Speech Transcript – Paul Flessner, Microsoft Worldwide Partner Conference 2003

Remarks by Paul Flessner
Senior Vice President, Server Platform Division, Microsoft Corporation
Microsoft Worldwide Partner Conference 2003
New Orleans, Louisiana
October 10, 2003

PAUL FLESSNER: Well, good morning. Thanks for coming out this morning. We really do appreciate you coming out from the various businesses around the world. We know it’s a big time commitment away from your business and family, and we very much appreciate it.

I hope you believe me when I say our business model depends on you. We really are a software company, have some solutions, but you provide a lot of the solution that makes it right for the customer in the end. And without all of you, our end-value proposition to the customer isn’t there.

I also know that you depend on us, that your businesses don’t work unless we’re successful. And so we take that very seriously, and we are working very hard to make sure that we are successful. And honestly that’s what a lot of what this week is about, about helping to educate you on what we are doing, so your business can better fit in. So we very much appreciate your attendance, and hope it’s a productive conference for you.

Servers. The last 10 or 15 years — I’ve been doing this for about 23 years — the last 10 or 15 have been a wild ride, right? We had the ’90s — a period of historic economic expansion around the world, fueled at the end of the decade by a dot-com boom and also the Y2K, just this prolific growth in businesses, customers worried about merger and acquisition. All they wanted to do is talk about scalability, scalability, scalability. And, boy, we worked on that, and we worked hard on it. I remember lots of conversations about role-level locking and things we were doing for big SMP to scale up. The customers really wanted to talk about scalability.

But toward the, I don’t know, ’96, ’97 timeframe, it was pretty clear to a lot of us at Microsoft that eventually we would get there. We couldn’t ease up. We had a lot more work to do to finally get to the highest point of scalability in terms of our competitors and what our customers needed. But there really wasn’t going to be a way to differentiate around scalability. So we started working hard on total cost of ownership, ease of use, integration. Certainly I’m not here to declare victory on that today. I think there’s a tremendous ways we have to go, and I’m going to talk about that. But we started doing things, like in SQL Server 7.0 where we worked hard on self-tuning, self-optimization, self-healing. I had this motto for the team of no DBA for standard operations. Let the DBA work on higher value application building. We really pounded on that theme.

And honestly there was a period of time in the late ’90s where the dot-com boom was just exploding where I thought maybe I goofed it. I don’t know if you had these — remember in a lot of these conversations where you’re talking to customers —
“I don’t want to hear — I don’t care how much it costs, I just — you know, I’ve got to get the thing in, and I’ve got scalability problems I need to talk to you about.”
You probably remember some of those conversations — I know I did — where the customer calls and says,
“Hey, look, if you give me your software for free and build the application for me, I’ll let you put your name on my Website”
— you remember that business model?

OK, so overall we kind of worked and got through that, and we really did focus and continue to focus on TCO. And honestly I think it’s why our business has done well in the last seven years, when the times were a little tough. You know, as you entered into ’00s, I guess you call this decade — 2001, 2001 — the economy cooled off a bit, kind of the hangover from the dot-com — customers had a lot of software and complexity still remaining in their environment. Our business has actually done quite well.

And I want to throw up a few numbers. So, overall — I don’t like to do a lot of momentum numbers, but I think these are important numbers for all of us to know, because your business certainly depends on the success of our business, so we are going to brag a little bit for a few minutes.

In the world, on the planet, every server, every different operating system — X86, non-X86 — there are about 19.4 million servers today. OK? In the ’01 — come on, Microsoft fiscal year — I apologize, that’s where my numbers are.

So if you think about the second half of ’01 and the first half of ’02, the year-over-year growth rate in the server business actually declined about three points. So we sold fewer servers in that time period than we had in the previous year. But Windows actually accelerated and was positive, up five points in that year. The next fiscal year, fiscal year second half ’02, first half ’03, we saw an acceleration back into the positive of about five points of growth all up, and Windows continued to outgrow the market. We want to put the expense of UNIX, the expense of Novell, who are giving share, and Windows is taking share, and also Linux is in there.

This year we are expecting more continued growth. We think the market is going to stand about 8 points, and we are expecting Windows to continue to take share in that market. So if you think about that 19.4 million servers worldwide, what do you think the market share of Windows is? All right, maybe take a second, put a number in your head. Today Windows server runs on about 62 percent of the world servers. Now, that number is a little higher than the number we normally use. I generally think about — you can talk about more in the 50s, low 50s, but that’s paid versions of Windows — and factor in some of the geographies around the world that aren’t quite there yet in terms of paying on the different business model. If you look at all the servers in the world that run Windows, we think it is in about the 60 percent range. We are going to talk a bit more about some of the work we’re doing and what we are doing with Linux. And, honestly, to our surprise, 2003, very early in its life cycle is starting to take some share back from Linux, and we’re going to talk about that.

And then we’ve got Exchange. Exchange today runs 60 percent — I think enterprise still — we are talking about companies that have greater than 500 PCs, who have a very high market share. There are about 1.2 million Exchange servers on the planet today; 115 million mailboxes, and we have 30 percent growth last year. The enterprise sector of the market — still room to grow, and lots of room to grow in the small-medium sector of the market, and we are focusing hard on that going forward.

And the last of the big three in the server business, SQL Server, which has just been a blow-out success. When I became responsible for the SQL Server business in 1995, it was a $65 million business. Today it’s over $1.5 billion. We sell over 1,000 copies of SQL Server every day, seven days a week, 365 days a year — 37 percent growth in that business last year.

So all up, the big three in terms of our server business, kind of the core of the platform, we really are doing quite well. But there’s much more we can do. There’s more we can do for the customer. The customers are still having a very difficult time overall.

If you look at this data — and I spent 13 years in IT before I came to Microsoft — and all of you work with customers every day — IT is drowning in a wave of complexity — some of that complexity introduced by us, security issues that we have had, past management — all of that. But however you look at it, I see business cannot continue to grow every year. And customers — if we’re going to enable customers to continue to buy new solutions for their business, we are simply going to have to do something other than continue to ask them for more money. Seventy percent — I don’t know, give or take — 60 — I’ve seen some numbers at 60, some numbers as high as 80 — but somewhere in that range, 60 to 80 percent of the IT budget is consumed by ongoing costs of operations. OK? Only leaving kind of 30 to 40 percent available for new opportunities.

What could we do to help shift that ongoing cost of operations down? You know, together we can target specific areas that we could work with our customers on to reduce the ongoing cost of operations. That’s an incredible source of opportunity. That money that they would otherwise be spending on ongoing costs and costs of operations could be plowed back into new system opportunities and new opportunities for their business.

As a CIO I’ve learned that one of the most unpopular things that you can do in a company is continually say no.
“No, we can’t do that, because we don’t have enough money.” “No, we can’t do this new system. We can’t expand.”
It’s an unpopular thing. CIOs are constantly measured on what are they doing to increase the efficiency of their operation? What are they doing to lower their total cost of actual operations? It’s not total cost of acquisition — cost of acquisition — that’s only about 5 to 10 percent of the total cost of any system over its lifetime. It’s TCO, total cost of ownership over the lifetime of that technology.

So let’s look at the manual reality. You know, what is it that is consuming all of that? This slide is probably the biggest opportunity slide we have got in the deck, right? Look at that staff cost. You know, what could we do by working together to bring down that staff cost? Certainly downtime is a biggie. But this overall it is just dwarfed by overall staff costs. The quality of our software, better security, better integration, overall total cost of ownership — those are the things we want to help work out over the overall ecosystem in terms of costs.

Now, many of you are in the server business, and I’m not here trying to put you out of business. We are trying to shift services I think to higher-value things, but let’s get it out of the overall just ongoing cost of operations. And to put a kind of finer point on this, look at this next slide — security management — 62 percent manual, all the way down. Every one of these, always over 50 percent of the cost is in people cost. Right? Now, the security one — you know, we are all over, and Steve talked a bunch about that yesterday. It’s really, you know — certainly we have to do better in terms of securing our products. We have to do better in terms of alerting customers when there’s a problem. We have to do better in terms of patch management, which is what sucked up a lot of that. And all those things are under way. But that’s not the only opportunity. Look at old storage down there. I forget — I saw a number the other day, $12 billion a year just in the U.S. is spent every year recovering files that were lost that you just couldn’t get access to, that sort of thing. Changing configurations. Come on. I mean, we can’t make software learn to tune itself? It’s something that’s a huge opportunity for us. So this is an area of tremendous focus for us going forward.

I want you to see a couple or three things on this slide. You’re a platform vendor in the server business. OK? We want to provide a great platform for Great Plains and thousands and thousands of other solutions in the market. It takes a lot of software to cover the platform market. And I — if you look at this picture down at the bottom — Windows server — we believe it takes an operating system. And an operating system is certainly more than just memory management, CPU and disks. You know, we believe you should be able to program an operating system, and you should be able to run applications. That’s why we have always had our application server built into the operating system for no extra cost. We believe you should be able to enumerate resources and find them and manage them, which is why we have a directory in the operating system for no extra cost. And you can kind of read down that list and go on and on. An operating system is more than just memory management and CPU. OK?

We also have to continue to add value above that. You have to manage operations infrastructure, you have to help the people in the data center manage the infrastructure, help them secure it. You have to provide a great experience for application developers, line-of-business applications, SQL Server — all the e-business servers — and absolutely, you see that big orange in the back — an integrated set of development tools to support this environment.

And the information worker. Certainly the information worker and line of business — that line blurs over time. But information workers have special needs, and we want to address those as well.

So all of this together is what it takes to be in the platform business. And that green dotted line kind of outlines the brand that we are bringing out to market, and really starting to talk about, as the Windows server system.

And what’s the core value proposition of the Windows server system? Integration. Integration means lower total cost of ownership. When you are faced with a competitor who is pursuing good-enough at a lower price point, which honestly is how Microsoft entered the market with Windows competing against UNIX — you know, get good enough, get in the market. We have a different business model, lower price point — we can get in there. But when you’re faced with that, you have got to innovate, right? If you just sit there and let good enough be good enough, guess what? — that competitor is going to take you. But we also need to innovate. We can innovate around integration.

Tell me, would customers like the ability to do end-to-end tracing across the entire stack of software from the client through the network up to the front end HTML server, into the ad server, back into the database and then back out? Absolutely they would. That’s not an easy feature to clone. What customers like more in terms of the Windows Server 2003 role set up — you want this to be a DHTP server, click here. We ask two questions, and you are done. Of course they would. Would it be better if we could put an intelligent cache on the ed servers so that performance could be, you know, 5,000 times faster, because they can go to memory locally without having to go all the way back to the database to get more data? Of course they would.

So all of these things are areas of innovation that we are going to work at. Right? We can’t stand still. We have to move forward. We compete with Linux, to compete for the customer’s business, in three simple ways: offering a broad platform; making sure we have an integrated experience that lowers total cost of ownership; and we continue always, always, always to push innovation, because we have to draw a higher-value proposition into the business.

So this is an important slide for you to understand how we think about the business, how we plan to compete with Linux. The Linux community has proven that they can copy features. Actually they do it incredibly efficiently. They may find themselves in trouble for doing it, but they do it incredibly efficient. They can’t copy this level of integration. They can’t copy the way things have to work together. The ecosystem doesn’t work that way. I’m sure they can do some, but believe me we think this is an area of competitive advantage for us going forward, and we’re going to pursue it aggressively.

It’s not just about integrating in the platform, all right? Windows isn’t the universe. We’ve got a big market share, we have to be responsible and manage that properly, but that’s not it. When you’re sitting there trying to reach out and touch customers in a B-to-B environment, they are on all kinds of platforms. UNIX is still there, Novell unbelievably still there — market share dropping every year, but still three million Novell servers out in the ecosystem. AS-400 mainframes — 20,000 mainframes in the world for the last 20 years not going anywhere. So overall Web services are still incredibly important, and we’ll talk a lot about it. Sanjay is going to come out here in a minute and talk about a developer story, and you’re going to hear Web services a lot. A better standard based on XML that Microsoft has pushed hard into the market to make sure we get great integration. So integration across platforms is something we take very seriously. Yes, Microsoft decided that we run on Windows, and that’s our value proposition. But it’s clear, absolutely clear as an enterprise vendor, we have to interoperate. And there’s a better way to interoperate than traditional EAI methods, and that’s through Web services.

So how is Windows doing in the market? That’s a question I get asked a lot. You know, how is Windows Server 2003 doing compared to others? Actually it’s doing incredibly well. What you have to measure in these things is the same time period from the previous release. Over the last 90 days, sales of Windows Server 2003 are triple what they were of Windows 2000 as compared to the mix of all Windows server sales. So it’s doing quite well. Web servers, we are getting a lot of work — or, excuse me, customer upgrades in the Web server space. That’s the specific area where we have been really very competitive against Linux.

Apache had the ability to run many, many sites per server. IAS didn’t have that capability until this release. We’ve got new capability, with better process isolation, improved performance, better security. We are getting an incredible uptake in that sector of the market.

File services — a lot of server consolidation, remember, out of that nine and a half, ten million server units we have in the market, lots of those are NT 4s; still about 50 percent, 40 percent of the overall install base is still based on NT 4, and a lot of those customers are upgrading. So there’s a huge amount of momentum in this product, and we did work hard to promote it, aggressive betas — you know, talked about 2.4 million evaluation packages sent out. So the momentum of this product is good and we expect it to continue good through the year.

We also know that we can’t wait to continue to add value till the next release. It’s something that, you know, as you get a larger and larger install base it takes more time to make sure you get the right level of quality assurance for the next big release. So we are going to work hard to make sure that we continually add value through the life cycle. Group policy management — I don’t know, this will be very hard to set up. It’s not a new feature. But it was very difficult to set up — much easier. Automated deployment services — substantiating a complete clean copy of Windows in any server in the network in a few, 10 to 15 kind of minutes. The resource manager — your ability to isolate and prioritize IO and CPUs, so that you can do better priority-setting on applications that run on the same machine.

And you can go down the list there — lots of value coming into this server as its life cycle progresses over the next couple of years. Information worker, SharePoint Services, another one of those innovations that we have to continue to prove around file services. It doesn’t take much to clone a file server today, but if you add more functionality, more collaboration, more integration with Office, better work that we can do in terms of just making sure that people can search and find and collaborate. You are going to see us add value. And then application infrastructure. So the point of this slide is that there’s more, more, more to come.

Now, this is an interesting slide, something that we should all be incredibly proud of. I talked about scalability, right? I was kind of the guy that got beat up in these days, all the — Oracle was always hammering this idea to a lesser extent. But this is a slide of a benchmark, standard database benchmark. I don’t know how many of you are aware of this TPC, Transaction Processing Council. It’s a TPC — www.tpc.org. It’s an organization that’s independent — we belong to it, hardware vendors belong to it — Oracle, IBM — and we do these standardized benchmarks, and we take a sample application that is incredibly well-defined — it’s an order-entry application — and we run that application on our software and hardware, and we measure through-put. And then an auditor from the TPC comes and audits it, and then it gets published.

So back in ’96 we were doing 3,904 transactions per minute, and feeling pretty darn good about it. But the competition was far ahead. At that point the UNIX systems were considerably ahead. They were in the 64-bit environment, and were able to add lots more memory, and they were outperforming us. And I understand Gord Mangione, who runs the SQL Server team in our organization was here last year, and he made a firm commitment that we would top the best UNIX numbers in this benchmark before the end of the year. And I am happy to report that we were able to succeed in that goal, with a 64-bit Windows and 64-bit SQL Server. So we were able to meet that.

And we haven’t stopped there. We recently just published another record benchmark where we overtook UNIX to be the top-performing SMP benchmark in the world at 786,000 transactions per minute. (Applause.) This is an amazing set of numbers. Thank you.

There’s all kinds of comparisons and that sort of thing. I can’t remember all of them exactly today, but this is like ten times the volume of the stock exchange for a year, and just huge, huge volumes of transactions that can be pushed through. And all this was done at a price point that has continually come down. That 3,900 number was at $110 a transaction. Now we’re at $6 a transaction. So it’s 201 times — not percent, times performance at one-seventeenth the price, and the battle goes on.

Now, I’ve got to be fair. UNIX has recently just popped up above us a little bit in the last couple of weeks. It’s a leapfrog game now. But there’s nowhere for them to run or hide anymore. They’re not going to hide about performance. If they do a big number, we’ll come back and do a big number.

Let me show you the most recent one, which I’m announcing today, NEC Corporation, on a new Itanium. This is a 64-bit machine. But this is only 32 procs. It’s not a world record as high as 77 but this is half as half as many processes as that 786,000, okay? And it’s, what’s, three-quarters of the through-put.

So as you see, if this were a 64-bit machine, we’ll be more tuning and more tuning, and yet we’ll break a million transactions in a minute, and all that stuff will happen. There’s no more place for UNIX to run and hide in terms of scalability. That game is over. This is why it was so important for us five or six years ago to start really trying to differentiate and separate ourselves around total cost of ownership, which is certainly our goal going forward.

And remember we had three pillars we talked about in the Windows Server System: operations infrastructure, line of business, and also the information worker. You just saw a talk from Jeff a few minutes ago, so I am not going to hammer on this too hard, but we certainly have a server component of this, right?

Exchange. If you want to always be online, you want integrated communication, Exchange is an incredible environment.

Outlook — it’s there, you know, we’ll connect to it, and we’ll talk more about some of the Exchange innovations. If you are out browsing and roaming around, don’t have your PC with you, oh well, we’ll keep you connected. If you’re on a mobile device, we’ll keep you connected to your mailbox. If you want to find whether somebody is online, and you want to publish that, live communication server. If you want to right-click on that person, do an IM with them in a secure way, a live communication. It’s all about integrated communications collaboration.

And that’s our whole focus, taking Office, which is kind of an individual productivity worker piece of software in the years past, and making it more a part of the collaborative and integrated communication world.

Exchange Server 2003 is an important release for us — probably one of the big innovations, you know, that we haven’t really trumpeted yet. We realize the Office launch comes up. But in terms of just pure usability, this ability for Outlook to sense and understand the speed of the connection you’ve got and do this what we call off-line cache mode, so that in the background it’s working and trickling mail down all the time, and keeping it locally, so you don’t have this uneven level of performance when you’re jumping back up to the server. So getting the messages down locally where your performance is more consistent, and doing it in a smart way, only bringing headers down when you’re on a 56 pay line, and then trickling the data back in the background — just being smart about how we get the information there.

One of the features that people are so excited about is the ability to have Outlook just connect through the Internet in a secure way without having to go through a VPN. When you punch that VPN hole through your firewall, that’s an exposure. Honestly, that was one of the biggest way that the Slammer virus propagated was through VPN connections — an unknowing developer running MSDE at home, punches a VPN hole in to do mail, and got MSDE running — bam, they’re infected. So not having to have a VPN connection, being able to do mail on the Internet.

OLA — OLA is amazing now. Looks just like Outlook — much richer experience. We did a lot of work with the Windows team and the Office team to do a joint release criteria as to higher level of quality — much easier to migrate into Exchange 2003 from previous versions. The active directory upgrade is much easier to do. Better mailbox level restore and a whole host of features around this product that I think will give a superior experience.

The big opportunity that all of us need to be thinking about is that 40 to 50 percent of those million-point-two servers out there are still on Exchange 5.5 on NT 4, and that’s a big upgrade opportunity for all of us. One of the big things that the customers are saying to us is that we need the rest of the ecosystem to get on 2003, because we need to round out the back-up solutions, the virus, the spamming.

Exchange 2003 also has much more integrated spam, both at the Outlook site and Exchange, and a better API for AV vendors to plug into. So there’s a huge amount of opportunities for you here.

The line-of-business stack, the application stack, if you will. This is about two audiences: developers and the IT pro, right? You have got to take care of both. The develop ecosystems and the tool set that we have got around it are phenomenal. Over 50 percent of developers in the world when they come to work, work in Visual Studio every day. That’s an incredible asset for our platform.

SQL Server, by far and away the highest market share product in the Windows environment in the world today. Unit share — revenues in the 40s, high 40s; unit share would be much, much bigger than that. So we really feel great about our position in this space.

Jupiter. We talked about Jupiter before. It’s the combination of our e-business products that we are bringing forth in a new release of it, and I’ll talk more about that on the roadmap in a few minutes. But what we have found is when people are trying to put together a portal and were trying to do aggregation that they were always integrating our products together, so we probably take that and reduce the total cost of ownership of that solution to the customer and integrate that ourselves.

It’s better depicted on this slide. A lot of our customers do in their daily businesses — try to unlock all the information that’s in the back, right? — whether it’s in a database, whether it’s actually in a system, an application — so trying to get it and project it out, whether they’re projecting to employees, the partners or the customers, or trying to get that information out there.

And customers don’t want to have three different infrastructures — one for their employee portal, one for their partner — you know, for B2B, and one for their customers. They want one integrated experience. They want to be able to pick and choose from a set of services, whether it’s EAI, whether it’s work flow, whether it’s authentication. And they want to be able to install the pieces that they want. The whole vision around Jupiter is to stop selling individual products, Content Management Server, Biztalk Server, Commerce Server — we’ll do that integration for the customer. Let the customers pick which components they want to install, and be able to provide a common infrastructure for the employee, partner and customer experience.

We’re also including some components out of SBS, so the customer will get a full portal and collaboration and integration experience with this product as we move forward. But, again, all in the eyes of trying to reduce time to market for the customer, and reduce total cost of ownership.

BizTalk Server 2004, soon to be released, beta today. A lot of what BizTalk is about is, again, about getting connected, right? Business to business, getting the folks if you will, the small-medium business, put to the hub, and trying to automate that process. Biz Talk has built-in messaging. It has built-in EAI and it has built-in orchestration to make all that easier — easy.

One of the things that developers talked a lot about, what — it’s more difficult to use and develop on than I want. So we did an integration experience with Visual Studio, so you can do all the work you need to do to set this thing up in the IDE.

How about using Office for real-time monitoring, business activity management? That’s the integration that we’ve done in the next release, so that you can fit and understand — think about it — if you could just automatically monitor and manage all of your business activities and flow with your training partners and see it all pop up inside Word or Excel.
“Oh, we had five vendors today, or five customers that purchased over a million dollars — here are their names. Hey, I would like to send them a customized e-mail to thank them for their business.”
All that’s possible going forward.

And of course making sure it’s integrated with Web services, so you can legacy applications, speaking legacy APIs, and project them through the customer, utilizing the most up-to-date Web services standards.

Business intelligence. This is a hot area three to five years. It’s really been hot over the last three to five years. People have so much data, and they are trying to continually arm it for information, whether it’s leads, whether it’s ability to just do better selling or customer service. It’s a huge area of growth and opportunity in the market.

A lot of people don’t even know we’re in the business. I guess shame on us. But back in SQL Server 7.0 we decided to integrate our OLTP, our relational DBMS, with our analysis server, our OLAP server. And we put it in the same box, no extra cost. We put in data transformation services, so you can scrub data from other sources — Oracle, ERP systems — and pull it into the database for reporting and analysis. And with the next release, the next slide I’m going to talk about, we are incorporating reporting services. So SQL Server is more than a relational database. It is a complete data platform, and we’re projecting it as such, and we are going to talk much, much more loudly going forward about our BI solutions that are integrated in, as well as the full set of management tools to set up, manage and build a data warehouse.

All you need out front — sure, you can buy other tools that have more sophisticated analysis, but the Office team is really engaged going forward for Office 12 to become a great landing pad. Their XML design service really — or management and UI service, editing service — really gives us an advantage in projecting BI data out to them so that we can do more powerful analysis with Excel, with the Office Web component, things like MapPoint, and certainly projecting it out on shared — like SharePoint Portal Service Servers.

MBS: This is an incredible opportunity for the MBS partners. All of the data that is stored in SQL Server for the line-of-business operations– there’s tremendous opportunities to leverage these technologies, to do customer recording and analysis for your individual customers in your geographies and their given industries to make the overall MBS experience more tailored to what the customer wants. If you haven’t had much chance to try this technology, I really encourage you to do so.

Reporting services is a very hot feature. We had this feature targeted for Yukon, but it was one of these deals where you put it out in beta, and customers were like — you know you had a hit, right? Customers would come back —
“I need this. I need it today.” “This is the most exciting thing I’ve seen coming out of the BI space in a while.”
You know,
“How can I get my hands on this sooner?”
It’s honestly a pretty simple piece of technology, and it solves a very critical problem. We’re going to demo it here for you in a minute. But think of how many times you extract data — you run this big, long complicated query in your database, you extract the data, you have it in this row set or data set — what do you do with it? How do you format it? How do you project it? What if I have a report that I want to go to lots and lots of people? Do I have to print it out and mail it out? You know, you want to actually project it up on a Website or put it inside Office. You want to mail it around. You want people to be able to drill in and do further analysis on it without having to go back and run that more expensive query. All that would be just another feature of SQL Server. And rather waiting for Yukon, which is about a year away, we are going to put it right into SQL Server 2000 and make it available to all customers for no incremental cost. This is a hot feature. If you haven’t tried it, it’s something — if you’re a SQL Server buff, this is something you need to dig into. There’s over 16 ISVs already committed to embedding this technology into their products, and we are working closely with them every day. The momentum on this thing is phenomenal. I think probably by the end of this year I will release this into the market.

Yukon: Big, big release, been in development for over three years. And we’re probably — we’re now in beta, and probably got about another year to get fully through the quality cycle to make sure we have got the right provide in the market. Huge advances in terms of the developer experience — much tighter integration with the IDE in Visual Studio; incorporating the common language run time, so that you can right your procedures in any language that the run time supports; HTTP connectivity so you can make SQL Server kind of a first-class Web services citizen. Lots and lots of activity in this space.

In terms of Enterprise data management, a lot of more availability features and great TCO — more self-tuning around the optimizer, continuing to reduce configuration parameters; data e-range positioning online index bills. And something that I think is going to be a breakthrough for serviceability — one of the big problems today we have in People Server is when something starts to go wrong, whether it’s a long query or some kind of problem in an application, see what is going on inside SQL Server. And with this new innovation around, called Virtual Cables, you are going to be able to go in, and we’ll reserve resources and connection space for you to make a connection and query these virtual cables, so you can see what the server is actually running at the time things had slowed down, so you can do more diagnostics and analysis to see what’s going on in the server as it runs.

Business intelligence we talked about — lots more marketing push, lots more visibility. This is a huge area of growth and revenue opportunity for all of us, and I think reporting services will spur that on. And always, always, always, security and quality — you’re nothing — nothing in the database environment unless you’re rock-solid reliable. People Server has got a good reputation. We had a bit of a tarnish with Slammer. Again, you know, it’s just this whole maturity and growing up and understanding how the world works today around security vulnerability on the Internet. You know, nothing that we even thought about several years ago, but it’s something you have to think deep about now in terms of exposing ports for other purposes. So really locking the product down and make sure it’s a robust and rock-solid product.

Well, boy, have I been talking a lot, huh? Lots of different features, covered a lot of ground. You know, we talked a lot about integration. What I’d like to do now is bring out Jon Rauschenberger from Clarity Consulting, a partner in the Chicago area, to have him give us a demo of all of this technology, or a lot of this technology coming together to solve some problems for its customers. Jon, how are you today?


PAUL FLESSNER: What are you going to show us?

JON RAUSCHENBERGER: What I’d like to spend a little time talking about this morning is how Clarity is leveraging the integrated platform you guys are providing to deliver solutions to our customers at the lowest possible cost. A lot of our customers fall into that small and medium business space, and what we are finding more and more is that when we go in to talk to them the number one thing we’re competing against is their budget. Can we deliver an e-business solution to them? Do they have budgets to actually afford our services and the products they’ll need to actually implement them?

So what we have done to demonstrate that we can deliver solutions like that on Microsoft’s integrated platform is to put together a demo that I want to walk through today. Now, the demo is built around one of our favorite fictitious companies, Kontoso. For this morning’s demo, Kontoso is providing e-business solutions to health care providers — so doctor’s offices, physician’s offices, that want to track electronically the services they’re providing, the claims they’ve submitted for payment and the payment they’ve received back.

So the scenario starts with the physician doing an exam on a customer. At that point we collect some information from them. That gets sent into the system, and a claim is generated from that service that was performed. That claim needs to be reviewed by the claim administrator, some additional information provided. The claim is then submitted back to Kontoso for payment, where it is going to run through their business processes to validate that it is a claim that they can indeed process. Once that’s been completed — so once we’ve gone through Kontoso’s back-end business processes, they will forward the claim onto the insurance companies, via EDI, or whatever other mechanism the insurance company is accepting claims over. At some point in the future those insurance companies will then send payments back through Kontoso that will ultimately flow all the way back to the customer.

And, as you can see, we have used a variety of Microsoft server products, all included in that integrated platform to build this solution. The last piece we needed to deliver — we use the feature that Paul just talked about — so we need to provide a rich reporting experience for the customers. They need to be able to look at these reports online, work with them offline in Office. So we leveraged SQL reporting services to implement that functionality.

Now, for the whole solution we elected to build it using Small Business Server Premium Edition, which gave us access to all of the products you just saw throughout the entire scenario, and enabled us to build a solution on that platform that we can deliver out to our customers.

So let’s go ahead and take a look at the demo. We’re going to pick up the demo scenario after the physician has completed the exam, and the claim administrator now needs to go in, review the services that were performed, and send that claim off to payment. And what you are seeing here is the intranet site that is part of the Kontoso solution. This is actually a BizTalk server business services activities site that integrates directly within the Windows SharePoint Services; so a very familiar interface for the customers that are using Windows SharePoint Services.

I can go in here, for example, and click on
“patient claims,”
and what I’ll see is a list of the outstanding claims that I as the administrator need to review and submit for payment. Now, the claims are stored as an XML document, and Windows SharePoint Services does a great job of document management.

So what we had the challenge of doing is giving the claim administrator a form they can use to review the information, fill out missing pieces, and then actually submit it back. And this is one of the areas where traditionally we probably would have had to build a custom solution, so use Visual Studio .NET to write a custom application.

In this scenario though we elected to use InfoPath. So when I click on that link, the Windows SharePoint Services knows that this XML document is actually going to be manipulated using an InfoPath form.

When I click on the link there are some interesting things happening. InfoPath launched automatically. The form that we built to deploy with the solution was downloaded into InfoPath, and the document that I want to manipulate was loaded into that form. So all the deployment headings are taken care for us. The form is auto-deployed when you launch InfoPath.

Now, here we can see the claim form. We can scroll down and take a look at the services that were performed. We have a total charge of $215. We have some missing information that I need to fill in, the insurance company that the customer wants this particular claim submitted to. We have got two for this customer. We have got Fabricam and Northwind. I’ll select Fabricam. InfoPath then automatically pulls in the associated information that the customer provided for that insurance carrier. We’ve got everything we need. I scroll back down here, and I can see the services that were performed. Again, the total claim is $215. And I can go ahead and submit the claim.

Now, submitting the claim kicks off the next step in the business process, which is actually to post the document out to Kontoso. They are going to run it through their back-end business processes. And then ultimately that claim will be submitted to the insurance companies via EDI for payment. So we can see now that our claim has been submitted. It has been sent off to Kontoso for processing. And this brings us to another interesting challenges we deal with, with small companies which is with a solution like this they now need the ability to track the status of that claim as it goes through that business process that we have implemented using BizTalk server. Again, traditionally we probably would have had to resort to some form of custom solutions, built on some sort of a reporting engine, some sort of data inquiry application, where they can go in and look at that claim to check the status of it: Has Kontoso accepted it? Have we received any payments against it?

By building that business process in BizTalk server, we can leverage the business activity monitoring tools that BizTalk 2004 provides for us. So if I return back to the home page of our intranet site, I can take a look at, for example, the list of customers — or partners, excuse me — that we are exchanging information with in this solution. And here we see Kontoso. I have visibility to the items we’ve sent to them and to the information that they’ve sent back to us, so I can drill in, for example, and take a look at all of the documents we’ve we sent out to Kontoso. And here we see all of the claims, including the one that I just submitted. We can return back and take a look at all the documents that have come back from Kontoso. And here we’ll see the acknowledgements and the payments coming back from Kontoso as that claim goes through the business process.

This is useful information, but realistically for a business user this isn’t all that useful. This is just telling them XML documents are going back and forth. Well, because the process is implemented in BizTalk, BizTalk understands what these documents mean, so I can very easily build a query, for example, that will show the business user the status of all of the claims, based on the XML documents that are going back and forth.

What we see here is again a query integrated directly to BizTalk server, delivered through Windows SharePoint Services, that shows all of the claims that I’ve submitted. It shows the dollar amounts of those claims, when it was submitted and when it was approved by Kontoso. Again, that approval is coming back as an XML document.

We can see here we have a claim that was submitted several days ago that hasn’t yet been approved. So we know we have some sort of a problem. At this point we can pick up the phone, call Kontoso, and start to find out what’s wrong with that claim. Again, though, we didn’t have to build this infrastructure. We got it for free by implementing that business process and using BizTalk server.

Now, one of the challenges of a solution like this is that once we’ve got it up and running, we’ve built it, we’ve written all the code, we have defined all the XML documents and the business arrangements between the partners, we need to deploy this solution. Now, in the past that has always been a painful process. Kontoso has the solution finished; the customer wants to get it up and running. That might take days or weeks. What we can actually do now is for customers running Small Business Server that are using BizTalk server, we can package that entire solution up into a single cab file that has all the custom code that we wrote, it has all the XML schemas and the XML documents, as well as the business relationship that Kontoso and the customer have formed, implemented as an XML document. The customer can take this one cab file, get it onto their small business server, and be up and running in a matter of hours with this type of solution — rather than days.

PAUL FLESSNER: That’s a really key point. I don’t want to go too fast. Lots of small-medium companies live in fear that their big buyer someday is going to come to them and say,
“Hey, look, I mean, you have got to get up online. I am trying to do just-in-time manufacturing”
— and these poor small-medium business guys don’t know what to do. The Small Business Server 2003 we are putting in what we call the partner edition of BizTalk, which is a very lightweight, again easy-to-configure — and if they want to get up and trading — you know, if you are a supplier for Ford or one of the big auto companies, all you need to do is click and accept this cab file, and you can be up and going. I think this is a huge opportunity overall for all of you to essentially plug in. The goal of this technology is to get B2B real. And that’s hopefully by getting the folks active in all the small-medium businesses that need to get connected, we can really get some momentum behind there.

JON RAUSCHENBERGER: And from a partner perspective, our value added is what’s in this cab file. It’s not getting that up and running. We have done all the work. We can now package it up, and we don’t have to waste those precious service hours we have available configuring the servers and doing those sorts of things. We can simplify that process.

Now, the last step of the process that we took a look at in the slide was that reporting interface. Paul talked a little bit about people reporting services. I share the excitement with the product. I think this is just an outstanding product from Microsoft. This enabled us to deliver a very rich reporting interface for the customer without having to go out and purchase a third-party product or build something custom. Again, we are just using what’s in the box.

So what we see here is the home page for the reporting site. I can drill down here and see a list of available reports. Here’s one that looks interesting: I can see claims by payer. When I click on that, the report is going to be rendered directly in the browser, so we are seeing an HTML representation of the form. I can get the report, I can give myself a little more room here. I’ve got full drill-down capability, so I can drill down and take a look at all of the claims we’ve submitted to Fabricam for payment. I can see —

PAUL FLESSNER: This is all queued up in the server. You are not having to go back and touch the database at all. That’s why the performance is super-fast?

JON RAUSCHENBERGER: Yup, yup. This is all cached, and nice and quickly accessible for us. We see here we have a claim that was rejected. I can actually drill through this report into another one, and take a look at the details of that claim. We can see it was rejected because it exceeded the threshold for that particular customer. We can now see all that information and work with it online.

Now, for a report like this one though, we may want to work with it offline. We may want to pull it down into Excel, or e-mail it to someone else. Directly from SQL reporting services, I can export this report, for example, into an Excel spreadsheet, go ahead and open that Excel spreadsheet. And now I’m working with that exact same data, with all of the formatting retained — but now I’m working in Excel, so I can do my own formulas, I can do whatever I want with this data. A very cool feature though is I still have that drill-through capability. So here’s our rejected report. Even from directly within Excel, I can still integrate in and call back out to that detailed report and bring up that claim that was rejected. So I don’t lose any capability when I pull that back into Excel. Very cool feature.

Now, the last thing I want to show you is this is the experience for actually building these reports. So this is a development tool. This is something that you have to do, go in and do some work for, but the development experience is incredibly simple — again allows us to focus on adding value, not necessarily dealing with infrastructure.

So here’s that report we were just looking at, opened up in Visual Studio .NET, with the reporting services integrated. I can, for example, add a chart to this report, simply by dragging and dropping that control onto the form. I select the data source that I want to work with, and now I’m just working with a typical report, Visual Editor. So I can go ahead and select the data that I ant to see on the chart. So we’ll just throw a couple of claim amounts onto these graphs, and then we’ll break it down by month — we’ll just throw month across the top here. Once I’m done getting it looking the way I want it to — so we’ll just put it on the page over here — I can go ahead and deploy that report from within Visual Studio .NET. So I just simply right-click on that report that I was just modifying, select
The integrated reporting engine will go ahead and compile the report, and automatically display it out to the report server that I am connected to. Once that deploy has succeeded, we can return back to the reports page here — let me just collapse things. And assuming everything is going correctly with our demo, I’ll just simply refresh the report, and we will now see the data rendered with that graph I added onto the report automatically included. No deployment. I just simply can do that directly from within Visual Studio .NET. (Applause.)

So, Paul, hopefully what I was able to demonstrate here was our ability not only to deliver a feature-rich end-to-end solution very quickly; but what I think is more important is that the time, those precious service hours that our customers have to spend on our services, are now going towards building functionality that helps their business not having to build infrastructure. That’s the big value added we see from the integrated platform.

PAUL FLESSNER: Looks great, Jon. Thanks a lot. (Applause.)

So kind of the last leg of our journey is the operations infrastructure. And, boy, do we have room for opportunity here, right? Think about what we do in terms of the state of the industry managing our environment today. What do we manage? Do we manage applications or whole systems? No. We manage individual machines largely — oh, the CPU is hot. Oh, the disk is hot. Oh, the network is hot — and we send alerts. People scurry around the data center, going,
“What in the heck is going on? Why is this hot? Why are we getting all this traffic?”
But we know so little today about the application itself. The application doesn’t understand itself. It doesn’t talk to the operating system to let it describe what is going on. We have too many products.

We have, you know, I think, just a huge amount of opportunity in the industry today. Most management products today spend most of their time not managing but trying to find all the different arcane places where we hide the description about the application. Think about the directory — or, excuse me, the registry. Think about all the different private files. Think about all the places where the apps themselves hide things. It’s an incredible discovery process to try to find all this stuff today, and that’s not where the value add is.

You are going to see a lot more about — well, overall, in terms of what we need to do. We need to be able to have the applications understand themselves — you know, how am I supposed to be deployed, how do I need to be operated? I can score that in the application so that that becomes part of the application and travels with the application? I need to be able to tell the operating system, and the operating system needs to be able to listen to how I want to be deployed, so that the operating system can then act on that and do the right instantiation of the right server.

The data center of the future will be a set of machines that literally have no idea, you know, what they’re about. We can dynamically instantiate an instance of an application on a machine, and just as quickly take it off and replace that with another application based on what’s going on in the environment and where we need the CPU power to be applied. In order to do that, applications have to self-describe. Operating systems have to understand and listen. And then the more sophisticated management tools have to monitor that and help us diagnose what’s going on in the end and application — all the way out to the client, so we can do intelligent things and do better monitoring management and patch management of that environment.

We’re calling this the dynamic systems initiative — okay? And I talked about it. You can see it a little bit more graphically. You see the little document flying up. That’s the application describing itself: What cables do I have? What DLLs? Where do I execute? How I’m partitioned — telling the operating system that, and all about itself, as the operating system knows how to deploy the application very quickly, very dynamically, whether you’re trying to instantiate or remove a given image of an application.

Think about big data center people don’t even know where the servers are physically in the environment many times, because they can’t, you know, track it down. It’s a very complex process — automates the operation of it itself. That’s kind of — that’s step three, where if we have all this information, if we know where to go to and discover and be able to understand the schema of an application, we can do some intelligent things about it.

See down on the bottom left of the slide you have this thing called the Systems Definition Mode — SDM. That will be previewed at this year’s PDC in the next couple of weeks — so we are going to talk a lot about it — and the next version of Visual Studio will support this, so that applications can begin to describe themselves. Longhorn will be our first version of our operating system that will be able to acknowledge that, and the Longhorn wave of products will be the first set of operations management products that will be able to benefit from that. It takes time, but it will be a very, very big step forward in terms of total cost of ownership of our platform, and one that I believe will bring competitive advantage in the marketplace. Taking a holistic approach at this is the only way to solve the problems, and I am very, very excited about this work.

Man, oh, man, this has been a ball game, huh? Patch management. We didn’t really — you know, part of the ecosystem for a long time has been a need for this, and it’s kind of shame on us for not getting there more quickly. The Windows Update has been a big breakthrough. I don’t know, a lot of you probably have one and multiple PCs at home — I know I do. In my life as a LAN administrator, if you will, got better at home when I turned all of the machines on to Windows update. You know, keeping the machines up to date, letting them update themselves, has been a big breakthrough. But it’s relatively kind of focused on the consumer market.

We have introduced SUS — Software Update Services — today kind of limited in terms of what it can update. And it is also kind of targeted to the small-medium business. It’s probably one of the best-kept secrets. I bet if I asked you to raise your hand, 40 percent of you probably never heard of SUS. But it is an important set of services. We will have another release of it very shortly, and it will be important in terms of delivering patches if you will out to a server inside the customer’s firewall, where the customer can sort through them, do their testing, and then stage them out to their environment more readily.

And then of course we have SMS. And another release of SMS that promises to be much better integration and much more reliable in terms of knowing what patches have been distributed and applied. And we still have kind of three services in the environment today for doing this.

So next summer, in essence, we will eliminate one of those services — actually two of them — and call — rename one of them Microsoft Update. We’ll have Windows Update, where you will get patches for Windows. And then you’ll have Microsoft Update, where we’ll get patches for all other products for Microsoft. And potentially in the future we are talking about potentially distributing third-party patches there as well in certain circumstances, and that’s something we’re still exploring.

We still have SUS in this timeframe, but much, much broader support for a product line. And of course that’s MS. And then in the Longhorn timeframe the picture gets much cleaner. Many of the characteristics that were in SUS — which is no cost product from us — will be merged directly into Windows, and this picture gets much cleaner. So everybody gets patch management at no cost — the consumer, the corporation, small, medium and large. And if you want more sophisticated types of management and monitoring, you can buy systems to enrich the experience yet further. But a lot of work and a lot of focus on patch management in the coming months, to make sure that we can do better here.

Now, this is something that I get asked for all the time, so here’s the roadmap. Roadmaps within the 24-month timeframe are pretty clear. That three- or four-year timeframe, the horizon starts to come into play and it gets a little foggier, but this is our best attempt at planning right now. You see the first column is ’03 and ’04. So Windows Server 2003 was the big dog in terms of the server environment, and that shipped already. Exchange Server 2003 is shipped, and we talked about that — live communication server, a great RTC platform. You can kind of look down at the host of applications that are assumed to ship in ’03 — actually I can point those out more clearly for you — SharePoint, Project, BizTalk, SMS — they’re all in the late ’03 timeframe, and then the others fall under the ’04 time — Yukon probably the long pull on this slide, in late ’04.

In ’05 we talk a lot about Longhorn, and Will Poole is going to come out in a few minutes and talk a lot more about Longhorn and the client experience. And there will be a wave of products — that’s when the e-business suite gets factored across those services from individual products. Live communication, SharePoint, another release, and there most likely will be another release of Office that will be very close to the Longhorn client timeframe.

And then Longhorn Server kind of swells to 18 months later than the client. It takes a little longer to rev the server, and the right quality assurance — and a lot and a lot of change that we are looking at in terms of the overall management that we just talked about. Another release of SQL Server in that timeframe, and then Systems Center, which was a combination of the MOM and MSN product into a more integrated experience, and many of the features of applications center migrate into Windows. So that’s the kind of all-up roadmap going forward for the server business.

A lot of times, you know, customers come to me and say,
“Hey, look, I love the different options. I love the flexibility of your platform, but man, am I confused. What do I use for what solution? How many different options are there? You know, how can I get some better guidance?”
We have to do a lot better job of cleaning up the taxonomy of our documentation, do a better job of including the customer in the community, and let them provide information, including the partner. We are spending a lot of time and effort getting a much clearer taxonomy for all the Windows server system products in terms of the product content that releases with the product, talks about features, performance tips, configurations; and solution guidance that talks about how to put this stuff together and make a solution like Jon showed us. And then overall learning and training — a lot of investment in pulling all this together and making it more readily available to you and to the customers.

In conclusion, we talked about a lot of things — got a lot of competition in the marketplace, we have got a great position. You’ve heard us talk about our value proposition in the market and how we differentiate ourselves. We are a company that is offering a broad and complete platform. There’s a tremendous amount of value in that. We are a company that is focused on integration, which should result in better TCO for our customers. And we certainly you know continue to talk about innovation. Innovation equals value for the customer, and it really propels and moves us forward in the marketplace. So with that, and with your help, we are ready to help IT succeed, and small, medium and large businesses all around the globe. Thank you for your business and thank you for your time. Have a great day. (Applause.)

Related Posts