Competing to Win
Financial Analyst Meeting 2004
July 29, 2004

KEVIN JOHNSON: I am delighted to be here to spend the next 30 minutes talking to you about my favorite topic—winning customers. Now, in the last two weeks, I've had the opportunity to address our entire field, sales, marketing, and services organization at the Microsoft Global Briefing, and over 5,000 partners at the Worldwide Partner Conference. Our priorities and goals are very clear: Win customers, drive satisfaction, and grow the business.

Now, I'm not going to spend time on all the great work we're doing to drive satisfaction and grow the business. I want to focus on “win customers.” Now certainly it's a very competitive landscape out there. It is crystal clear to our field and to our partners that we are focused on winning customers by putting the customer first and focusing on the value we deliver, differentiating that value from competitive alternatives, and supporting that value with evidence. It's crystal clear the competition and the competitive alternatives of Linux and open source, and over the last year or two years we've seen the debate really shift. We've seen it shift from an emotional debate over developmental models—open source development models, commercial development models—shifted then to a debate of technology features and functions. And here over the last year it has shifted from that to a focus on value, business value, and using factually-based evidence to support good business decisions.

Now, I want to just provide you a little bit of market overview. But then I want to set the stage with real customer examples and how this plays out in the field. Over the last year I traveled over 120,000 miles, working with thousands of customers and partners, and our field teams around the world, and I want to base this discussion on those interactions to real situations that happen on a global basis.

Let me just first start with some market trends, market indicators. The first thing we look at in terms of some numbers is, look at the last 12 months. Windows Server units have grown faster than x86 server shipments. So if you start to look at what's taking place there, you see that 62 percent of the x86 server shipments are Windows Server units. And that's up two points over this last year. Linux is at 18 percent—they too are up two points. But fundamentally what's happening is between Windows Server and Linux—it's taking share that's declining in UNIX, proprietary UNIX, UNIX on x86 servers, and Novell.

We now shift to PC shipments. This is a Gartner study or analysis that shows fundamentally flat in terms of share shift over the last 12 months on PC shipments. I look at Windows bootable operating system units: They grew faster than PC shipments, and that's fundamentally part of the work that Will talked about in terms of unlicensed PCs and the work that we're doing. But we're holding flat relative to that share of Windows bootable operating systems relative to PC shipments. The third set of data I show you is the study that's done by Management Insight, a consulting firm that goes out and researches every year what value-added providers have to say about what they see in the market and their intentions going forward. And this particular thing looks at server preference. It asks I think 15,000 VAPs their server preference going into this year. And this is sort of a leading indicator of what we see taking place. And you see that Windows Server preference is up 7 points. There's a set of momentum and shift that is building due to a number of different trends and issues in the industry.

Now, some of the market perceptions. Certainly everyone agrees on the shift to x86 servers. You know, proprietary UNIX platforms don't provide the price, performance, and flexibility that Intel or x86 servers provide today. Certainly Intel architecture hardware providers are benefiting from that shift. And it is a significant amount of demand in the marketplace for UNIX migration. In fact, there are some partnerships where we've launched a UNIX migration offering—with one of our partners we launched a UNIX migration offering in the United States, and within like a 45-day period we had built a pipeline of about 300 individual opportunities in the customer mix. The interesting thing was that about a third of that pipeline was actually Linux-to-Windows migration. So there is this shift taking place, the customers are really trying to drive to get to x86 servers. And when they do, why do they evaluate Linux? Why would they consider Linux? Well, it's that desire to migrate from UNIX platforms, where they might say, okay, it's going to be easier to move from a UNIX to a Linux. But really a lot of it has to do with this perception, the perception that there's a lower cost of ownership on Linux. And that perception is not reality. And that's really where we've been grounded in focusing our efforts and discussion with customers.

Now, certainly IBM was promoting Linux and certainly you talk to customers, whenever they have a Linux proposal or a Linux solution from IBM, it is really wrapped with IBM Global Services. It really is viewed I think from a customer perspective as a tool to really help drive IBM Global Services in the solutions they bring to market.

And there's also this market perception around what I call the commercialization of Linux. It starts with IBM's investment and focus. It's Red Hat announcing its enterprise services and support price, you know, stopping providing the free distributions of Red Hat Linux for desktops, Novel's acquisition of SuSE, the financial reports around Red Hat, and Novell and the other commercial companies really building this business around Linux. There is this trend of commercialization of Linux.

Now, how does that play out from a customer perspective? Well, there really are a set of considerations that customers use when they are evaluating or going through this decision process. First they look and say, total cost of ownership. You know, over the last three to four years, customers have been finding ways to tighten up their IT expenses—run their infrastructure in a lower-total-cost-of-ownership way. And the perception right away is that either the acquisition costs or the total cost of ownership on Linux would be lower than Windows. But certainly total cost of ownership is one of the considerations.

Developer productivity. You know, over the last few years as IT budgets have been very tight, application backlogs have grown. Customers have a backlog of projects and applications that probably is the largest backlog they've had perhaps in the last decade. And it's because they've tightened up a lot of that spending over the last three to four years. So developer productivity becomes very, very important. Interoperability—looking at a platform that interoperates with existing systems. Certainly security is top of mind for customers, top of mind in the industry. And more and more I'm hearing from customers the consideration around indemnification. And this really started a bit over a year ago when Gartner published the advisement that customers really think hard about putting mission-critical workloads or critical workloads on Linux due to the intellectual property concerns that SCO had raised with IBM, and I hear this more and more from customers in roundtables. So when you look at those sets of considerations, let's start with total cost of ownership.

Now, over a year ago there was a study done by IDC that fundamentally benchmarked total cost of ownership on five different workloads. They looked at it over a five-year period, and found that Windows was between 11 and 22 percent lower total cost of ownership than Linux. Now, right away everyone said, "Well, Microsoft, you commissioned that study." That's a fact: We commissioned that study. But it's also a fact that IDC and these analyst firms are reputable third-party objective analyst firms that based their report on the facts.

Now, since then 16 studies have been published. And in fact the last five have been noncommissioned. These are just independent studies done by these analyst firms. Take the one around Forrester. You know, Forrester shows that their results said that Linux is 5 to 20 percent more expensive than Windows. They supported the fact that Windows is lower total cost of ownership. The Yankee Group did three waves of surveys. They found that Linux was three to four times as expensive when they did those surveys of real customers with real data. Gartner published their report around their desktop study. Their desktop study showed it would be significant incremental expense, higher total cost of ownership of Linux versus Windows. This is all fact-based research. There's methodology, data behind this.

Now, when you get into a real customer discussion, this becomes very real. Let me give you an example: Grand Expeditions. Grand Expeditions is a luxury travel company. For those of you who perhaps have gone on a sailing trip with a company called The Moorings—that's part of Grand Expeditions. They employ about 900 people. You know, it's a midmarket type of a firm. And after September 11th, the travel industry went through a huge financial challenge with much, much less travel. So in that period, Grand Expeditions went down a path to deploy Linux, because they were trying to reduce costs, and they thought because Linux had an acquisition cost that appeared to them to be free or lowered total cost of ownership, they deployed Linux. What they found was over time it became more and more expensive. They could not maintain it, they couldn't keep up with the reliability needs that they had. It was causing their business pain. So just in the last six months their CEO/COO went to one of our partners and said, "We need help." They benchmarked with that partner Windows Server and our Windows platform relative to Linux. They made the decision. It was deployed within 60 days. They now have increased reliability. They are getting the benefits of lower total cost of ownership. In fact, in just a short period of time their chief operating officer has quantified at least $200,000 of savings. And they've dramatically improved their reliability and ease of use.

This is the kind of discussion that happens every single day around the world. Our field sales organization, our partners, understanding the customer needs, what the customer is trying to accomplish, painting a compelling value proposition that Windows has a lower total cost of ownership, differentiating how we deliver that from Linux, because we have the systems management, because we have the integrated innovation that enables us to help customers run their IT infrastructure more efficiently and more effectively, and supporting it with evidence, third-party evidence and customer evidence. Now, this happens in small customers, mid-market customers and large enterprise customers. And just to give you a sense of the types of customers that have gone through this exercise, looked at the facts and done the analytical comparison that showed Windows is a better value proposition, let me just show you a little video montage of some of these customers we worked with over the last year.

(Video segment.)

So we're taking this approach of really focusing on the customer, value proposition differentiated from Linux open source supported by evidence. Now, if you look at our Get the Facts Web site, we post all of those analysts reports, and we have over 75 customers who have either evaluated Linux or have used Linux and done their own analysis and come to the same conclusion that Linux is lower total cost of ownership. Those are 75 customers that are willing to be public, willing to talk to other customers, willing to share their experience and the facts around the business decisions that they've had to make.

Now, you may say, "Okay, Kevin, that's great. We understand the total cost of ownership, but there are other criteria that customers make these decisions on." I agree. Let me take you through some other scenarios.

This example is an example with a company called Analog Devices. They are a signal processing chip manufacturer that employs 8,000 people worldwide, with revenues of $2.5 billion. It is a company like many customers that over the last several years has had to tighten up their IT budgets, that has left a number of key applications in their backlog, key applications that deliver them business value. It's also a customer that had a number of systems on UNIX that they said we need to migrate off this proprietary UNIX platform onto something that gives us a better cost benefit.

Our approach here, really focusing on first of all the total cost of ownership facts, which Analog Devices understood and agreed with. But then it was the other criteria: developer productivity. How could we help this customer reduce that application backlog, deliver applications faster, and deliver that business value? So we pointed to some of the other evidence here, the Giga Study. Giga benchmarked building an internal application portal using .NET on the Microsoft platform as compared to J2EE on Linux. And they found that building this on .NET on the Windows platform was between 25 and 28 percent lower total cost of ownership, and they could build that application faster with much, much less developer resources.

The other question Analog Devices had was around interoperability. They knew that these systems had to interoperate with their other applications. Recent analyst report by Jupiter Research: They researched hundreds of IT professionals, really doing the survey to find out who had which vendor, which platform, had number one support relative to interoperability, interoperability whether it's network interoperability or applications interoperability. The Windows platform and Microsoft came out number one in that study and that survey: 72 percent of IT professionals surveyed rated Microsoft as the best in interoperability.

Now, we proved this at Analog Devices. We prove this by enabling a product analysis research process application that helped save this customer $1.5 million. They were able to avoid a $150,000 one-time cost for some of the UNIX maintenance on UNIX systems they had to do. And they were able to build and enable these applications in less than six months. So again it's another customer example. Total cost of ownership was better on the Microsoft platform. The developer productivity was better with .NET and the Windows platform. The interoperability was better. They went down this path and they enabled the solution. And you can see the quote from the customer here. It was about building these applications on the Microsoft platform, because it would provide rapid development, lower cost, and easier integration with their other applications. Total cost of ownership, developer productivity, and interoperability.

Now, many customers in this analysis say, "Okay, we understand total cost of ownership, but we're really trying to focus on how we save some short-term acquisition costs.” Another piece of factual evidence I'll present to you: a study done by BearingPoint, which wanted to go out and look just at acquisition costs, fundamentally said let's compare in some different scenarios—enterprise, midmarket, small business—a few different scenarios on server acquisition.

So in this particular scenario they looked at Red Hat with full support—and full support, by the way, included all their servers and the 24-by-7 needs that the customers have. They then also benchmarked that with the Red Hat distribution with limited support. Limited support was only support on 10 percent of their servers and install base, and it was only 8:00 A.M. to 5:00 P.M., Monday through Friday support. They then compared that with the cost of those same customer scenarios licensing Windows Server and our premier support offering, so that they had full business support around the clock and the acquisition costs of the licenses. They used Software Advantage and the value proposition we have in Software Advantage. They spread that out over a five-year period. Not only is there lower acquisition cost in year one, but throughout that five-year period that particular study showed a 76 percent difference just in the acquisition cost of licensing and the services support when comparing Windows with support to the Red Hat distribution mentioned above.

Now, how does this play out? Another customer, ADC, which is a broadband access provider, 5,500 employees. They were a customer running on Windows 98 and Windows 2000. They were making the decision, should they move to Windows XP and upgrade, or should they consider Linux at that point in time? They evaluated the switch to Linux relative to Windows XP. They did their own internal TCO analysis. They used the Gartner methodology. They applied consultants to go through and do the benchmarking. And they found the Microsoft solution was 78 percent lower total cost of ownership. It would have cost them something like $3.9 million to make this transition to Linux with the same functionality and the same feature set that they were getting with Windows at a cost of only $900,000. And with that switch to Linux they would have lost some of the functionality and capabilities they needed on the desktop.

So once again Linux is not lower cost to acquire. When you look at the commercialization of Linux, certainly when you do the benchmarks as BearingPoint did, it shows that Windows provides the lowest acquisition costs in these scenarios, the lowest total cost of ownership as well.

Two other areas we hear from customers. Security: Certainly the question customers raise—”Well, is Linux more secure?” A couple of pieces of data I'll share with you. Internet Security Systems recently did a study looking at vulnerabilities—vulnerabilities on the Linux platform and vulnerabilities on the Windows platform. Their finding was that the Linux vulnerabilities are growing at a 21 percent faster rate than the vulnerabilities on the Microsoft platform.

You can then also compare that to a Forrester report entitled, "Is Windows More Secure?" They went and looked at a few different approaches to benchmark this. They looked at something called “all days of risk”: the number of days between when a vulnerability is made public in some forum and when a software update is available. Certainly in this particular study it showed that Windows provides the lowest days of risk as compared to these other distributions of Linux. They went further. They looked at the number of high security flaws. And for high security flaws they used a definition provided by the National Institute for Standards in Technology, an ICAP project that was done that defined high security flaws. They found once again that Microsoft and Windows was lower than these other distributions of Linux.

Now, the discussion we have with customers starts with our focus on security mobilization. You heard some of the discussion earlier this morning that Will shared around the work we're doing, whether it's in the area of isolation and resiliency, the work we're doing to automate patch or software updates, and the work we're doing to reach out and provide more prescriptive guidance. Customers are responding to that. But what's also important is when they look at the facts, the facts provided by these third parties and these independent studies—it supports not only are we engaged, helping make a difference on security, but when you compare that to the current situation relative to Linux, from a customer view there is a Microsoft advantage and a Microsoft differentiation that has value.

The other area that surprisingly now comes up more and more in customer discussions is indemnification. Customers realize that when they license software they're looking for two things, not only the license to the code to install and run that, but also the license to use the intellectual property, whether it's copyrights and patents. And it's come up, starting with the Gartner article that was published about a year ago, flagging this for customers.

So in roundtables and customer discussions it comes up often, and clearly from a Microsoft standpoint customers would sit and draw this little grid that you see on the screen here. They would look and say, "Okay, is there indemnification for patent claims? Is there indemnification for copyright claims?" And "Is the company going to stand behind and pay for any legal fees and damages that may incur?" Clearly in our volume licensing, the work we've done around indemnification has proved us to be really standing behind our patents, copyright claims, and the legal fees required to cover those things.

Customers then build the model and they start looking at different distributions of Linux, or they start looking and saying, "Well, Hewlett-Packard has announced some limited copyright claims relative to the SCO litigation, and Red Hat has some things on legal damages capped." And many customers come back with a question mark when asked, you know, what is their relationship with IBM. You know, many customers asking the question, "Where is the distribution coming?" And in a big project being led by IBM Global Services, what is IBM's commitment to them? And it is a big question mark.

You know, so clearly when we look at a set of topics, whether you go around and say total cost of ownership, acquisition costs, security, developer productivity, and indemnification, our field teams and partners were very focused on ensuring that we are articulating the value that we deliver to customers, how that's differentiated from Linux and supported by fact-based evidence.

Now, one of the examples that I'll share with you that this whole set of sequence and discussions unfolded was that a large national retailer here in the United States—this is a large national retailer with 3,400 stores throughout the United States, a large national retailer that's had a very long and deep relationship with IBM, they were looking at their next-generation store systems for their core store systems. And this is something in our competitive situations we really study in the areas where we win and where we don't win, how this discussion unfolds. In this particular case the proposal was from IBM partnered with Novell, and a Linux proposal that they gave to the customer that they said would provide the customer lower total cost of ownership. The Microsoft team and our partners went in and we showed the facts. The customer benchmarked it and said, "Microsoft platform has lower total cost of ownership." Then the discussion shifted to security. The competition may say, "Well, you know, you should look at the security related to the Microsoft platform." And clearly the customer saw the work we're doing on security mobilization, the work we've been doing to reach out to them, to help automate software updates, provide descriptive guidance. The work that we've done has made a significant difference for them. When they looked at the evidence done by some of these other third-party studies, the Forrester report specifically, they were convinced that actually the Microsoft platform had a better security value proposition to them than Linux.

Then the discussion shifted into a question around intellectual property, and so there was a question the customer asked IBM relative to their commitment to indemnify for concerns that the customer had related to IP. I'm not sure exactly how that discussion went or what came out there, but our team was focused on the fourth element of this, which was really the business application. This was the customer that was trying to improve their retail operations where it counted the most, in front of the customer.

So the work that we had been doing around our Smarter Retailing Initiative, and the set of partners that we had and the other customers that we focused on this with, whether it was 7-Eleven or RadioShack and the work we did with them, we had a clear value proposition that went far beyond just total cost of ownership, far beyond security, far beyond developer productivity. We had a set of business solutions that helped them get closer to their customer, help them improve their supply chain, help them improve their operations in store.

So at the end of the day, when the customer added all this up, they made the decision to go with the Microsoft Windows platform. And again this is a customer that had a very long and deep relationship with IBM. It's a customer that appreciated the fact that from a sales and marketing perspective we focused on factual-based evidence that they could use to benchmark their environment and make a very logical business decision. It sort of exemplified the type of work that's happening with hundreds and thousands of customers around the world every day, focusing on a value proposition, differentiated from the competition, supported by evidence.

The final area I want to touch upon is the work we're doing in government and education. You know, certainly you may say they have a different set of needs and requirements than commercial businesses. The work we've done over this last year has been very, very focused on that particular niche, the work we've done to reach out much more to government and education. We've established a worldwide public sector vertical. It's led by Maggie Wilderotter, a very talented executive, with significant field resources that we put into public sector organizations and every subsidiary around the world.

We've really reached out to government to really focus on their priorities, whether it's things such as digital inclusion, education, national security, working to grow the economy in their local country. Examples of the work that we've done through different programs—something called Partners In Learning: We've signed over 67 Partner In Learning agreements in 67 countries around the world where we are partnering with government to help bring technology more to education. It's not just software grants that we're making. We're partnering with local business to help through something called Fresh Start to recycle PCs that we can get into education, combined with software grants, and we're training teachers and helping teachers better use these technologies so that we can help these countries bring more technology into their educational systems.

Work that we're doing around topics such as digital inclusion. Will Poole mentioned the two pilots that we're doing relative to Windows Starter Edition, and the work that we're doing reaching out to government policy-setters and government leaders on the set of social issues that they're trying to address, such as digital inclusion.

On national security, we put together something called the Government Security Program, which is now signed with 34 different countries. We provide them access to our Windows source code, tools to read that source code, and things that directly link into our Microsoft security specialists so that we can help governments be more secure, and that they can feel more comfortable relative to critical national infrastructure that's important to them.

You know, other things around economic opportunity. Economic opportunity, whether it's the work we're doing in China to partner with the government to help build a stronger local software economy, the work that we're doing in countries around the world really focusing not only on the technology industry in that country but how technology helps support other key industries in those countries.

You know, clearly Will touched upon the opportunity in emerging markets. A recent study that was performed showed that over 50 percent of the growth in the install base over the next several years will be in these emerging markets. So the work that we're doing not only in China, India, Russia, Brazil, but we've basically taken approximately 150 countries defined by the World Bank as developing and emerging markets, we built a set of taxonomies around those, and we have a program and a set of teams that are reaching to each of those countries to really focus on how we can help governments, whether it's in the area of e-government applications, digital inclusion, education, security, or economic opportunity.

Now, finally I'll just close with when we think about competing to win. It is about putting the customer at the center of what we're doing. We listen to customers and their needs, whether it's TCO or security, developer productivity, better interoperability, or their needs around indemnification, and we will continue to focus on the value we deliver, how we differentiate that from competitive alternative—and we support that with evidence, third-party evidence and customer evidence.

Now, on this Get the Facts Web site—I encourage you to visit it—you will see all of the analyst reports that are up there, all the facts that support the evidence I've shared with you today. You will see that there are over 75 customers, specific customer examples where they have gone through these benchmarks, they have looked at the facts, they validated them in their environment. They're willing to be references, they're willing to talk to other customers. And we have a pipeline of another 200 customer references that we're working on getting published on that site. We continue to build in terms of the momentum and the focus we have. And, again, we are going to continue to focus on the customer and continue to focus on a very fact-based discussion around value.

So, with that, I look forward to winning lots and lots of customers over the next year. And I encourage you to go look at the Get the Facts site. And I'd certainly be very open to any feedback or thoughts you have for me. My e-mail alias is kevinjo. If you have other ideas that can help us win customers, I'd love to hear from you. Thank you very much.

(Applause.)

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