Designing XP amid antitrust scrutiny
Jim Allchin, VP, Microsoft Platforms Group
To lend a helping hand, Microsoft is pulling out the stops to the tune of a $1 billion marketing campaign--$200 million of it coming directly from Microsoft--for the product's debut Oct. 25.
But it remains unclear just how much of a fillip Windows XP is going to provide. At the same time, the prospect of an operating system integrated with Microsoft services like Passport and the upcoming HailStorm has fueled complaints from critics that the company is returning to the same kind of behavior that triggered an antitrust lawsuit by the government once before.
Thus it falls to Allchin to present the benefits of Windows XP to the wider world. He recently celebrated a milestone when Microsoft shipped the software to computer manufacturers.
Allchin, who was in town this week to address the Intel Developer Forum, used the occasion to proclaim that Windows XP will indeed go a long way toward pulling the industry out of its slump.
After his address, he sat for an interview and talked in more detail about one of the most closely watched Microsoft projects in years.
Q: How about a general overview of where you are with Windows XP?
A: We released (XP) to manufacturing on Friday and are super-happy about it. It's the best operating system we've ever done...Windows XP is different because we surfaced the technology in a way that people can actually get more value than what they've been able to get out of their machines. And that's what I feel proudest about.
It's a cleaner interface, without a doubt. But what gives you confidence that the installed base will upgrade in any meaningful way? Basing their appraisal on the current macroeconomic climate, analysts are skeptical about it being a world-beater in the immediate future.
Well, I can't do anything about the world economy. The only thing we can do is build what we think is compelling value. And this beats any system we've ever done in terms of value. This is a hands-down winner over Windows 95. Obviously, the world is different today than it was then...But I've been surprised. I thought businesses might be the ones who said, "Ah, I don't know if we should upgrade." It's not what I'm hearing.
What do you envision for XP adoption rates?
OEMs (original equipment manufacturers) are going to switch over full bore--that's where the bulk of our royalty revenue comes from. On the retail side, that's not the huge bulk of our revenue, but we'll see a pop in terms of revenue, and we have some stuff planned that we haven't told people about yet that will also help.
Like what? In terms of marketing?
No, in terms of products that we still have coming. And we're obviously going to spend a lot in marketing because we think the product sells itself, if we can get people to touch it and see the value.
Can you be a little more specific?
Nope. If you mean in terms of marketing, I can be more specific.
When you say product, do you mean related to Windows XP?
Add-ons or some things that we don't know about yet?
Yes, things that you don't now about yet. We haven't disclosed.
And that would be launched with the product?
Are these features of XP?
No. They're additions to--some free, some not free. But the point is to show off some of the capabilities that are in the product. Given the marketing we're going to do, which will be significant, I'm very optimistic.
But to finish up on the point: A lot corporations seem very keen on upgrading?
More so than I expected. I know several that are talking huge orders, hundreds of thousands.
Can you share that info with us?
I don't think any of that is public yet. But you have to think about how many old machines exist. There's at least 140 million that are at least over three years old. My own personal opinion is it's much higher than that. But those are the numbers from IDC. We don't think there's been enough value in the PC and the software to drive upgrade sales.
You were about to offer more specifics about your marketing plans.
We've said we'll spend $200 million over four months on a worldwide basis. We have a program going on with Intel to highlight--if you take the Intel line and Windows XP--the value you get. We have a set of programs with a variety of partners in addition. And of course we're going to promote through non-monetary means, on our own Microsoft.com site and our other properties, the capabilities of XP. So we're certainly going to make a huge push to let people know about the product.
Passport integration and Windows
Can you talk about the integration of Passport-enabled services into XP? It's my understanding that XP users must use Passport to gain the full functionality of the OS. Correct?
Well, let me explain. There is a set of capabilities in the system that you need Passport for. You're not obligated to use those services if you don't want to. You'll still have a better experience on XP than on any prior operating system. I consider it to be night and day between XP and Win 98.
Presumably, the success of XP would also build a subscriber base of Passport and HailStorm users as you go along.
I don't want to talk about HailStorm because I think there's tremendous confusion about what the world thinks it is. I want to talk about what's in Windows XP and what it talks to on the back end. There are meta-Internet services we talk about which we consider to be pretty fundamental, architecturally, for building and making the Internet a little easier for people to use. Authentication and presence--in the future, we may have others--but those two, for the present, are core. And we're trying to support both of those in Windows XP.
We're not trying to make a business on this. I think people are very confused about that. You have to enter very little data; you're not asked for credit card information or anything else. Many of the people using our services today already have those. We're trying to be very strict about the privacy of that, and I think we're actually trying to take the leadership role.
But as you're adding more functionality, the message to developers and content providers is that this is the future and that they, presumably, should adopt it. So isn't the conclusion that Microsoft is leveraging its desktop monopoly to present an unassailable Internet position?
Well, do you think there's only going to be one authentication site? I sure as heck don't. Do you think governments around the world are going to have us storing their credentials? I don't. I'm looking for a way to have an interface there to talk to lots of different authenticators...You have lots of other Web services and lots of other servers that are going to exist.
I happen to believe that a key set of functions of domain controllers have to exist on the Internet. I also don't believe they're all going to come from Microsoft. There are going to be others there. Somebody doesn't want to use it? They don't have to.
A key point, though, is the control of the desktop and the monopoly control that you enjoy. The user starts off using Windows XP, therefore Passport and therefore the rest of this stuff follows. You said there will be other alternatives--obviously. But with the control you have over the desktop, those are more theories than practical scenarios.
I believe Microsoft has to continue to add value or there will be stagnation and no sales. I'm working very hard to figure out what people want, and I'm trying to make that available--and I'm trying to do it in an open way.
I've heard all these debates, like, "Well, you put these items into the list for photo printing. How dare you!" OK, now here are my choices. I empty the list--I can do that; I can just flush the list and there's nobody in it. Is that good for users? Because I can't find them all that exists on the Internet. Or do I try to put some in there and try to do it in what I think is a fair way?
We've chosen to think about it from the customer side because we also want to have enough value in the system so that people are excited about it. If you're making the point that we have a leadership position and that we need to act like leaders in it, I hear you. I think we are, and I take it very seriously from that perspective.
When you're creating fundamental technologies like Passport, though, you also have people like VeriSign and others who do authentication--why not work with them? Why not create an open standard?
Stay tuned, stay tuned.
Well, why not do it now? You could avoid all this criticism where you look as if you're trying to derail authentication--which is going to be the basis of everything.
But anybody can do this!
Still, they have to make it compatible with you.
Why? What's to be incompatible with? Just add the stuff.
VeriSign tried that for four years, but it didn't work because they didn't have the momentum you're going to have with XP.
But I'm still confused. There's Windows XP--add whatever somebody wants to the system. We spent a lot of time building Windows XP. There's a lot in it. We're trying to add value across the board with plenty of opportunity. We've created so much opportunity for the industry that it's stunning.
Take Web services. Last year, it's my understanding that $13 million was what was spent in online printing, particularly photo finishing. It's a $40 billion business! There are going to be a lot of people who try to enter this market. That's just new opportunity. Make no mistake: I have no doubt that AOL and others are going to try to create authentications. More importantly, other governments are going to create authentication authorities; you're going to be able to plug this in.
Limits to definition of the OS
You keep saying add it in rather than bundle it in. The user gets XP with the readymade bundled service rather than adding VeriSign or other authentication technologies. That's the fundamental issue--and has been. If the past is prologue, then obviously Microsoft wins by default. They own the desktop. They can bundle the stuff, and the user has easy access to it. It's about ease of use.
Well, I certainly hope we did it in an easy-to-use way. But you're going down a complete slippery slope of what is it that we shouldn't do. I mean, Oh my gosh, we added a line of code. Ooh, boy, this line of code--you know, I don't know; maybe we used too many variable letters or something. It's very, very hard to know--which will eventually play out here--beyond people in this room how that'll play out.
But as an engineer, I don't know how to do things if you don't listen to what people want and keep adding value. I don't know how you stay competitive. Cars are continuing to change. The car that we just drove here today is shouting out directions to the driver. They're going to continue to make innovations like that.
But is there any limit? You're saying there's no real line between the operating system and the rest of it; we're just adding value--albeit with a 95 percent installed base.
Well, I guess this will all be played out in another part of the country. Software is software and silicon is silicon; without integration, I don't see how the world goes round and round.
The heart of these questions is the stewardship. You say you took it seriously, but there's a lot of evidence to show you have not always taken it seriously. So tell us what it is you've learned from the past and how you intend to operate going forward so these situations don't come up again.
I think the way we're operating with other people in the industry is evidence of our cooperation. We're giving up a tremendous amount of intellectual property to standards organizations, working with others to try to create the new foundation for programming. It's good for growing the overall market. I think we're doing much better at listening to consumers. We will compete. That's the name of the business here--but making sure we listen to people so they can see any opportunity that we create.
We are spending a lot more effort working on opportunities that exist--whether it be on the hardware side, peripherals, Web services or new applications--trying to make sure there's enough opportunity for people. Those are some of the things that I think we've learned.
Coming out of the trial, do you feel chastened? Do you go about the way you do business any differently?
We're close enough here to the trial, so I can't--though I'd love to--go there. When it's all said and done, we can have another discussion. Obviously there's change. To be clear, we got what we considered to be an order and we changed to try and address it as we read it. Even before that came out, we changed the way the agreements were done. That's more on the specifics issue. On the global thing, independent of the legal thing, there's a question of understanding we are a leader and the role we need to take. I think if you talk to people at Microsoft, they grok that now.
How has the trial affected the launch of XP?
The trial has had zero to do with it--absolutely zero.
But you raised a lot of questions that perhaps you were trying to affect whatever might be going on in the courts.
(Laughing.) You know it takes time. You know we've had this RTM (release to manufacturing) in this time frame. We had no choice! If we were going to make Oct. 25 (the launch date of XP), we had to...and it has zero, zero, zero to do with anything going on in Washington, D.C.
How have you worked out the icon issue on the desktop with the OEMs?
Well, I am a super-hard-core believer that the desktop should be clean. When the (appellate court) ruling happened, we changed and said basically, "You can do it the old way or the new way. And the old way means that there are going to be icons there; we're going to put some there, and you're going to put some there. If you don't put any there, we won't put any there." So, that's the way it is.
But the start-up screen will have to be Windows?
What do you mean by the start-up screen?
When you boot up.
The fundamental goal is that we would like Windows to be intact and the user then gets to choose what they want. So when the user gets to Windows, they can choose--and it can bring something to totally change the shell, totally change everything, the song, whatever.
But it has to be user invoked, not by the OEM?
Honestly, it's quite involved...We do want to have a holistic Windows experience and let the user select it. If we provide the user with the choices or the OEM does, then we're very happy with it. There are situations where the user doesn't necessarily get the choice--and that is what I consider to be bad.
What would be a good example?
Well, something we don't permit and the court said they understand was that the user interface can be replaced without the user deciding that's OK...We think the user should make that choice so we have a coherent Windows experience when it comes up.
So you're saying OEMs don't have the liberty to mess with the user experience?
I didn't say it. I said there's a whole series of ways the system can be customized. And those are defined in a tool called the OPK that they use, and the OPK follows the license that the OEM has.
So the OEMs can, but they have to follow certain guidelines?
That's correct. and we try to give a lot of flexibility. Just take the Start menu: There are total of eight slots there. The OEM, if they so choose, can basically control five of the eight. It's our product; we think it's very fair.
And the five of eight are what?
They have control of the top two. And then they have control of the three at the bottom.
Are any of the features at launch going to be related to .Net?
Not heavily. There are just some nice, cool little toys and the like. But we expect more platform APIs (application programming interfaces) to become available that we haven't pushed on. Within a very short amount of time we're going to show how to build some cool applications on what's already in Windows XP. The technology's already there; we just haven't promoted it yet. And we're going to promote it and show some apps taking advantage of it.
Charles Fitzgerald of Microsoft has talked about XP being an endpoint for HailStorm. What was he referring to when he said that?
I think he was saying that a Windows XP system being able to have Web services--whether they be Microsoft or other--we're trying to make it so that you can write apps to access these Web services. We are trying to put the tools and the technology in place in Windows XP so that you can process XML and do it blindingly fast, as an example--and be able to talk to Web services that are published in UDDI (Universal Description, Discovery and Integration) and are using SOAP (Simple Object Access Protocol) interfaces and be able to do that on the client. So it's basically an endnote, a client into a cloud.