Microsoft CEO Steve Ballmer spoke with eWEEK Senior Editor Darryl K. Taft last week after his keynote speech during which Microsoft officially launched Visual Studio 2005, SQL Server 2005 and BizTalk 2006. The interview was two days before Microsoft released internal memos about the sea change the company was going through in its effort to attack the world of services. Ballmer, however, discussed that issue and a lot more.
When you guys began your enterprise play, a lot of your competitors said you werent an enterprise company and you couldnt compete. Youve proven that you could. But now youre entering the tools space with an enterprise toolset, which is not your heritage. How do you think youll be able to bring that famed Microsoft high-volume, low-cost, mass-market story to this arena?
Ill be shocked if we dont have super-high market share within a year. Ill just be shocked. If you take a look at the total number of customers that these high-end development suites have, its tiny.
We looked at Rational before IBM bought it. The number of actual seats and users they have is tiny. And so if you can take some of the good concepts and put them in an ease-of-use package and at a price point that you can get out, I think developers want this stuff you just have to make it easy enough to use and at the right price.
And I think with Visual Studio Team System we have that. And I would expect to see our share of high-end software life cycle seats to really climb quite dramatically for the next year.
So youll eclipse, and pardon the pun, but youll eclipse Rational?
Id be very surprised if we dont. Who knows? But I think weve got great products, and I think those things are so low-volume that in a sense it should be easy to bring those kinds of capabilities, I wont say to the masses, but to a much larger mass of people than those guys have reached.
Well, its always been an area sort of focused on heavyweight stuff for heavyweights.
Yeah, if you go back to the old Bachman and all the guys whove ever competed in this category, none of them has ever gotten to critical mass. But I think by the way weve done the integration with Visual Studio, I think we should get there.
You started out with sort of a joke about timing and product cycles, what do you think you have to do to get more agile in terms of delivering products?
Well, I think we have had an interesting Visual Studio release in between. So I wanted to take the issue head-on, but I think I maybe even over-accentuated in the remarks I made to the customers.
I think we have to decide that we want to turn the releases faster, and we can. What we did here is we said were not going to do our next release until we have a whole big bunch of stuff done, including the integration of the .Net runtime into SQL Server, which was a huge piece of work for both of those two teams, frankly. So you tie them both up.
And what we have decided is that were going to keep releases coming on a much more regular schedule. And if we have some big, thorny, hunky things that we need to do, were still going to do them, but were not going to necessarily tie the next release to them.
So well have a way to bring regular releases to market, while at the same time working on some things that have a longer cycle.
In some senses, I think we need three cycle times. With Community Technology Previews, were going to have some stuff that people see in the six- and nine-month cycle. We need our regular two-year cycle. And well probably always be doing some big, thorny stuff that is two releases out. And thats a change in mentality, but I think we can get there fast.
Weve always wanted to just do huge things, and we tie all the next release up with the huge things. We may have to do some of these huge things in a way that if theyre ready they go, but if theyre not we still have an exciting release.
Models Evolve
So what are we looking at in terms of the next rev of these products [SQL Server, Visual Studio, and BizTalk]?
Short. Its short… Whether that winds up being more like one year or more like two years, well have these products turning much more rapidly. And the teams know that.
I mean with the launch of these products … Im not going to get into whats in the next releases and blah-blah-blah-blah. But were definitely committed to a much closer cycle time—not just for the next release but from here on out.
Do you think the industry is moving away from an economic model based on selling software licenses to alternatives such as software as a service—like your Office Live, and subscription-based services like Red Hat?
Well, I think youre going to have all three models. Youre going to have what we call the transaction model where you sell something. Therell be a subscription model where you pay for something as you use it.
And therell some things that will be funded via advertising or essentially as part of a bigger idea. I mean what are we doing with our Express editions?
The fact that theyre available for free download doesnt mean that weve given up being a profit-making company.
But we do recognize that theres just a budget limitation students have and yet we want to create a funnel so that over time some percentage of those people, themselves or the companies they work for, wind up buying a full Visual Studio, a full SQL Server.
So I think youre going to see a variety of business models, and were embracing all of them. We grew up certainly on the transaction business model. But what we really talked about last week is were going to have a variety of things we do that are ad-funded and a variety of things we do that are subscription-funded.
But how much weight do you see you placing on each, how much can you get out of them?
Well, we already sell almost $2 billion a year advertising, so I cant tell you its small. And depending on how you look at some of our enterprise agreements, a number of those are kind of subscription-based, so its hard for me to tell you thats small. I think theres a lot of economic promise in all three for different scenarios. Each has its own place.
Do I actually think ad funding is going to be the primary source of revenue for mission-critical applications? No, I dont think so. I think thats still going to be either transaction-, or perhaps some time in the future, more subscription-oriented revenue.
Will there be an equivalent or similar “Live” strategy for tools?
When theres something to say on that, of course youll be the first to know. [Laughter] But I think its probably fair to assume that were working on a “Live” strategy for every part of the product line. We talked about Windows and Office and Xbox last week, but let your mind run wild.
You said Visual Studio, SQL Server and BizTalk make up the Microsoft application platform; how do you see that platform competing with the competition?
Those are core components; I also tried to show Office and Windows…
Well, who would we call competitors, because everybodys got their own view? Oracle competes with a piece, but only with a piece. I mean really, theyre only a serious competitor in the database.
IBM competes in the high end of the enterprise with WebSphere and DB2 a little—but DB2 has so little traction.
In the hosting market, we certainly compete with Linux and PHP. Thats why we really wanted to take a look at what weve done for shared hosting and with the Web Developer Express Edition, because I think weve done some amazing work that will help us build share against that competitor.
But I would say youve got those guys and then youve got the old Unix crowd, such as Sun, HP, etc. Those are probably the four competitors.
None of them have as complete a stack as we do. None of them have as integrated a stack as we do.
Were open—you can plug our stuff in with other guys stuff. But at the same time, our whole is bigger than the sum of our parts.
And none of the competition tries to go from student to enterprise, and from one-person, two-month projects on up to 30 people for a few years. So I dont think theres anybody whos trying to compete across the board. And we are in fact trying to be best of breed.
We want to be the best platform for shared hosting, we want to be the best platform for the enterprise. Despite the fact weve got this breadth, we think we can have breadth and be best in class. And we know thats important to the developer crowd.
Next Page: Competition in tools.
Competition in Tools
Youve been leading in the tools arena for some time, but who or what do you see as your primary competition?
On Windows its unclear who you would say. Youd probably say its a Java runtime with probably Eclipse-based tools. Thats probably the number one competitor. But its a different platform and different tools.
I would say if you go off of Windows and look at Unix environments its sometimes Java, its sometimes Apache and PHP. You get kind of a mix of things.
What about down the road? Do you see that mix changing? Do you see Eclipse becoming more of an issue?
Theres certainly some momentum, at least among the non-Windows platforms building around Eclipse. But its sort of like everything else in that environment: What really happened? How quickly? How chaotically?
And how much money they can make. Right now I dont think anybodys made much money off of Eclipse, but somebodys going to break through that.
I dont know. Nobodys made any money on Linux, and nobodys broken through that. So its not clear to me that anybodys going to break through and make money around Eclipse. I mean, IBM wasnt making any money around it, thats why its now open source. They kind of gave up on it.
As open-source continues to drive value out of commodity middleware, will Microsoft have to move further up the stack into vertical applications?
I dont believe open source has done anything to commoditize commercial software in terms of what it provides as core infrastructure. So my answer is were not going to allow ourselves to be commoditized.
Weve got too many innovative ideas for that. And if you take a look at it, over the last few years weve gained market share in the server, weve held market share at the client, and I think the customers have said clearly, “These guys are innovating in ways we care about.”
I was at a small event on the eastern shore of Maryland where Brad Smith [Microsoft general counsel] spoke and offered what many saw as the first olive branch Microsoft held out to the open-source community, and since then weve heard of a lot more youve done to open dialogue. Whats at the bottom of that? What are you looking to get at?
We all coexist in one big world here. We compete, but with everything … Well, open source is a little funny because you dont know the company—theres not a “company” to go talk to.
If you look at for every competitor we have, we compete vigorously, but there is actually a level of cooperation.
With IBM weve worked hard on interoperability standards, with Sun were now working hard on interoperability standards, with Oracle weve worked hard to make their database work well with our operating system.
So the real question you could ask is…and its a little bit harder to know where to turn, but were trying to ask, where do we need and what does cooperation look like despite competition with open source?
And we get smarter about that every day. We do compete with open source, but therell be a “coopetition” angle, and were trying to figure it out.
Well, would you expect to have any tools that support other environments?
No. No. There are plenty of people that are doing that. Nobodys making any money doing it. Our job as a company is to make money. I dont see anybody making money at the low end the medium end or the high end around open source.
So we need to outrun it, but at the same time, doing tools that support it doesnt look to fit with our strategy.
Im not sure if customers would expect it from us or want it from us. And certainly at the end of the day I dont know how to make any money from it—at least not now.
Whats the bigger risk Java-based programming models continuing to gain market share at the expense of .Net or supporting .Net frameworks in the form of Mono or Rotor on Linux and growing the size of the potential .Net market?
Well, Javas lost market share. So your question sort of says, what are you afraid of, continuing to grow? And I start with the fact that Java has lost share for the last five years. Five years of market share loss. Go get your own favorite survey, but .Net just keeps gaining share.
So I guess I can say, sure thats a competitor, but right now its a competitor were doing very well with. If you take a look at the .Net clones… You didnt use that word, but the guys you described are essentially trying to clone our intellectual property.
You always have to worry about clones. Weve had clones of Windows. None of them have ever gotten any traction, but theres a set of things you have to do to stay in front to protect your intellectual property and were doing all of those things. So I think well compete effectively with both of the scenarios you described.
Whats Eric Rudder going to be doing now?
Hes doing some advanced development work … some important advanced development work that I think is probably not appropriate for me to get into with you at this time. [Laughter]
But hes still a key guy, a go-to guy in the company?
Oh, yeah, Erics a great guy. Hes a great talent and a great guy. Hes fantastic.