Bill Gates Keynote
NationsBanc Montgomery Securities Conference
San Francisco, CA
January 27,1998

Good morning. It's a very exciting time in this business, and I think now more than ever we can say that the companies that really take the long-term view of R&D, of relationships and how they work with customers, are going to be the ones that get ahead. Our philosophy of working on a long-term basis is one that anyone who follows our dialogue with the financial world has seen. We're very, very careful to point out what a competitive business it is, and that it's a business with risks. And because of that, I think, we've had a great relationship with the people who have been involved with our stock from the beginning.

The PC industry is the fastest-moving industry on earth. It's delivering better products at lower prices with more capabilities. When I first did a conference that David was in charge of, a PC cost $5,000, and it wasn't good enough to do a presentation from, so I used slides at the time. Today I'm using a machine that costs about one-quarter of what that would have cost, and I haven't seen a slide for many, many years.

The miracle that really drives this is the exponential improvement in chip technology, and that will be faster in the years ahead than it has been in the last decade. We're going to have a lot of performance to put into these systems, and rather than simply think about what we're doing today and doing it faster, you have to think about entirely new capabilities and whole new ways that you'll interact with the machine. And that's where our R&D is focused and what I want to highlight here this morning.

It's not just hardware prices that have come down. If you take almost any software category, the incredible number of machines out there, and the presence of a standard operating system (in particular, Windows) has meant that the software industry can thrive, while at the same time bringing down prices. And those lower prices have been fantastic for volume.

If you take the marketplace where Microsoft has its highest market share, which is in office productivity software, prices have come down a lot. In 1990, just to take a few applications, three applications would have cost more than $1,000, and there really weren't any discounts around to get below that. Today if you go out at retail and buy the five products now, all of them dramatically improved, the highest price you'll find is $499, and for corporate customers or educational customers, the price is dramatically lower than that. And that's a good thing, because it's opened up new markets and for us the size of the business is big enough to allow us to do as much R&D as we want. In fact, the only thing that holds back Microsoft R&D is our desire to be very careful about who we hire. So it's not a financial constraint, it's just the ability to bring in great people and make sure we retain the quality that we've had throughout our history.

Besides very high-speed processors like the 330 MHz Pentium II that Intel introduced last week, we have other areas of innovation. And these are critical because they will change the character of the machine. For example, I think people still deeply underestimate the impact of low-cost, high-quality flat-screen technology. At a meeting like this, very few people are using a computer in this session. I predict that in five years, the majority of people in a gathering like this will have a tablet-type device, the same size as a tablet of paper, but with the ability to take notes, look at information and recognize your handwriting. It will be something that will have more power than today's personal computer, and yet be dramatically more convenient.

Now, there's a lot that Microsoft needs to do to create software that enables that new machine. The user interface has to change; the ability to work with large documents easily has to change; and we're hard at work on pulling together those advances.

The storage industry is to be congratulated for moving their capacities up even faster than the exponential improvement we have seen in chip performance. In terms of their financial results, it makes it a very tough business to be in. They are providing so much innovation that, in some cases, the market doesn't catch up to what they make available. In the long run it will, because stored digital video, stored digital images will use immense amounts of capacity. Today it's hard to buy a drive with less than 4 gigabytes. You don't get any discount when you go down to 1 gigabyte, or hundreds of megabytes. And for people who have been in the computer industry for a long time, it's kind of mind-blowing because you grow up with a sense of storage being very scarce and thinking, "Well, I can't waste a megabyte, that's very precious." And yet, as we make design tradeoffs today, taking advantage of that large storage really is the right thing to do. It's a new world when the computer you buy can store more than everything you could type in your entire lifetime.

What that says is there will be new things that will take advantage of those advances. For example, digital photography combined with the PC will be an explosive area. It's not quite ready yet, the pieces are still falling into place in terms of the cameras and the software and the printers, but five years from now, that definitely will be in the mainstream.

If there's any area that I have a little bit of concern about at proceeding fast enough, it's communications technology -- being able to connect these machines up at very high speeds. In order to achieve that vision, we need more than just a phone dial-up connection. Even the 56 KB that's possible doesn't allow people to get information quickly enough, or in a rich enough way, to allow people to move on to what we call the Web lifestyle. There are some advances, both in cable modems and with ADSL, which could get us there. In fact, Microsoft is very involved in both of those things, encouraging the companies, helping to set standards, and making sure our software has the built-in support.

There is a new low-cost form of DSL that has just come up recently that it looks like will have the support of the entire industry, and that could be a breakthrough where we can build the modems into the computers and get the monthly fees down to where we can create the large market. The only reason I'm more cautious about this part of the technology world than the other pieces is that it is a regulated environment, so there's always a question of whether the right incentives will be there for the phone companies or the cable companies to charge ahead at full speed.

Now as people get all these advances, they're going to have to be packaged up in a simple fashion. They're going to have to be made so they are easy to install. And there are certainly some challenges there. In fact, we have put together a little video that talks about what this installation is going to be like. So let's take a look at that.

Well, there's still some work to do there. The computer industry has become dramatically more competitive in the environment of the PC where you have hardware substitutability. And it's been interesting, as the PC technology has taken over more and more of the computing environment and found a way to coexist very well with the existing devices out there, most companies have really had to change their strategy. As a result of that, Microsoft has had close partnerships with virtually all the large computer companies. In fact, we view most companies in our industry as pure partners, with two exceptions. IBM is part partner, part competitor, and then you'll see Sun on there who's just a pure competitor. But everyone else, we're doing exciting work to extend our software to take advantage not only of the product innovation that they're doing, but also the services that they bring out to the marketplace.

Now the key strategy for Microsoft has been to get lots of software written on top of our platforms. It's interesting, we don't restrict what software people write for Windows. It's not something where you have to sign an agreement. Anyone can do it -- you can go to a local bookstore and get thousands of books on how it works. So it's completely royalty-free and open. But we still feel like putting hundreds of millions into evangelism, into getting the tools out there and getting the latest information out there is a very smart strategy for us, because it helps developers take advantage of the latest versions of the technology.

We're excited that investment in the software industry is at an all-time high. We're around 4% of the software industry, so we can grow a little bit without being too big a part of this picture. And part of the reason for that is there is an incredible variety of software out there. People like SAP, BAAN, PeopleSoft- those are partners of ours, where they are exercising their expertise in businesses that we don't have any expertise in, so it's perfectly complementary.

Now the challenge here is to be a company that obsoletes your own products. Customers' expectations and what's possible is changing fast enough that no existing product will last out into the future. It's what happened with word-processing, going from typewriters to Microsoft Word. But it's clear that what Microsoft Word does today will not be enough unless we can make the product understand natural language, unless we can integrate great speech recognition and grammatical understanding, and unless we adopt it so that electronic mail collaboration is very easy from in the Word product, we will be replaced by another word-processing product. So we are hard at work doing that, and our real asset there is the feedback we get from customers and the quality of the developers that are doing that work.

So our big strength isn't some intellectual property; it's not today's products. It's really the ability to develop software with a great team. We do have a lot of product momentum, particularly if you look at the migration of 32 bit on the desktop. Now we're at about 95% of new PCs shipping with a 32-bit system. The installed base, now, we've passed over where the majority of the installed base is 32-bit, and very rapid migration compared to what most people expected.

Windows NT has been our great growth product over the last year, and that drives not only the Windows NT sales, but also the sales of the BackOffice products that run on top of that: Microsoft SQL Server, Microsoft Exchange, and the other elements in that combination.

We do feel that our research has been going very well. We're very pleased with the hiring that we've been able to do. We're articulating, now, two visions that are very important to us: Digital Nervous System, and Web Lifestyle. Digital Nervous System says that a company ought to think very hard, not about individual applications, but about the basic information flow inside their company. Do people use electronic mail? Have paper forms been eliminated? Do people who want to look at sales results have that information totally up-to-date in a way that they can dive in and see the patterns, which is simply not possible if it's printed out at the end of the month and put into a notebook. It's got to be live on the computer screen so you can mail it around to other people and dive in and understand what's going on as early as possible. So a Digital Nervous System is a network of up-to-date personal computers where you've thought about the key processes in the company, and you're dealing with the planned events like personnel management, project management, sales budgeting; and the unplanned events, like a competitor doing something you didn't expect or a customer being unhappy. All of that literally at the speed of light, because these things connect together in an integrated fashion.

It's my belief that the companies who do this well will have a huge advantage over the ones who do not, because the quality of the decisions and the efficiency of the way they operate, will be much better. And most companies are already buying the building blocks here, the PCs and the networks. The big difference is that some are not using them very well to change the information flow inside their business, and we want to help people seize that opportunity.

Looking at the individual level, the vision here is what I've called the "Web Lifestyle." It's when you don't think it's surprising that the way you pay your bills, the way you get news and sports scores, is to sit down and use the Web many times a day, either with a full-size PC device, on your TV set, or with hand-held devices. You simply take it for granted. In the new house that I've built, I have a 45 megabit-per-second connection to the Internet. I've got a lot of nice, high-quality screens, so I claim to be living the Web Lifestyle. So far I'd say it's pretty great being able to call things up at any time, and not having to wait for things to dial up; not having to wait for it to be slow, and just knowing that information is easily available makes all the difference.

The company has a few key focuses. In the product area, obviously updating the Windows product is central to us. This will be a big year for that because around the middle of the year we will have the beta-2 of Windows NT 5 and we will also have the shipment of Windows 98. We will have a major new version of Microsoft Office coming out late in this calendar year, and many updates in the BackOffice area, in particular some updates to our Microsoft SQL Server product and the Microsoft Exchange Server product.

Now, spanning all of these products are some key initiatives. The Internet: we've come a long way towards building that into our products. There is still more to be done. With things like videoconferencing and phone-quality conversations across the Internet, the standards are not there yet. So Microsoft is one of many companies that is working through the committees to get richer Internet protocols.

Internet security: there is a lot that has to be done to make sure that you can trust that your information is not only private, but that the transactions are being accomplished in an effective way. We have some issues there with government regulations, in terms of the encryption software that we can export, and there is a lot of time going into making sure that this issue gets explained.

Scalability is a big thing for us. People often come to us and say, "How far can they go using PC technology?" For example, the transaction speeds in our database products are now ten times what they were a few years ago. And so we are getting to the point where we can take on very, very large tasks, not only by interoperating with the older machines, the main frames and the high-end UNIX systems, but also by moving those applications onto Windows NT. And making sure that customers see that progress and they know when they can move those applications is a very important thing for us.

The progress is very rapid, partly because the hardware advances are there and also because we're using more processors in every system and we're clustering these systems together. So we have architectural improvements, chip improvements, and software improvements that all compound so that you can get more than your standard exponential type improvement.

I'd say the most important initiative for us now is the third one, simplicity. It's too complicated to work with a PC. Updating your software, knowing where your files are -- a lot of people are reluctant to go out and get the latest software because they're afraid that their machine is so fragile that they'll break something if they do the update. And the model of being able to use a high-speed connection and just use a few clicks to get the latest software, that will be a breakthrough which will grow the entire software industry, whether it's in the business environment where people will get updates on a flexible basis, or in the consumer environment where there is a tendency for many buyers not to ever go beyond the software and the versions they buy with their machine. So we can use technology to help us a lot with simplicity.

One of the interesting things that we're doing right now is going through these scenarios, seeing why people get so confused. The file system is very opaque. The error messages are very opaque. We sat down and went through all the error messages on our products, and found that even we thought they were a little cryptic. For example, here's one that you'd love to get: "The DHCP client could not contain an IP address. If you want to see DHCP messages in the future choose Yes, otherwise choose No." Well, you don't really know what to say there! Sounds like it could be a bit of a problem, but it's certainly not intuitive. And so we have to use the technology to hide the complexity and make sure that people can get done what they want to get done.

We're very enthused about getting these upcoming product releases. The themes of these really are Internet, simplicity, and scalability. We've got a new foundation for developers in the way we're evolving the COM architecture to make it easy to build distributed server applications, and that is a big breakthrough. I think that over the course of this year people will see that building rich applications is far different than it was in the past. Take for example, building a transactional website. To do that today is very, very hard, and we're going to make sure that it's very, very easy.

Things like our Transaction Server, having that built into the operating system, is a key step in that direction.

Now, Windows NT 5 is by far the most important product release we've had for a long time. Windows NT 5 represents our breakthrough in terms of cost of ownership: the ability to take all the user states, and instead of having any of that be on the local machine, and instead store it centrally on the servers. That means that if a machine breaks, you just put in another machine, and immediately you're up and running. If you want to go to somebody else's machine and use that, you just log in, and all your information is there. It means that making users homogeneous is a simple command by an administrator, and so the administrator can be sure that the applications will work the same across all those machines.

The technology that gives you the best of both worlds, that is, the best of the PC with its responsiveness and portability, and the best of the centralized world, with being backed up and easy to administer, is called IntelliMirror, and that's built into Windows NT 5. You have to have a Windows NT 5 server, but you can get some of the capabilities without going to the Windows NT 5 client on the desktop. To get the full capabilities, you do have to move to Windows NT.

A big push for us will be getting the business desktop moved from Windows 95 to Windows NT. We expect that will take many years, but we'll have some pioneering customers who are doing that in an aggressive way with the rollout of Windows NT.

Windows NT 5 is a big advance for the developer platform, and it's a big advance for the directory. The kind of rich security that lets you share information across organizational boundaries is what we've built into Windows NT 5 with the Active Directory.

You're starting to see us not only expand our product line on the high end, up to clustered multi-processor servers with partners like Compaq, Digital and HP, but you're also seeing us take our product line down into very small devices. And this is a very important thrust for us. In fact, I've spent a lot of time in the last year on this part of our strategy. We introduced at the Consumer Electronics Show just a few weeks ago what we call the Palm PC. This is one here that happens to be made by Philips. We have some great hardware partners in these businesses: Casio, Sharp, Hewlett-Packard, Philips, Compaq--quite a list of companies that in the same way that they build the PC hardware and get the software standard from us, they will be building these devices. They have a lot of flexibility in terms of how they configure the device: things like the communications options or extra applications they put in. We think this market will be one that grows very dramatically.

Part of this is making sure these devices work well with your PC and that you'll automatically share information without having to give any commands.

Another category of this hand-held PC we've seen a real explosion of sales in the last month, as these color devices have come out, is this one from Sharp. There's also one out from HP, and they're having a hard time making enough of them relative to the demand. They're relatively expensive compared to the other handheld PCs; they're about $900, but with the 2.0 software, which has richer applications, including a PowerPoint-type capability, they're proving to be quite popular.

Another interesting category is the computer in the car. This is the first time you're seeing this use speech recognition as part of the interface, so that you don't have to look down at the computer screen instead of paying attention to your driving.

Another category is using Windows CE in the intelligent set-top box, and we reached a big milestone on that a few weeks ago when TCI decided to include Windows CE in every set-top box that they buy. It was an order for over five million units, and it's not just TCI; it's all the different cable companies that they've been working with. And so that's a big boost for Windows CE.

What are we going to do here? We're going to get lots of applications just like we did with Windows. Our tools are quite strong here, and our evangelism is just kicking into gear, because we need to get the volume out there to initiate that bootstrap. But we see that as a very key area that over time will be a significant revenue generator.

We keep growing our R&D. Of any large company, we are the most R&D-intensive company you will ever see. If you study our P&L, you have to say thank goodness COGS keep going down. Thank goodness we have G&A (General & Administrative costs) under control. Thank goodness we have scale economics in our sales and marketing expense, because even with our sales increases, we have managed to drive up our R&D percentage. A few years ago it was down to about 13-14%, now it's up at about 17%, and if we're able to do the hiring the way we expect, we'll get that to be around 20%.

Why are we spending so much on R&D? Well, because there are great opportunities. Computers today, they don't see, but they should see. They should know who's sitting down in front of the machine and what kind of expression they have, or what kind of gestures they are making. Computers can't talk. You can't talk to your computer. It doesn't understand the semantics of the information, so you get things like you saw in the video, where you're searching for information out on the Web, and you get just pure string matches that don't fit what you're interested in.

Moving up in graphics, moving up in scalability, changing the way we do software development to make our software factory even more effective, and getting rid of some of the manual testing that goes on today: these are all breakthrough areas that I think we're making fantastic progress in. And so, because of that long-term outlook that I mentioned, we're going to continue to do more and more in this area.

A key point, though, is that all of this will depend on the partnerships that we have. And we have lots of different kinds of partners. PC makers, chip makers, communications companies; but perhaps more importantly, the thousands of small companies out there that do the solutions, that actually put the systems in place, that do the consulting and the training. And we have special programs to reach out to every one of these partners.

Microsoft's business plan was kind of a wild business plan when we got started, because we assumed a complete restructuring of the computer industry, where a company could be a specialist in one area--building high-volume software products--and then rely on other people to work together and do all those other pieces. And that model has worked extremely well for us. One example of this is that we've chosen not to be in the service business. We've chosen not to create a profit center around a large consulting group or an installation group, and although we have consultants, we measure those consultants by their ability to transfer knowledge to the customer and move on the next project, which is the opposite of what you would do if you had created consulting as a profit and loss area. If you look at our revenue breakdown and compare it to some other people, we get less than 2% of our revenue from services. Now, other partners are in much higher percentages, with Oracle being the extreme case. They have really decided to be a service organization, and that is a perfectly valid strategy. But our strategy allows us to work on a much closer basis with the large service organizations, people like EDS and Andersen, and, perhaps even more importantly, with small entrepreneurial consulting organizations that have really latched onto the new technologies and picked customers where they understand their needs. And they are really the change agents in this marketplace. They are out driving those new technologies.

Developers are a special type of partner that I have talked about. We have a set of CDs we send out; it's a lot of information, because we're updating it quarterly; but there are over 700,000 people who get those CDs on a regular basis, and that's got all the beta software, it's got all the rich new tools, all the information about what we're up to is there. And so we're able to serve these customers through the Internet and that CD distribution, better than we ever have before.

Our tools, if you go across the board: Visual Basic, Access, Microsoft Excel...We have over 75% of all development that gets done, gets done with unique Microsoft tools. If you look at the remainder, most of that is still C today and it will stay C for a very long time, because in terms of very efficient development on the code basis that is out there, that's very significant.

We have a strategy of supporting all different development approaches: C, Visual Basic, Access, Java, you name it. We're going to work with developers to let them use whatever language they want, mix those languages together, and take advantage of the Windows platform.

So let me just close on a very optimistic note. I think the software industry is just at the beginning. The competition has us all moving forward very aggressively, the volume is allowing us to put more into R&D while providing products that lower prices, and the specialization in this business is working. That means that partnerships will continue to be the right strategy and one that we pursue. The feedback of customers is really amazing. People are interested in software. They love doing beta tests. They love telling us how we can make it better. I don't know of any product where you can get the customer so engaged in helping you move it forward.

And that's partly why I love this business. It's partly why it's not only the fastest moving, but also the most fun business to be in. And certainly in the next decade, I can assure you you'll look back on the computers we have today and ask, "What did we use them for?" They couldn't see, they couldn't listen, and they couldn't even learn what you're interested in. And it will be the magic of software at the center of those systems that is making that all possible.

Thank you.