Keynote Interview with Steve Ballmer, CEO, Microsoft Corporation
Gartner Symposium/ITxpo 2005
October 19, 2005
ANNOUNCER: Ladies and gentlemen, please welcome Vice President and Gartner Fellow Tom Bittman. (Applause.)
THOMAS BITTMAN: Good morning, and welcome to Day Three of Symposium, and the second of our Mastermind Interviews. Please join me in welcoming my Dave Curley, Research Vice President. (Applause.)
Your IS organization depends on dozens, maybe hundreds, maybe thousands of vendors, and few of them are as pervasive or as critical to your organization as Microsoft. Please join me in welcoming the President and CEO of Microsoft, Steve Ballmer. (Applause.)
Steve, as we usually like to do, we've asked the audience what they would like us to ask you today, and let's show a few of those videos right now.
That's a nice start.
STEVE BALLMER: I've never, honestly, thrown a chair in my life.
MODERATOR: We nailed it down. We knew we had to get your blood boiling just to get started, so we thought that would be useful. Steve, let's start with one that's really at a high level. Talk about strategy, in the '80s and '90s market dominance was about API control, DOS and Windows gained that control, and gained the support of many application vendors, but that's changed. In the Internet world, it's just more about Web services, it's much more about Web presence, it's much more about interoperability than it is about API. So, in this changing world, how do you retain control, how do you retain strength? And what's your new strategy?
STEVE BALLMER: Well, I might characterize things a little differently. In a sense, many of the best businesses that people have built in information technology are actually about bringing somehow together massive communities of suppliers and massive communities of users. So, take Windows and the API set of Windows, we brought together essentially a massive community of people who built applications and peripherals with a massive community of people who wanted to use those in some sane and integrated fashion called end users, consumers, the almost a billion people in the world who use PCs.
In the Internet, you see eBay, for example, bringing buyers and sellers together through a platform. It has some protocols, and it has some APIs, and it's got blah, blah, blah, blah, but the fundamental economic value that eBay has created is in the way it's bringing buyers and sellers together.
If you look at what's going on in Internet advertising, be it with search or portal, be it with us or Yahoo and Google, it's actually about bringing a massive community of advertisers and content producers to be able to deal with a massive community of end users. But, I actually think the formula, I don't know about control, that's your word, not mine, but the formula for massive value creation in some senses is around creating technology that helps bring together massive numbers of suppliers of something important, and massive numbers of consumers of that same good.
MODERATOR: Let me build on this control issue, actually, because I think there is a control aspect to it. Whether it is benign control or what-have-you, with the Windows API fit, you controlled and drove the direction finding what consumers need and packaging that in IT and then providing it to developers to try and deliver solutions. If we look at this new world, can you maintain that same degree of control, and that same model, or does it need to change in this more Web-centric, multi-user, multi-vendor world?
STEVE BALLMER: We have a phenomenal opportunity as a company to create new value for our customers, new value at the desktop, new value at the server, and new value out on the Internet cloud. That value is going to come from both new applications and services that are not only created by Microsoft, but created by third parties. And so the question of how do you allow third parties to create interesting scenarios that partly live on the client, on Windows, on the server, Windows, and out in the cloud, I think is a very important one, and certainly one we're focused in on. And you can see a lot of it in what we're doing with MSN today as we open up new capabilities and make them available to developers.
MODERATOR: The shift from APIs and the platform being the box in front of me to more of a Web world in some sense is driving change in Microsoft. And one of the big questions that I think the market is trying to understand, and our users are trying to understand is, how do you stay agile in that kind of world? How do you maintain innovation and agility in your organization itself?
STEVE BALLMER: The top priority for us is to be an innovative company. At the end of the day, we make all our money based upon our ability to innovate. If we don't innovate, we don't have new versions, nobody needs to upgrade, nobody needs to buy, so you're asking me, in essence, what do we need to do to make sure that our job-one skill, innovation, is fulfilled. We actually have to have a variety of ways to innovate. We've got to have the ability to invest in things that will take a long time. We have to have an ability to invest in things that will take a few years, and we have to have the ability to invest in things that literally need to change and move and rev every six months or nine months, very frequently.
We've gone through a period where, shall we say, we haven't had all the muscles working evenly. Our MSN team has been pumping rapid innovation into the market. Office has been moving every few years. Windows has had a longer gap between its major releases, with the exception, of course, of the major security release we did, XP SP 2. And I think the important thing we're focused in on across Microsoft is how, through the combination of both products and services on top of those products, Internet-based services, all of our major businesses can have a short twitch capability, call that every six or nine months; a medium twitch capability; and at the same time we can't stop doing the R&D that will actually take three or four years to get done. We just can't make our customers wait three or four years for the things which should have been on more interim cycles. So, we're trying to pace ourselves, not pace ourselves, but to orchestrate ourselves so that we have innovations coming on all three of those cycle paths.
MODERATOR: Can you point to an innovation, something that you're working on now that might be somewhat public that you're proud of?
STEVE BALLMER: We're in the middle – we're at the beginning, I should say, of 12 months of the greatest innovation pipeline our company has ever had. So, I can point to Vista; I can point to what we're doing in IE 7, which I think is absolutely fantastic; I can point to our new server releases, Windows Server, SQL Server, the tools, BizTalk Server; the Windows Mobile stuff, this is the new MotoQ, small, thin form factor, brilliant screen, great connectivity to Exchange. I mean, I can point to a lot of things, Office "12," I think Office "12" will be the most exciting Office release that we've done basically since Office 4, 13 years, so we have a really rich – Xbox 360, not that anybody here would care, but maybe there are some children who care a lot about that innovation that's coming to market, the interactive TV stuff we're doing. We're in the middle of the best pipeline, frankly, that we've ever had as a company. And now the key is to make sure that for every line of business we have the things that pop every six or nine months, pop every couple of years, pop longer than that.
MODERATOR: Microsoft is a very large organization, one of the people that asked the question on the tape asked how you stay in touch with your users as a big organization? And then others have said, Microsoft has gotten a little bit of big and bureaucratic, and that's fed into this problem of delivering innovation. How do you drive that agility and stay in touch with your users as a large organization?
STEVE BALLMER: The first thing you have to do, and we're a company that believes in self-criticism, and sort of continuous improvement. So, I can tell you the good, I can tell you the bad, I can tell you the ugly, but it almost never matters where we are. We know we have to push, push, push, push to be better. So, I want us to have three cycle times, every business, three cycle times. We're going to continuously improve and push ourselves. Job one is having a mentality and a culture, and essentially a self-imposed requirement that speed and agility is important. We have made some big bets. If you go back four or five years ago, why Vista and "Longhorn" took so long, it's because we decided to make bigger, longer-term bets, instead of having innovation cycle on these three time frames. So it was a conscious decision. We had a different conscious decision, and we'll move forward with agility
In terms of really staying in touch with users, the more users you have, the more you actually have to use electronic techniques to stay in touch with your users. I know I've asked this at Gartner forums before, but I'll do it again. Has anybody ever got a message that says something like, an error has occurred, do you wish to send to Microsoft, yes or no? Does anybody in the audience ever get that message? Statistically, anybody who didn't put their hand up is misrepresenting the facts. (Laughter.)
I could say I ought to be embarrassed about that, but what we're actually saying is we want to know statistically the kinds of experiences people are having, so that we can get after the issues that are most important to our users. We're going to extend those technologies beyond what I would call crash analysis to usability.
In the Asian countries, we've extended it to the quality of the recognizers that we put in to help people type Chinese, Japanese and Korean characters. So the general notion of collecting feedback statistically from customers, staying in tune with it and responding is important. I certainly say XP SP 2 is about agile responsiveness. We got a job-one priority, security, and we got after it.
MODERATOR: Let's talk little bit more about some agility that you had to show last year. You went through something that's been known as the reset for Vista. Could you describe to the audience what that was about, and what's different now, what you did to shift your direction?
STEVE BALLMER: We made a call 14 months ago, I'm going to say, it would have been August of last year, so 14 months ago. We made a call that the integration challenges of trying to bring together a new operating system that had a new presentation service, file system, user interface, communication system, and to have all those things be co-dependent was not going to time out. We needed to introduce those innovations on a different cycle and in a different way. We made that decision last August. Once we made that decision it took us about two months to reset the team. And between the time we made that decision and the time we ship what will now be Vista, Windows Vista, we'll have about a two-year cycle.
During that same period of time, which I consider a pretty rapid cycle for an individual release of Windows, considering we still have the most major Windows release probably ever, certainly since Windows 95, Vista is that significant. But, we also retooled our entire engineering system. We componentized the architecture of Windows to give it more flexibility for rapid release going forward. We've changed our build test systems to give us greater agility in the way we do development internally. Wall Street Journal did a great article actually about that a few weeks ago. So we've really done a lot of things with pretty great agility since we made that call 14 months ago.
MODERATOR: OK. So agility, but many of our clients, and they've raised the questions in the video, are still concerned about security, and change, and a change in the development process just raises concerns about stability and security in the release. So why should they believe Vista will be more secure?
STEVE BALLMER: The changes we made to our development process are really all about security, reliability and quality. We were losing major cycle times essentially by not capturing security defects and quality defects early enough in our design cycle. Everybody here who runs a development shop understands the advantage of finding things early. We changed our process. We changed our tools. We know have tools that we've written that will post-process a piece of source code, will look for potential security defects. We've made that rich and integrated part of the development cycle. We've put a security set of dates and checks into that cycle. So while it's making us more agile and fast-moving, it's doing it primarily by putting security and reliability earlier and more appropriately into the product.
MODERATOR: So they should trust Vista day one?
STEVE BALLMER: I'm going to trust Vista day one, and I'll bet most people in this audience will trust it day one on their home computer, and then we'll decide what day to trust it at work.
MODERATOR: All right. Fair enough.
STEVE BALLMER: And I encourage you to do it early, but I must be honest amongst friends.
MODERATOR: Steve you've talked about some things you're doing to change the development process after we get Vista and Office "12," these big releases out next year. What should the audience expect going forward, in terms of the rapidity with which they'll get new releases and new innovation? What should they expect?
STEVE BALLMER: I think you should expect that each of our products will have perhaps its own cadence. But, we need to have service offerings associated with each of our products that allows us to feed innovations that are appropriate to the market on, let's just call it six, probably more realistically a nine-month cycle. We'll have other innovations that we feed on a few-year cycle. Call those major releases, as opposed to innovations that we're feeding from the service. And there will be some things that we're talking about and working on that may take us even more than a two-year type cycle. But, we're going to want to give customers visibility to them. "WinFS" is in this category today.
MODERATOR: How many years is that? Never mind.
STEVE BALLMER: This time it's three years, this time. We've worked on advanced file systems before, and our customers should be happy to know we don't give up on something that's important to them even if we didn't get it right. This generation we've been working on next-generation filing technology for three years. We initially thought when we announced it we should announce when it ships. That was probably not the smartest decision. The decision we've made now where we're going to tell you, it's coming, we're going to provide beta releases for people to play with, and we're really going to have done more of the customer focused interaction before we set a final shipping schedule. I think that's spot on.
MODERATOR: You made it very clear, Steve, that you want everybody in the audience to go out and put Vista on their desktop, and for virtually everyone in the audience I'd say, the Windows environment is their user environment, it's the place they boot into and live every day. But, for an increasing part of the world it's not just about the Windows desktop. In some cases that's just a launching point, and it's the Web, and various Web sites, and search bars, things of that sort that define their end user environment. Those companies like Google and others that live in that world are now extending down to the desktop with toolbars and search. How do you deal with that competition and that changing view of what the user's environment is?
STEVE BALLMER: In a sense we start with the leadership position we have. Where do more people spend time online than MSN, globally? Globally nobody spends more time online anyplace than they do with MSN. It's because of the strength that we have with our MSN Instant Messenger product, and with Hotmail, and with Spaces. We've got the leading blogging site, the leading online e-mail product, and globally, not in the United States, but globally the leading IM site. So we're very used to this notion of services that people are spending their time in.
When we look at Windows, we say to ourselves, what do we do in Windows to enhance those kinds of service-delivered experiences? You look at what we're doing with IE 7, security, browsing. You look at what we're doing with search integration in Windows Vista, and Office "12." You look at the kind of connectivity that we're building between Outlook and Hotmail, and the instant messaging software with some of the technologies that we have coming. We understand that link. You look at Windows Presentation Foundation, which is essentially a next generation technology in Windows that can enhance Windows apps, but can also enhance Web sites. All those things are about extending an experience that has a local component and has a component that, as you say, lives out in the Web.
MODERATOR: You know, your organization, the re-org that you just went through where you took your client organization, your server organization and MSN, and put them together, points the way of the future.
STEVE BALLMER: Yes, we think that there is an aspect of what we do that's got to be strong client, server and service. And most of our interactions, certainly with the customers in the room at work, has been all clients and server, we think service will be important from the consumer market on up to the enterprise.
MODERATOR: You made a statement about MSN a little bit earlier, that people spend more time there than anywhere else, but a bit piece of that is the instant message. Talk a little bit about search as one of the questioners asked at the beginning? People spend a lot of time in Google search, that makes it very visible to users, more visible, I would say, than instant messaging. So what's your strategy on search, and how are you going to compete in that market?
STEVE BALLMER: Well, I'm going to have to pick a bone about the more visible than instant messaging. People who use instant messaging love instant messaging. They're all separate services, is really the point, and I'm not sure of the importance of comparison is
MODERATOR: I think the Google brand, because they're going to Google, and it's even a verb, I'm going to go Google something. People don't say, I'm going to go use Microsoft
STEVE BALLMER: But, people say I'm going to MSN; that would be a way people speak in and , not in this country, but in countries outside the United States, and that's because AOL has such a strong position here. But, if you say if the real question is, instead of the comparison just, hey, how are you guys going to get after this search opportunity, the answer is, the good old fashioned way, with innovation.
Come on, let's do a little audience survey here, in general 50 percent of all searches do not lead to a desired outcome. I ask this audience, is there anybody in the room who believes that the search experience is not going to be dramatically different and better 10 years from now than it is today? No, people can't find what they're looking for, number one. Number two, is there anybody here who is going to deny the fact that it would be nice to get a level of integration between what happens searching inside the enterprise and outside the enterprise? I think not.
So I think between the opportunity to innovate in general, the opportunity to innovate in a way that's meaningful not just to an end user, but an end user in business, we see plenty of opportunity and plenty of opportunity in things that, who knows, people may or may not do. I mean, if you read the newspapers today, other than curing cancer, Google will do everything. But, the truth of the matter is, we've got our heads down, we've got very, very smart people just working on innovation, innovation, innovation, and we think that gives us a great opportunity.
MODERATOR: I'm absolutely glad to hear you say that, because innovation is key. So tell us what you're going to target more specifically to drive differentiation in that space? When we're back here next year, what is it you're going to do between now and then to really not catch up in some of these areas, but leap forward, and trounce your competition?
MODERATOR: But it is general innovation.
STEVE BALLMER: We don't trounce our competition, we compete. David.(Laughter, applause.)
MODERATOR: You embrace and extend.
STEVE BALLMER: No, just take relevance. If we can say a year from now, 70 percent of the time people use the MSN search service, they're getting the right answer, and that number is 50 percent with competition, that's a big step forward.
MODERATOR: Relevance is a good one.
STEVE BALLMER: Relevance is job one for the user. Now, for revenue it turns out there's a whole -- for the advertisers there's a whole bunch of other things that are very important, but for the end user, you know, relevance is really the starting point. Relevance to me, is it relevant to my data, is it relevant to things that I do, does it know about specific domains of expertise that I have, does search remember anything about me and use that to be more relevant, does it respect my privacy in the process of learning about me and trying to be more relevant.
We've got a list, seriously, as long as anybody's arm of areas in which we can deliver more value and more relevance. Natural language, still today search engines don't really parse what you type and try to make sense of it, they do a fairly dumb matching of words to key data that's been intuitive out of the indices of things that goes on, on the Web.
So there's a huge amount of are in which to innovate and if you look at the relevance measures today in the United States or the U.K., there's not much difference between Google and Yahoo! and MSN, they're really very close. That's not true in , that's not true in , but in the and in the U.K. you'd see very little difference frankly, Google is a little higher, but you wouldn't see a big variation between us and Yahoo!
MODERATOR: So you're telling us you want to break out in terms of that relevance ranking?
STEVE BALLMER: That's key.
MODERATOR: Let's talk more about innovation here. You talk about not trouncing but competing with your competitors. So one area --
STEVE BALLMER: We're also glad to build share, Tom.
MODERATOR: One area where there's a lot of innovation taking place right now is virtualization, and virtualization is hot. In fact, let me ask the audience: How many people are looking at or already deploying virtualization solutions with Windows? It looks like most of the audience or a lot of hands.
MODERATOR: Significant amount.
MODERATOR: A lot of hands.
You had competitors who were putting virtualization solutions underneath Windows four years ago, you have Open Source looking to do it next year, you've announced a hypervisor that's going to come out in the "Longhorn" wave, which means maybe 2008 or so. Are you too late or are you just right for the market?
STEVE BALLMER: No, we have our Virtual Server obviously in the market today, we're going to extend the capabilities of our Virtual Server product in what we call our Release 2 timeframe, which is here in the next several months. So we have a product in the market today, we're going to expand its capabilities significantly here in the next couple of months. We have virtualization at the client today, and we're going to extend it fairly significantly in the Windows Vista timeframe. And then we're going to continue to move forward, including the introduction of a hypervisor layer in the operating system itself.
But if you look at what people are doing with virtualization today, whether it's from EMC or from us or from anybody else, we think the Virtual Server product we have in the market will be quite competitive, frankly, in this Release 2 timeframe.
MODERATOR: The other issue here is that there is innovation taking place in other places, so we mentioned in open source. One of the questions from the audience again is how you're dealing with open source. The Linux competition, especially in the server, has mainly been about who's going to capture business from some other places. So we've seen a lot of migrations from UNIX to Linux, we haven't seen as many migrations from UNIX to Windows. So are you losing some business there and what are you going to do to stop that? Why should they go to Windows instead?
STEVE BALLMER: I think we have four big opportunities to take business from Linux, and we will. And again why will we take it? Because people will take a look at the tools and technologies we put in the marketplace and decide that they deliver better results at lower costs.
What's the first? High-performance clusters. High-performance clusters is a thing that has been a Linux stronghold, it's about 20 percent of all Linux systems. We're coming out with a compute cluster edition of Windows Server, we're coming out with new development tools that help people write applications that make sense in that kind of scientific computing environment, we see a great opportunity to thrive with innovation versus open source, versus Linux.
The same thing in Web hosting. With the new work we have coming out in Visual Studio and ASP .NET, we see a great opportunity to offer hosters, lightweight hosters a better solution than Linux, and that's been a traditional stronghold.
The same thing in server appliances, primarily for security.
MODERATOR: Are you seeing success? Are you seeing success yet, especially in UNIX migrations?
STEVE BALLMER: We're seeing a lot of interest right now in the cluster work we're doing, and we've had success in migrations on hosting. People are moving UNIX applications off of Sun and HP systems, we are not winning more than we're losing. I mean, I'd say more of those are going to Linux than us, and yet we're winning a significant percentage; maybe 25 percent of the deals that we engage in, when somebody wants to move off of Sun or AIX, we're winning. Sure, I'd like to win 50, 55, 60, 65, whatever the appropriate percentage is –
STEVE BALLMER: – and that means we still have work to do. The day I can come in front of a Gartner audience and say we have a better UNIX than Linux; that will be a good day; we're not quite there yet.
MODERATOR: That will be a very interesting day. (Laughter.)
STEVE BALLMER: The day you let me say it will be even more interesting.
MODERATOR: That's probably true. (Laughter, applause.)
I want to get you back to that ASP .NET and ask about an interesting thing with the Windows environment and its evolution. You've traditionally, particularly over the last few years, positioned Windows as this rich client environment, and the browser and the Web is the simple client environment. But we've been seeing a little bit of a change in the market; Ajax, which I need to tell the audience is a way of delivering a rich client experience down to these Web browsers in a dynamic fashion, has been gaining some traction, closing the gap on that user interface between Windows and these downloadable capabilities. Your own projects around "Atlas" and other areas are building on top of that.
So I've got to ask you where these two things fit strategically for Microsoft in the future. Do they converge, do they stay separate? Where do they fit?
STEVE BALLMER: We think most users, most end users, appreciate the benefits they get out of a rich local environment with rich visualization, the advantages of processing, the advantages of storage, et cetera. I don't think that's going to go away. We've been through this debate a hundred times; I don't think it will go away.
MODERATOR: Can you talk briefly about "Kahuna," which is the codename for the front-end to Hotmail that uses some of these technologies and looks a lot like Outlook, that from the user looking in seems to be really driving a convergence of these models?
STEVE BALLMER: Yeah, I do think that there will be advantages of service-based and an advantage in using rich functionality in the client, and we will lead the way, as you indicate, in the work we're doing around the new version of Hotmail.
MODERATOR: We should talk about service-oriented architectures then.
MODERATOR: That's exactly what I wanted to go to next. One of the questioners also asked about service-oriented architecture, and it's been gaining a lot of hype on the market, many people are talking about it, we had many presentations here. But Microsoft hasn't been nearly as visible, at least in terms of the message, and we see at least with mindshare, Microsoft falling behind a little bit.
Tell us what the real strategy around service-oriented architecture and service enabling your environment is and what Microsoft is going to provide, particularly to enterprises to exploit and benefit from service-oriented architecture.
STEVE BALLMER: If you look at the actual stats out there, I think more people have written service-oriented architecture components, that is, components that speak Web services, more of those have been built in Visual Studio with .NET than any other platform. I'm not going to dispute with you what the PR dialogue looks like, that's in the beauty of the beholder. But the number of our customers who have used .NET and have used our Web-services infrastructure, the work we've done in collaboration with the Suns and the IBMs and et cetera of the world is really quite incredible.
I think we'll see a ramp up this year with our new Visual Studio, with the new ASP and with our BizTalk product, which helps you stitch together, if you will, a number of these components that have been written to a service-oriented architecture.
So we're doing a lot in the area, it's an area of great innovation, it's also an area where I expect to see essentially a simpler, lower, lighter weight kind of service-oriented approach also break out that uses more RSS-like protocols, and we'll certainly lead the way there, too.
MODERATOR: I think one of the splits that may exist is in the enterprise-class focus around service-oriented architecture, which oftentimes has to take into account the mainframe and complex applications and complex database environments, which doesn't really seem to be as much Microsoft's focus versus some of the other vendors. So can you hone this down a bit and talk about what Microsoft is trying to do long term to deal with that enterprise class service-oriented architecture in the general environment?
STEVE BALLMER: I think if we're going to be blunt, every vendor focuses in on the platform that they own, and the beauty of the Web Services interoperability stuff is it essentially gives you a layer at which to put things together.
We're clearly doing all of the pioneering work with Web services on Windows. IBM is clearly doing all of the pioneering with Web services on the mainframe. Three or four guys are trying to do the pioneering work with Web services on UNIX. And that's why the work, the standards work that we embrace with Web services interoperability is important.
It is not our job to lead on the mainframe any more than I would say it's IBM's job to lead on Windows. It's both of our jobs, despite the fact that we compete, to work strongly on interoperability standards for Web services, and we've embraced that wholeheartedly.
MODERATOR: You talk about working on your own platform. One of the changes in strategy at Microsoft has been in management, in the back office. So traditionally in the past you've relied on third parties to provide management tools and you grew a rich portfolio of third parties, and now you're taking it more front and center yourself. Our clients have seen a large number of initiatives and discussion about TCO, but they haven't seen a lot of reality, a lot of vision there. Where are you going with management and TCO at Microsoft?
STEVE BALLMER: Well, if you ask sort of from an IT perspective, what's the number one thing we get hit on today, frankly, it is TCO even more than it is security. That's different than two years ago when I would say it's security. Security hasn't fallen off of our plate, but TCO has certainly risen up.
We've done some work, actually, with Tom on what we call the Infrastructure Optimization Initiative, designed to take a look at what does it take to drive costs out of the kind of infrastructure with our platform that people run. If you take a look at the work we've been doing in Windows, in Windows Server but also in our Systems Management Server, our Operations Management product, we've got new products coming out to help people manage storage, we have a new product coming out that will help with capacity planning here over the next several months, you will see us build out a very rich suite of management products essentially that work very well on our platform and in conjunction with partners can be used in a heterogeneous environment.
MODERATOR: We talked a little bit about this last year, and when we were asking you about your priorities, management was on the list but it was further on down. Is it moving up or is it still down on the list?
STEVE BALLMER: Well, management is high on our list. Total cost of ownership is high on our list, management is job one in a sense for total cost of ownership.
It turns out that almost anything you talk about, we talk about these days as it relates to enterprise customers, the key is management. Virtualization, what's the most important thing? Management tools to help you manage virtual machines. Security, what's the most important thing? Not just the protection technologies but management to help you watch and secure an entire environment. So I'd say the management agenda is quite high.
MODERATOR: So TCO, we also have to pay for this, and we had a number of questions about pricing and licensing. With the hardware technologies mainly in the server changing the way software needs to be priced, multi-core processors, virtualization, tools such as that, how do you see licensing and pricing changing in the future, how is Microsoft going to help change it?
STEVE BALLMER: Well, we made a major announcement last week in virtualization, including the way we're thinking about licensing in a world of virtual machines, we made an announcement last year in terms of the way we're thinking about licensing in a world of multi-core processors, and I would say not only from Gartner but broadly through our customers both of those were well perceived, well taken as being let's say favorable from a customer's perspective from a price perspective.
We still account for, in the average, large enterprise less than 1.5 percent of total IT spend, and we think we deliver a good value. We're going to continue to work to broaden our portfolio. We're working now in the portal area, we're working in BI, we're working in management, we're working in search, we're working in workflow, we're working in document management. So we want to expand our footprint with these large enterprise customers, but we're going to continue to work to offer very good value.
MODERATOR: It sounds potentially complex. I want to go back to the question that got a big round of applause from the audience: What are you going to do to simplify this licensing world?
STEVE BALLMER: The simplest thing we have today, and we have a lot of customers, about half of our enterprise customers find it quite simple, is our Enterprise Agreement. The best thing we can do frankly we find with customers who want to go this direction is put together a license agreement that essentially takes out a lot of the counting and a lot of the complexity. It's not appropriate for everybody, but we think we have a form of our agreement that does take out a lot of that counting and complexity.
We still know that there are things people struggle with, so we recently simplified our license agreements. According to these standard statistics it used to take two years of post-graduate education to understand our license; that's down to 9th grade now. (Laughter.) I'm not sure how many 9th graders are going to go read the license, but that's at least what the reading specialists tell me.
There are things that we're doing to help our customers understand license position. There are things that we're doing to try to help customers understand how an architectural decision drives the licensing decision.
So we know we have a lot more work to do both in terms of the tools we give our customers and the license forms.
MODERATOR: How do you make sure that some of this world of service-oriented architecture that we talked about earlier, and creating all of these granular components, doesn't further complicate the licensing issue? What's your thinking around building this service world into it in a simple way?
STEVE BALLMER: I think the last thing you really want to do is move backwards and license components. What customers really want to do is license things they understand and license them at a fair price.
And the thing we struggle with right now is do people understand what they're licensing. Most of our feedback is not overall on cost, although cost can be an issue, most of the feedback I get is, I don't know whether I need another Terminal Server CAL for this configuration or I don't. So continuously working on simplifying the packaging and making sure that the price for the value, if you will, is in balance, those are the two key principles.
MODERATOR: Both in a virtualized world and in a Web services world.
STEVE BALLMER: In a virtualized world, in a Web services world.
MODERATOR: They all fit together.
STEVE BALLMER: Look, you know, there are customers who occasionally will come to us and say, "Look, we've got the following problem, and if we read your license correctly, it strikes us that this will cost us a lot more money than we're willing to pay." And sometimes there actually is a disconnect, we have not anticipated the customer's environment very well, in which case we have a team of people who are trained to try to accommodate and work with those customers.
MODERATOR: Let's go out into the future a little bit. So we've talked about a number of challenges and changes in technologies on the client side that may change how computing is done from a desktop, what is my desktop. We see technologies like RSS that are very interesting to be integrated across the Web and perhaps the desktop, virtualization that may change how we deploy servers and so forth. There's also competition, new competition that's getting more interesting.
So there's challenges there from a competitive standpoint and changes. You also have demand from shareholders for growth, you're also having demands to change how the company operates going forward.
So where are you taking this audience over the next five years on your journey to answer those competitive demands and also to grow the company?
STEVE BALLMER: Well, to me these things are not very disconnected. You get the following loop: You do things that are innovative and bring value; you respond to the big core trend, which is to move more towards a services approach, whether it's service-oriented architecture, whether it's on-demand delivery through virtualization or whether it's services running out in the cloud, there's sort of a theme there. So innovate, drive with value, make sure you're keeping up with the latest trend, which is let me call it just software as a service, if you will, and if you do those things well then you compete well and you grow.
So I'd actually say if you ask me, where are we trying to take all the folks in this audience, we're trying to take them where they want to go, because I think if we're doing the innovative, high value work that helps folks like this, we do get to grow. And I mean folks like this broadly in your professional capacity, in your personal capacity, in your home life, in your business life; we're trying to have a platform that supports you in all of those activities because all of us we're IT people, we're end users in our businesses, and we are people who want to have a life.
MODERATOR: So when we talk about innovation and change in the future, what do you want them to walk away with as watch-this-space, this is the thing that's really going to change the paradigm?
STEVE BALLMER: Well, today I'd leave you with just kind of two things to think about. Number one, I do think this notion of all significant software evolving to have a service element is important. And you see that in what we've done with a bunch of our acquisitions, FrontBridge and some of the other things, that would be one thing.
Then the second thing very short term that I'd leave you with is remember this is the year when we have more exciting innovations for this audience than ever before. Whether it's the new Moto Q – paid political announcement for this nice, thin, wonderful device to take your office on the road with you – whether it's the work we're doing here, the work Office will bring in BI and document management and workflow in individual productivity, new user interface, Vista; the thing I would leave you with is this is the year to really engage, because we think we can help you help your users make a quantum leap forward, make them more productive.
And if you believe at the end of the day IT is about putting the people who work in your companies at the center, making them more productive, making them more valuable to the company, I think that's a message we all need to learn and we've got the technologies to help you this year.
MODERATOR: Steve, thank you very much.
STEVE BALLMER: Thanks, everybody. (Applause.)