations. The big issue is, do we elect to make serious investment, whatever that means, R&D or sales and marketing, in the areas that become big businesses, and we do it soon enough.

I wish we had had the capacity to do Zune a year earlier, but frankly, we took some of the top folks out of the Xbox 360 effort, and right after 360 shipped they became the Zune team. Otherwise, we would have done it earlier, I suspect. We just didn't have the right talent capacity at the time.

So in terms of ROI, like I said, if we do a good job in online, there's going to be a good ROI, but that's one of these big-bang, serious, forms of investment. In healthcare, we've put in some money. We're after it. We're figuring out where the big opportunities lie. We passed on dozens of ideas in the healthcare area before we made the decision to go ahead and buy Azyxxi. Craig's got a whole vision for how professional healthcare providers meet consumer healthcare. We shared it with a bunch of CEOs, Craig did at the CEO summit. It's pretty jazzy, but we're not ready to put the pedal to the metal now and say it's time for big investment yet.

We're feeling our way forward. But, we are making some investments in the right forms of innovation, so that as we see the path, we see the business model more clearly, we can make those serious investments that require a serious discussion of ROI.

CRAIG MUNDIE: Just to add one thought, it really is both a bottom-up and a top-down process, and the thing that makes it work is people. Sometimes you decide you want to have a strategy, like we said, OK we're going to make a strategy for health, and then you go get a lot of smart people involved in trying to execute against it.

As you indicated, you've got a ton of smart people, they're going to come up with good ideas. The ability to bring those things forward, do triage on them and figure out which ones really have merit requires both a serious and ongoing process of the middle-management of the company, and then a mechanism for getting the top management to pay attention when one of these things pops out. That was something I think Bill always helped the company do, and we've now been systematically getting other people to be able to do that, too.

STEVE BALLMER: Just one other thing, if we were fund next year an additional 50 innovations at $5 million each, it would be a footnote in this discussion. We had $2.7 billion, roughly, of op-ex, and I'm not saying we have 50 more that aren't funded this year, but I'm saying it's not the funding of the small innovations that people come up with that wind up being at the heart of the ROI discussion that we have. And I think it's wrong-minded for us not to put some of the seed money around. So we're always trying to plant enough new seeds, nurture enough little saplings, but the big discussions come down not on the saplings, it's when something looks like it's going to be a tree, and we bring the big watering can, and the big hoses, that's usually when I think we wind up standing in front of you and having to talk about things.

QUESTION: (Off mike.)

STEVE BALLMER: That's right. It's the opportunity cost of the resources, I agree with that. Point of fact, some people actually have an opportunity cost, some don't. They don't work on any no, they have an area they're going to work on, and if they don't work on it here, they'll go work on it someplace else. I don't mean they're bad people, I just mean they have a passion and an energy in a certain area. Some of that there is an opportunity cost, and that's one reason I guess we both go we all go through a very careful strategy planning exercise. Chris has us focus not just on EVA as part of that exercise, in addition to all the other things which we naturally consider. And I'm glad we have the hiring capacity, frankly, that we do, but there are still choices, like the choice we made to finish Xbox 360 before we started Zune.

COLLEEN HEALY: Great. Thank you so much for your attention and for your questions. Thanks to Steve, Craig, Chris and Kevin as well.

If you're one of our winners, see Michelle at the registration desk to claim your prizes. The buses will leave at 7:15 for the airport and the hotels. And please join us in the McKinley Room for cocktails and the reception. Thanks so much for joining us. (Applause.)

END

Due to the varying sound quality and subject matter of tapes, the information in this transcript may contain inaccuracies.