Microsoft CEO Steve Ballmer visited Silicon Valley on Thursday for his annual meeting with top venture capitalists to talk about Microsoft product strategy. This was his first visit to Silicon Valley since announcing the search partnership with Yahoo in July.
I had a chance to sit down with him just after that event for a hour-long one-one-one interview. In the first 10+ minutes of the interview Ballmer gives his high level thoughts on major Microsoft products and strategies (including Windows, Windows Mobile, Internet Explorer, Bing, Azure, Mesh, Natal and others), competition, the future of search and search marketing, Microsoft’s “three screens and the cloud” strategy, the recent acquisition of Interactive Supercomputing and, yes, even his thoughts on Twitter.
A few interesting points from the video interview:
- On Microsoft’s “three screens and the cloud” strategy: Ballmer says it’s a “fundamental shift in the computing paradigm.” He added “We used to talk about mainframe computer, mini computer, PC computing, client server computing, graphical computing, the internet; I think this notion of three screens and a cloud, multiple devices that are all important, the cloud not just as a point of delivery of individual applications, but really as a new platform, a scale-out, very manageable platform that has services that span security contacts, I think it’s a big deal.”
He also says that Microsoft obviously won’t be the only player in the new market, and joked that some people “for whatever crazy reasons don’t want to be on windows, might want to be on linux:”
Ballmer: Now in our own case, you know we’re going to try to share technologies, so that we get kind of synergy from a developers perspective. Windows on the phone, you know, Windows PCs controlling TVs, the Windows PC of course itself, Windows Azure in the cloud, so we have a lot of work that’s trying to share technology, but obviously you don’t want exactly the same experience on a little screen and a very big screen and a mid-size screen.
Arrington: Ok, does it work – you talked about Azure in the cloud, but does it work if somebody’s using Amazon web services or something like that. Although we’re talking more about the developer side now, but are you planning to interoperate as much as possible.
Ballmer: As much as possible implies that infinite complexity’s a good thing. Of course, it’s unreasonable to say that you’re going to completely support only your own three screens and only your own cloud. I wish that it were true. We have to make our screens and our cloud first and best, but clearly there are going to be people for example who don’t want to be in the cloud, that want to be on premise, that for whatever crazy reasons don’t want to be on windows, might want to be on linux, for gosh sakes.
Arrington: Yeah, crazy.
Ballmer: For me, I’m allowed to say that. And we need to interoperate, but we do need to be first and best in support and in integration of our own platforms.
- On search innovation: Ballmer says that search innovation, both as a product and a business model, has largely stagnated over the last five years. He also thinks competition will drive more innovation in the future. “I think if you look out the next 10 years we’re going to see more innovation in search,” he said.
- On Why Microsoft won’t build a branded phone, as they have with the Zune and Xbox: Smart phones, like desktop and laptop computers and televisions, are “non-niche devices,” which he defines as 300 million or more units per year. These markets are large enough that there will be multiple manufacturers, and it’s unlikely that any single vertical vendor will dominate the market. It makes sense, he says, for Microsoft to be a vendor of the platform and services for these types of devices. So, don’t expect a Microsoft branded phone.
I’ll call anything that’s north of 300 million a year non-niche. PC’s are not niche devices. Part of the reason I think they’re non-niche devices is, multiple people can manufacture them, they all interoperate, they work together, etc. TVs are not niche. You know, there’s more than, well over 300 million of those sold a year. They interoperate in that case mostly based on standards, but with some innovation. Phones are not niche. The categories where, I think, a single player can control a large percentage of the volume are the smaller categories. What does Apple sell every year of iPods: 30 million, order of magnitude, something like that. What is the whole video game market is maybe 30 or 40 million in units a year. But when you get these categories that are 300 million, 500 million, a billion, a billion-five a year, the truth of the matter is you’re gonna want multiple points of manufacture, with a lot of innovation around it whether its supply chain, for geographic diversity, and our basic play with our software is to try and be super high volume. So I think you can have an Apple in the phone business, or a RIM, and they can do very well, but when 1.3 billion phones a year are all smart, the software that’s gonna be most popular in those phones is gonna be software that’s sold by somebody who doesn’t make their own phone. And, we don’t want to cross the chasm in the short run and lose the war in the long run and that’s why we think the software play is the right play for us for high volume, even though some of the guys in the market today with vertically oriented solutions may do just fine.
- On Microsoft’s acquisition strategy: Microsoft acquired 15 companies in FY2009. Ballmer says to expect roughly the same level of activity in the future. “I’m guessing we’re gonna want to buy about 15 companies again next year,” he says. He says most of those will be smaller transactions ranging from $50 million – $400 million. And those companies will have to “really fit well with our technology platforms and distribution,” he added.
- When I half-jokingly asked if Twitter fits well with Microsoft’s technology platforms and distribution, he responded “Twitter would be great, yeah. I mean, not that we’re talking about buying Twitter…” And he then went on to say that he thinks the Twitter guys are “fiercely committed to staying independent.”
For the rest of the interview we took a deep dive into each of these topics, and over the next few days we’ll have a few follow up posts on each area of discussion in detail. There is some absolutely amazing content to come. The full transcript of the video is below.
Additional Posts From This Interview:
Microsoft CEO Steve Ballmer On “Moving The Needle”
Microsoft CEO Steve Ballmer: Chrome And Safari Are Rounding Errors.
Full Transcript:
Michael Arrington: This is Mike Arrington, I’m here with Microsoft CEO Steve Ballmer, hello Steve. It’s been a year since I’ve had a chance to sit down with you, you’re back in Silicon Valley – what brings you here?
Steve Ballmer: We do an annual event where we bring together venture capitalists, and try to make sure they understand where we’re going, and we understand where they’re going, cause there’s going to be opportunities for us to partner with their portfolio companies, try to get their portfolio companies to build on and alongside of things that we do. There’ll be chances for acquisition, and we do that once a year, and we happen to be doing that today. I’m down in the valley probably 7 times a year, and this is sort of more of a valley day because we’re with the VC community.
MA: The last time you were in the valley, is that when you did the yahoo deal?
SB: Last time I was down here probably was the announcement of the yahoo deal. I’ve had kind of a quiet two months of travel since I was down here at the end of July for that.
MA: I have a couple questions about that but we can hit those later. You guys have a lot of new product initiatives, I think some are big ideas, big new businesses, possibly some are what you call an enabler, something like that. You’ve got big buckets – Bing is out, Windows 7 is coming out, Azure, I think you guys have said by the end of the year, Mesh is there, Project Natal, and others. How do you feel about big buckets and all of these products – your babies?
SB: Well, it’s great to have a year where you have a lot of stuff that you can kind of be excited about, if you lead a company like ours we have a lot of very exciting products for the consumer and frankly for the enterprise customer, although that tends to get a little less press attention I would say, but it’s a fun year. It also is a good year to say then that it’s a good year to go build business and it’s a good year to lay the seeds for the next generation of businesses that can be good, and whether it’s what we’re trying to do with Bing, or do with Natal, the next 12 months is shaping up to be very promising.
MA: One of the big things you talk about is something you call “three screens and the cloud.” I’d love if you could dive a little bit into what that means from a business standpoint but also from a user standpoint and what they’re going to get out of this when it finally comes together.
SB: Yeah, the reason I like the little phrase “three screens and the cloud” isn’t just that it’s true and it’s what we started talking about it at CES but I kind of like the alliteration (?) of “Three Men and a Baby.” Every time I say it I think, “Three Screens and the cloud,” “Three Men and a Baby” if you remember the movie, but I think what it really refers to is a fundamental shift in the computing paradigm. We used to talk about mainframe computer, mini computer, PC computing, client server computing, graphical computing, the internet; I think this notion of three screens and a cloud, multiple devices that are all important, the cloud not just as a point of delivery of individual applications, but really as a new platform, a scale-out, very manageable platform that has services that span security contacts, I think it’s a big deal.
You lay natural user interface technologies on there, and platforms on there, and then you start revitalizing the UI platform. What you’re seeing on phones and TVs, people want more than what’s called the classical graphical user interface: touch, voice, camera, gestures – all of that stuff whether it’s Natal or the touch stuff, in iphone or Windows 7 or whatever it is. It is the next big generational shift in the computing platform. And people are going to want applications, I’ll call them that, or services, depending on whether you like old fashioned words or new words, but they’re going to want things that service them across those environments. So when I’m away and just have my phone with me I still may want to check in on the action – my favorite xbox competition, or I may want to play games with somebody who’s in a different environment. We’re sitting watching television, and we want to share with somebody who’s not physically present – we want that to work to somebody who might be a family member, who’s on their PC in a hotel room traveling tonight. So you got to think about it as one integrated computing infrastructure. Now, whether it will all come from one company, and what are the standards, and what are the points of proprietary differentiation, all of that’s going to get kind of played off.
Now in our own case, you know we’re going to try to share technologies, so that we get kind of synergy from a developers perspective. Windows on the phone, you know, Windows PCs controlling TVs, the Windows PC of course itself, Windows Azure in the cloud, so we have a lot of work that’s trying to share technology, but obviously you don’t want exactly the same experience on a little screen and a very big screen and a mid-size screen.
MA: Ok, you talked about Azure in the cloud, but does it work if somebody’s using Amazon web services or something like that. Although we’re talking more about the developer side now, but are you planning to interoperate as much as possible.
SB: As much as possible implies that infinite complexity’s a good thing. Of course, it’s unreasonable to say that you’re going to completely support only your own three screens and only your own cloud. I wish that it were true. We have to make our screens and our cloud first and best, but clearly there are going to be people for example who don’t want to be in the cloud, that want to be on premise, that for whatever crazy reasons don’t want to be on windows, might want to be on linux, for gosh sakes.
MA: Yeah, crazy.
SB: For me, I’m allowed to say that. And we need to interoperate, but we do need to be first and best in support and in integration of our own platforms.
MA: Ok, fair enough. Search. Congratulations, Bing is now one of the major services, they’re saying Bing has now achieved 10% of the market share, up actually quite a big percent over 8, 8 and a half – that’s a big percentage jump, and it seems to be pretty quick, and it seems to be steady and sticking. So congratulations on that.
SB: Thank you.
MA: But, search. Bing is clearly, I think everyone agrees, Bing is a good step forward. I think people who thought they would test them out, a lot of them are sticking, but looking forward 5 years from now, search innovation over the last say, 10 years has been somewhat interesting but will the next 10 years of search innovation be more interesting than the last 10 years?
SB: Let me say something dramatic – I think the first five years there was innovation in business model, there was innovation in approach, give credit to competition, the last 5 years there’s been some, quote, innovation, which is really things like digitizing, maps and books, or whatever the case may be, adding the larger content base to the corpus of information. But in some senses the UI, the approach, the algorithms have changed less in the last five years, then more … so I think if you look out the next 10 years we’re going to see more innovation in search. And, of course, that’ll be best served by good competition in the market and, y’know, at this stage, hopefully with the government approval of our deal with Yahoo, the good competition better come from us. Otherwise I don’t think we’ll see some of that innovation. But whether it’s natural language, visualization, change in the UI, change in the business model… Business model on search is making life tough for other content providers, makes life tough for some of the merchants…
MA: yes
SB: We’re not an incumbent, we can play with user interface, we can play with business models, we can do some things that maybe the incumbent can’t do.
MA: Yes
SB: And the incumbent does some things pretty well too, and we’re gonna have to hustle to catch up, and they have a lot of years of tuning those relevance algorithms, and we’ve got a lot of work to do.
MA: Though you talk about UI, you know, and sometimes it seems like people talk about search problems as, first of all, the search engine understanding the query, and I think that’s where you’re talking about UI.
SB: Well, that’s where I’m talking about natural language,
MA: Right.
SB: Actually, UI I’ll talk about the presentation of the results
MA: Literally, the sort of… visual presentation of the app.
SB: Ya, and I think it matters, actually, search isn’t unlike every other application; the way you present things actually does matter.
MA: What will your market share be in 10 years, on search.
SB: Oh, I don’t know. I mean, making forecasts is sort of not, not a sane practice. A lot more than we have today!
MA: It would be great for me if you answered, but…
(Steve laughs)
MA: The Zune HD: A hit. Seems to be a great device. Still haven’t had my hands on one. Do you have one on you right now?
SB: I don’t actually.
MA: It seems like a lot of people really like the Zune HD, it’s selling out. When you look at the Zune and the Xbox, you seem to be more than capable of creating, successful end consumer devices that are hardware tied to services. When do we get our Microsoft phone? I know you guys keep saying “We will not build a Microsoft branded phone…”
SB: Well let me ask you a question. I’m going to answer your question with a question. Which is to say, look, lets just break hardware devices into two broad categories. Really high volume, and more niche. And I’ll call anything that’s under about 50 million a year niche. And I’ll call anything that’s north of 300 million a year non-niche. PC’s are not niche devices. Part of the reason I think they’re non-niche devices is, multiple people can manufacture them, they all interoperate, they work together, etc. TVs are not niche. You know, there’s more than well over 300 million of those sold a year. They interoperate in that case mostly based on standards, but with some innovation. Phones are not niche. The categories where, I think, a single player can control a large percentage of the volume are the smaller categories. What does Apple sell every year of iPods: 30 million, order of magnitude, something like that. What is the whole video game market is maybe 30 or 40 million in units a year. But when you get these categories that are 300 million, 500 million, a billion, a billion-five a year, the truth of the matter is you’re gonna want multiple points of manufacture, with a lot of innovation around it whether its supply chain, for geographic diversity, and our basic play with our software is to try and be super high volume. So I think you can have an Apple in the phone business, or a Rem, and they can do very well, but when 1.3 billion phones a year are all smart, the software that’s gonna be most popular in those phones is gonna be software that’s sold by somebody who don’t make their own phone. And, we don’t wan’t to cross the chasm in the short run and lose the war in the long run and that’s why we think the software play is the right play for us for high volume, even though some of the guys in the market today with vertically oriented solutions may do just fine.
MA: You just bought Interactive Supercomputing a couple of days ago now. Why?
SB: Well, one of the big pushes for us, we, we’re in the server business, if you .. if you look at all the things people do with servers today, we do pretty well at most of them. The two areas in which we have, are least successful are web, where about half the web servers are Windows and half are Linux, we’re competing, and workin’ at it; and the other one is high performance computing. Scientific supercomputing-like applications. And the truth of the matter is, this is never about price, it’s always about, have we done enough to solve problems, compared with what you could do with the roll-your-own of Linux and Open Source. We took a look at it and said, no, we really want to be in the high performance computing, super computing space, and we’re gonna get after it. We’re gonna get after it with infrastructure, we’re gonna get after it with tools and applications, and as part of that strategy, we made the, uh, deal to acquire Interactive Super Computing.
MA: Okay, great. And you’ve bought fifteen companies in fiscal 09 – I think, is that right?
SB: Yeah, about right, yeah.
MA: Does that stay steady? Increase next year?
SB: I’m guessing we’re goning to want to buy about 15 companies again next year, if, just sort of a guess. Most of the things we buy are smaller. We probably pay some place between 50 million and 3 or 4 hundred million. And then occasionally we’ll do something bigger. But most of what we buy, it’ll probably be about the same order of magnitude, it’ll probably mostly be companies that have 50, 100 to 200 employees, And really fit well with our technology platforms and our distribution.
MA: Would Twitter fit with your technology platforms and distribution?
SB: Twitter’d be great, yeah. I mean, not that we’re talking about buying Twitter… I mean, the Twitter guys want to stay independent, that’s great. Using – making Twitter an asset to one of our businesses, that would be the real question for us, how does it fit, but obviously, yknow, they’ve got a lot of kind of buzz and interest at this stage, but, no, I think they’re fiercely, uh, committed to staying independent, which, which I respect.
MA: Great, thanks.
