COLLEEN HEALY: Great. As the chairs are coming out, we'll run this process as we did our first one, which is we'll entertain a question from the CommNet, and then the Investor Relations team is working their way around the room with paddles and we'll go from there.

So the first question I'm going to take was submitted on the CommNet is around Home Server, Windows Home Server. And the question is, where does it fit in and how big could it be?

STEVE BALLMER: Well, I think Windows Home Server is an interesting product, and frankly I think has a reasonable amount of potential. And yet with the interest in software as a service and the cloud, I think we're going to have equal balance between how much for the local storage they want to have, how much they want to spend on that, how much they want to store and keep things in the cloud, what their kind of interests and concerns are about spacing and backup and the like. So I have a pretty positive outlook -- a very positive outlook for Home Server, and yet I think it's too early to give any kind of wild-eyed projections.

COLLEEN HEALY: Great, thanks. Let's go with mic two, please.

QUESTION: I guess Chris, or maybe Steve, can you maybe elucidate a little bit on your capital structure and how it relates to your acquisition strategy? In the past you've talked about carrying sizable cash balances to give you strategic flexibility. Then along comes something like aQuantiv. Do we view that or do you view that as a call on the balance sheet so you're comfortable with lower cash balances? Or is this something that we should think of as impacting the available cash flow you're going to have for next year for repurchases?

CHRIS LIDDELL: Yes, I'm happy to comment. When we look at the amount of cash that we believe we should have on board, there's a number of factors that we take into account, acquisitions being one of them. The inherent business risks that we have, some of the legal and other risks that we have, and some of the other balance sheet claims that we might have going forward. So there's a mixture of a number of factors that we have, and that's a dynamic situation -- it changes on a year-by-year basis.

Clearly with what I talked about was the fact that we are doing more acquisitions -- not only smaller ones, but medium ones. That's one of the things that we're going to build in from a cash flow point of view. We don't clearly keep cash around for an extremely large acquisition. Whenever one of those ever happens, we worry about that at the time. But, yeah, acquisitions are important.

The one thing I would say is that if you take aQuantiv aside, we've done $1 billion to $2 billion worth of acquisitions over the last two years. That's relatively modest in terms of the amount of cash that we have on board. So it's an important factor, but it's not one that makes a material difference.

COLLEEN HEALY: Great, thank you. Mic two.

QUESTION: Hi. I had a question related to Windows attach rates to enterprise agreements. That's something that you mentioned on the last conference call you started to see an uptick in. I was just wondering if you could talk a little bit about how we should expect that to unfold, how much room do you think there is for that to move higher, and also when should we start to see that play out given the higher AFPs they're getting with the enterprise edition of Vista, so that we can get to that point, Chris, that you mentioned before, where revenues would be growing faster than PC units? Thank you.

CHRIS LIDDELL: Well --

STEVE BALLMER: I have something I want to say, and then I'll let you really answer the question if that's OK.

CHRIS LIDDELL: Yeah, sure. (Laughter.)

STEVE BALLMER: Revenue growing faster than PC units. We have a nice upside, and Chris will talk about the upside in the enterprise edition and maintenance and the like. The numbers are really quite stark in terms of the rate of growth of the PC market, the so-called OEM business that we have, which is the bulk of our business. The bulk of our Windows business is not in aftermarket; it's in the core OEM business. And the rate of growth of new PCs in markets where we either have lower prices and/or higher piracy is really quite dramatic versus developed markets, and not likely to change in the next few years. So I wouldn't correlate the two events all that well. I think we will continue to see relative to PC growth challenges in Windows revenue growth -- not because we're not doing a great job but because things are coming, if you will, in markets where we have to do different kinds of things to get penetration.

That doesn't answer your basic question which I'll let Chris answer about attach rate, but I just -- I don't like the bundling of the two concepts that much.

CHRIS LIDDELL: Yeah, and as you know each quarter we disaggregate exactly those factors. I'll just go through the math very quickly. Eighty percent of the revenue in the client business roughly comes through the OEM channel; 20 percent through the commercial and retail. In the 80 percent through the OEM, we get a slight uplift in the units that we sell relative to the market overall because of piracy -- gains on piracy. It's been averaging about 1 percent a year for the last couple of years. But then, because of the growth in the consumer market relative to the business, because of the growth in the emerging market relative to the mature markets, the rate of revenue growth is lower than the overall rate of unit growth. It's just the natural math of the fact. And Steve went through that in February to some extent. And we've talked about that, I know, to a number of you. If you weight-average the ratio, which is the way that I think about it, it normally takes 2 to 3 percent off the revenue, the inherent revenue relative to the unit growth. That's just the starting math. From there you have the opportunity to upsell into premium.

To answer your specific question, it comes under 20 percent of the revenue that's in the commercial and retail side. Remember half of that's retail -- there was a big retail spike this year, so that's going to be hard to repeat year after year. So it's the smaller element of that.

In that area I think there's a very good opportunity to grow, but just keep the order of magnitude growing -- some proportion of the 20 percent and getting some upsale will be a nice plus. It will help offset some of it, but it's not going to drive a big growth rate on the 100 percent of the revenue. Sorry, that's a long answer, but I know I go through it each quarter math-wise.

COLLEEN HEALY: Mic one, please.

QUESTION: Hi, it's Adam Holt from JP Morgan.

I had two questions on operating margins or margins generally. And, Chris, thank you for the incremental detail on the structure there. And I wanted to actually drill down on one of the slides that you put up where you have core margins that were sort of 61, 62 percent for the last several years. And you suggested, I guess collectively, that we should expect to see limited margin expansion in the core. And I guess I was wondering, you know, after a lot of investments around getting the big Office and Windows releases out, why wouldn't we expect to see a little bit more margin expansion in the core? And I guess my second question is maybe for Steve -- this came up at lunch in our conversation a little bit -- but as we think about the medium term -- maybe the next three to five years -- understanding that you obviously want to invest in growth in some of the new areas, our dilute of two margins, how should we be thinking about the ability to take margins up on a going-forward basis over, say, the next three to five years?

CHRIS LIDDELL: I'll answer and then hand it over.

I mean, from the margin point of view, the most interesting for me is to get the unprofitable businesses to be profitable. As I said, for Xbox that's a challenge for next year. For our OSB business that's going to be a longer-term challenge, but the amount of value to be created there is huge. So I'm willing to say that's the right thing to do.

In terms of our core businesses, those are fabulous margins. Could we improve them? Possibly a point here, a point there. But at the end of the day we are going to be reinvesting for the next wave of innovation, and that's exactly what I talked about with the long curve. So inside those businesses we're not only thinking about the next release of Vista, the next release of Windows and what we're doing there; we're thinking about the trends of five to 10 years out, and we're willing to make investment decisions that we think create value over that time, even if it doesn't give us the short-term pleasure of allowing the margins to expand even higher in those businesses.

BILL GATES: And we may get some margin expansion. But, you know, in a sense if you take the last few years as a guidepost, we've invested a lot over the last few years not just in Windows and Office; we invested a lot over the last few years in building up a real enterprise, telephony and unified communications business. We've invested a lot in our business intelligence product line and product lineup. And I think those investments will now have a chance to generate revenue.

But I want to make sure we're making the next two, three, four, five, six investments in the next wave of net revenue growth opportunity.

If you go to the framework Kevin Turner talked about -- sockets and revenue per socket -- I don't want our growth and future to be restricted by what we sometimes refer to as socket growth. I want all the socket growth and I want -- we all want value growth against the sockets. And value growth against sockets only comes if we invest in new categories -- security, management, business intelligence, etc. And so you ought to expect to continue to see us make that kind of investment.

COLLEEN HEALY: Mic two.

QUESTION: Michael Cohen from Pacific American Securities. My question is for Bill. This has been an industry of periodic paradigm shifts -- client server, Internet -- that's opened up vast new areas of business opportunity, and it's been sort of boring for a while. (Laughter.) And I was wondering if you'd give your thought for what do you think the next area that would really open business expansion?

BILL GATES: Well, I'm not sure it's been that boring. (Laughter.)

The big paradigm shift that's taking place now is the cloud coming in as a peer -- as an equivalent, the important place for storage and computation as servers and clients. And that's a very dramatic thing. The ability to take cost out of IT to let people deploy services more rapidly to essentially be up-to-date using the latest, and have a full suite of software is way better when you're using the ability to have our centralized IQ looking at the operational license and having that be far more automated. So I'd say that's, you know, which we'd call software plus services -- that's the big opportunity. It's big for the business computing market -- that's always been about 80 percent of the computer industry or more -- and it's significant for the consumer side as well.

The thing that supercharges that a little bit is the idea of natural interface and device variety, where because we're user-centric, because we're always sending your state up to the cloud, your willingness to move from device to device is much higher in this environment than it's been where you have to kind of manually think, OK, let's move this file between this and this. Which program is there? Which patch is on this thing -- and making the user go through a lot of effort.

So that for the next five, six years is the big thing that's playing out, and all the players will get measured against how they adopt that model. No one is really delivering that model today. And we have things like Windows Terminal Server, Office Live, various things that point in that direction. Some of the non-big players have things that kind of point in that direction. But nobody is actually doing that. So this is not maybe like the browser paradigm or Internet paradigm where people thought, OK, there's - the company is born on it or actually doing that ahead. That is the big one. I don't see any other ones like that. It's kind of like during the Internet phase where we said that we were so focused on the Internet that if there was something else that came along that was a paradigm shift, we were certainly going to miss it because we were focused on what we needed to do to be the leading Internet company on both the development server platform and the client developer platform.

COLLEEN HEALY: Mic two.

QUESTION: Sid Parakh, McAdams Wright Ragen.

Just early on we saw a slide with a lot of new products that are scheduled to be launched next year. Can you help us understand what sort of revenue contribution should we expect from these new product launches - not just in the next year, but over the next few years? And how does that impact sort of the longer-term revenue growth rate for the company?

STEVE BALLMER: Well, we very consciously and conscientiously don't give out long-term growth rates, and I won't start now. So it was a good way to get at the same question.

The other thing though that's sort of always complicated is, you know, people want to say what percentage of your revenue comes from new products. And in a way the answer is always 100 percent, because our old products never wear out. So you could say Office 2007, new versions, and yet in a sense of course maybe some people would have bought the old Office if we hadn't had the new Office available. Some of those products are brand-new in the sense that we didn't have anything in the space before, and many of them are upgrades, line extensions, enterprise editions, things that are designed to drive more value. I think the guidance Chris has given for fiscal year '08 kind of stands for itself. It speaks for itself. And we won't give guidance that goes beyond that. But I think you can take our comments about long-term approach, investing in opportunity -- and you know it's all about investing in growth, and whether that's foo or that's bar, a reason why we're doing all of the things we're doing -- not the only reason -- I'd say we start off at times internally with the excitement of the area and the innovation and what we can bring.

But the business folks, including Chris, and me, and Bill, and many others, and the questions we're asking ourselves are: Can we drive revenue and profit growth? If we didn't think we could drive any revenue and profit growth, whether it's out of something like Xbox or what we're doing in search or Surface -- Surface is an interesting case study. It's a really interesting case study, because frankly it had a lousy business case -- lousy business case. Bill though was giving it kind of mouth-to-mouth. The team was wondering what to do. It's one of these things I could look at and say, Geez, with that business case what do you do? And then we announced the things.

And now people, as I said, people are kind of falling all over themselves to say what a fantastic in the customer community product. And actually what the shape of its growth, its profitability, its contribution is, it's very, very hard to predict. I think we're all extremely happy that we let the team that was incubating it continue it and bring it to market. And now we'll get a chance not only to see what the business case delivers, but we have a bunch of new business cases that essentially popped because of the positive customer response.

CHRIS LIDDELL: I think the other thing is the role the new products have in terms of the value proposition to our enterprise customers. And one of the things that we talked about on the fourth-quarter call, but didn't really get a lot of airplay last week, was the strength of our enterprise agreement renewals and the strength of our unearned, which was just fabulous. If you had to take a highlight from me of the fourth-quarter result last week it was the unearned, not the actual underlying results in fiscal year '07. And part of that is just the proposition that customers see in terms of the new products coming down the pipeline. It's impossible to disaggregate them because they happen in the future, but people are certainly renewing at a rate which is higher than our historic -- and creating new enterprise agreements at a great rate in anticipation of what people see. So I can never take a dollar and split it between the individual products, but in terms of what the field has to sell as a concept is an extremely important part.

COLLEEN HEALY: Mike one, please.

QUESTION: Hi, it's Kirk Materne from Banc of America. The question is around piracy, and just a two-part question. First, as you see companies in places like China and India trying to access capital and become, let's say, more respectable corporate citizens about licensing, I guess, how much of the growth in those areas do you see benefiting some of your revenue recapture around piracy? As those economies grow, do you see the trends there at least becoming better on their own?

And then, secondly, around your efforts on Genuine Windows, when do you think it's too early to start seeing if that's having an impact or not? And, Chris, just to remind us, was there any kind of positive impact in the client number for fiscal '08?

STEVE BALLMER: Do you want to start? Or I'll start with the other question.

I think everything that leads to what I would call rapid and more organized economic development in emerging markets is a good thing. I mean, as much as people like to talk about various aspects of Sarbanes-Oxley, any company in one of these emerging markets that wants to list, and list in a country where you have to comply with Sarbanes-Oxley -- the United States -- that's not a bad thing for us, because having an undisclosed liability for software is a very good thing. Let me just say it that way from our perspective. So you have some of that.

You've got the general trend in some of these countries to what I would call more organized economy. India is a very good example. India's retail infrastructure has been very fragmented -- small mom-and-pop. And if you look at what a number of the big companies are doing to try to build big, sophisticated, organized retail which depends on proper IT support, that's all good for legal -- it's not all about sort of small stuff. It's all very organized mission-critical computing, and it plays very directly into the kinds of things that we do. So I see a lot of good things -- government action, government involvement, government concern is still the number one thing. Though, governments that decide that they want to export intellectual property, want to protect intellectual property, that's an important phenomenon.

In terms of Genuine Windows, what we gave ourselves was some new tools to use to try to find the right balance point between piracy, customer convenience, customer acceptance, satisfaction. And I would say we didn't see any measurable results in the quarter that we just talked about, but we're now in markets getting experience and giving ourselves the tools to explore whether there are points in which we put up a little bit more of a block if you will to pirated use, see acceptable customer satisfaction and superior revenue performance.

CHRIS LIDDELL: Yeah, from a client point of view it was about a 1 percent difference. So, the PC market grew last year 10 to 12, we grew units at 11 to 13, but as I mentioned to Heather earlier, that's a unit number. So, we grew units faster than the market, mainly because of the anti-piracy work that we had, but bear in mind there are relatively low revenue units that we're talking about.

COLLEEN HEALY: Thank you. Mike two.

QUESTION: Good afternoon. Sarah Friar from Goldman Sachs.

Guys, you've proven that the aQuantive acquisition that you're definitely willing to get a bit more aggressive on inorganic growth, and yet I think most people in the room would still see Microsoft as an organic grower at heart. What would drive you to really think about a bigger acquisition? And could you give us a framework for how you make that build-versus-buy decision?

STEVE BALLMER: Yeah, aQuantive, I don't know how you want to think about it. I think of aQuantive as part of an organic growth, not an inorganic growth strategy. Not that we didn't spend a lot of money to buy something, but the value -- take the ad agency aside, which has a distinct value associated with it -- but the value in aQuantive and the thing we paid for and the thing that will either pay back or won't pay back, the way we thought is we bought ourselves the capability to do some other things. It wasn't per se the fact that the aQuantive advertising technology business was undervalued or something. We see an opportunity to blend it in, morph actually its strategy some, and morph our strategy some to do something bigger.

So, I don't know whether you'd call that organic or inorganic, but it's not inorganic growth the way I think banks, for example, would talk about expanding and acquiring other retail banks.

And as we think about this thing, I think in general our growth strategy is a growth strategy that is predicated on organic growth, organic development and synergy, like the aQuantive-style synergy.

Are there some big things out there that we could conceivably buy? Sure, there are some big things out there that we could conceivably buy. Anything that we would buy would have to meet the following criteria. Number one, we can't pay more than it's worth, and what it's worth is what it's worth on its own, and whatever we can get extra because of the way we morph it. I don't think you should expect us to buy things because we think we can squeeze out 10 percent of the cost structure. I got more feedback on breaks about how you'd like to see us squeeze out on our own 10 percent of our cost structure than I got sort of kudos on that, and yet we will -- and we won't overpay for things. We don't believe it makes sense to overpay for things, particularly big things. You can say that's a sign of discipline or a sign of the fact that we're basically organically minded guys, but nonetheless that's how I would characterize our kind of philosophy and view.

CHRIS LIDDELL: I guess there are two things I'd say from a financial perspective that we use. On every acquisition we do we look at a build-versus-buy analysis, not just a large one like aQuantive; if we spend $10 million we still look at a build-versus-buy. And we have a very disciplined approach about what's the amount that we would have to spend on a build and what would the market returns look like; if we buy, what does that look like, what do we have to spend and what does that look like, and what's the delta between those two figures, and that's how we determine how much we can actually pay for the acquisition. But we do a formal 10-year DCF model and look at the delta between a build versus buy, regardless of what we're looking at.

The second is the risk aspect of what we're talking about. So, even if you do that simple mathematical model, there are some areas where we think that over a two- to three-year period, the sort of time that it might take us for a build, there might only be, say, two or three winners in that particular space because of the network impact that we'd have. So, the risk inherent in a build in terms of the timeframe that's associated with doing it may be higher in some cases and it might swing us more towards a buy. But overall build is our fundamental -- that's our starting point.

COLLEEN HEALY: We have time for two last questions. We're going to go to mike three, and then we'll end on mike two. Mic three?

QUESTION: Yeah, hi, this is Taunya Sell from Ragen MacKenzie Wells Fargo.

And I had kind of a curious technology question. You know, we're seeing more cores on a chip, and Intel is talking about multithreading, and obviously we have the example of a cell phone where you have an application processor and kind of the DSP. But I'm just curious what a big software company like yourself thoughts are on this, and kind of where that might start, and does that start in the server, how does that play out? I mean, it becomes very complex, and I'm just curious what your thoughts are there.

BILL GATES: The place that it's most obvious is in the server, because there you're dealing with a set of requests that are inherently parallelly processable, whereas on the client your ability to keep any more than, say, eight processors busy is a very difficult thing. You don't generally have something that breaks itself down into discrete pieces quite that much. There are exceptions to that, some of the natural interface things, some of the gaming-related things, some of the financial calculation things. But if you look at the percentage of time on what percentage of PCs you have those naturally parallel things, it's actually pretty small today.

So, the server is the place that it will come in. It's just like 64-bit in that sense, that you'll see it there, and then it will come down to others.

The main reason you're seeing multiple processors on things like phones and PCs has to do with power constraints, where you want to keep certain functions running in the background while you shut down other functions. And the early generation of these parallel chips don't really let you turn units on and off and get a certain profile, power profile, out of them. It may eventually get to that point, but the way the things are fabbed, that's a fairly difficult thing to do.

QUESTION: (Off mike).

BILL GATES: Well, it changes -- on the server it doesn't change things a lot, because we can capture the parallelism through the nature of the multiple requests that we're serving at once. On the client it changes things a lot. And the way that you want to write software is different. So, you have to take the nature of what the new things people want to be doing, and understand what those are, and then get those crafted in a way that can use the parallel processor. And that's partly why you see us collaborating with Intel in somewhat of a deeper way than we have in most of our history of working together, because they have to do the chip architecture and some of the memory architecture, and how the different units connect up with each other, together with us doing software tools and evangelizing those tools, and even taking some of our big applications that run on the client and being able to deliver those in a form that can use this high-processor-count-type architecture that is the natural evolution of all those transistors that they have out there.

So, it's a very interesting problem, as Craig alluded to, in terms of the client. It really forces invention and using techniques that have only been talked about essentially out of the supercomputer space in the past. And we even have some of the real wizards who did breakthrough work in that space who have joined our team, people like Burton Smith and others, helping us do this new architecture.

STEVE BALLMER: We probably have someplace north of 200 people in a variety of different places around the company who feel like their mission technologically is to help us make the transition to the so-called many-core architecture: compilers, operating systems, incubation, application incubation, supercomputing style projects and their application, research. It's actually a reasonable size investment. And people say, well, why would you do that? That's a long-term investment. If we're going to keep Windows healthy, that's a very important investment area for us as an example.

COLLEEN HEALY: Great. Mike two, please.

QUESTION: Thanks. Last year at this time we talked about services and would it be cannibalistic to the business, would it be additive to the business. As you get ready to roll out your platform, and we've had another year to think about it, I'd be curious how you think about that issue now. Certainly as you host things like Exchange, it could be they get revenue over time, but it would seem there are also a lot of opportunities to address piracy and new opportunities for growth. So, if you could just sort of flesh out that issue a little bit for us and share how you're thinking about it.

STEVE BALLMER: Yeah, I tried to talk a little bit about that in my comments earlier, so I'll start just by repeating sort of certain parts with a little bit extra emphasis, and then go from there.

On the consumer side our ability to monetize after market a PC with application -- I mean, we do better than anybody else selling software. We sell a reasonable amount of Office. You can buy Office for home, student-type use for just over a hundred bucks. And yet most of the Office you would find on home computers has been pirated. And so the notion of being able to monetize -- and monetize -- I'm not talking about Office now, but in general, advertising is a potential better monetization source in a market where piracy is high; I think we have net opportunity, particularly consumer and tiny businesses, home businesses, et cetera, to actually increase revenue on everything we do that's application-related, because of the move to software plus services.

Once you get past the smallest businesses, so other small businesses, medium businesses, large businesses, we actually have the opportunity to add more value. It's one thing to give somebody a piece of software, and there's some value associated with that, but if you say not only do I give you the software, I operate the software on your behalf, I help you take cost out of your infrastructure in kind of some of the ways that Ray Ozzie described, that's of value. And at least my ingoing assumption, and we'll have to get market feedback, we've got to do a number of pricing discussions and negotiations, but my general belief is we can save our customers money and improve our net profit profile, even net of whatever the costs are of operating the datacenters, et cetera, in this environment. And at least so far that looks probably right, but everything is kind of early days, if you will.

COLLEEN HEALY: Great. Thank you so much, Bill, Steve and Chris, and thanks to all of you for attending. It does end our formal presentations today, and we'll look forward to seeing you directly across the hall for a reception. Thanks again.

STEVE BALLMER: Thanks, everybody, for the time and patience; we appreciate it.

CHRIS LIDDELL: Thank you. (Applause.)

END

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