Back in October 2013, I sat down with Satya Nadella, Microsoft’s cloud chief, and we ended up talking about Microsoft and its competition, the role of the company in modern technology landscape and what it needs to do in order to get its mojo back. While most of the conversation took place from the context of cloud computing, we did get into specific challenges facing the company.
The conversation was very telling, about Nadella and his approach to the large-scale overhaul the company needs to make in order to regain some of its lost sizzle. This week, there have been several reports that he is going to be named CEO of Microsoft next week — a great choice if you ask me — and the interview below will give you an insight into why I think so.
Om Malik: What does the future of Microsoft productivity apps and Microsoft’s Enterprise business look like, say in 2020?
Satya Nadella: Let’s even take a couple of different dimensions to it. I think everyone’s going to be in the cloud and no one’s going to be exclusively on one public cloud; that’s the dichotomy. So it means you’ll be using multiple SaaS applications that could be in multiple clouds, you might even still have a private cloud, not just private servers, but you may have a private cloud and a public cloud. So if distributive computing isn’t going to die, it’s actually going to be thriving even in the year 2020 or 2030, then (our goal is) imagining that future and making it easy for end users to use their devices that they have to be able to access corporate information and for corporations to be able to have controls that allow them to give per device and user access. That’s the future that we build for, not the old model of IT procures and provisions but the model where end users procure and IT governs. That’s the model that I see.
What is your biggest challenge right now from a cloud and enterprise standpoint in the company?
The way I look at it, in order to be an enterprise company of the future, there are three things that I believe one has to do to be relevant. First is, you need to have hit apps, there’s no point by starting by just infrastructure because if you start with infrastructure without the most horizontal ubiquitous application efforts you probably will get the infrastructure wrong. So that’s where the tip of the spear for us is Office365 and the fact that we are off to the races on it, in terms of our hit rate versus Google apps is pretty high and we’re very happy about that.
Now the second piece is Azure itself, which is the public cloud infrastructure. You need to have an honest-to-God public cloud infrastructure which is at global scale. It has to even be available in China which is something uniquely we do and in fact, Office365 runs on Azure so it keeps us honest about your point about what parts of Microsoft actually run on Azure. All of our first parties run on Azure and that’s a big piece to it.
But the third piece I think is the most interesting one is it’s not enough to even have our own massive first-party public cloud but you need to enable others to build their own cloud. So that’s where our servers come in.
So I look at that three-part framework and then score ourselves, which is on number one, we say, “Hey, we compete against Google and Salesforce,” and I’ll claim that we have more usage than Salesforce because of the nature of the category of Office and I’ll say in the core of the enterprise, we have much more success than Google apps has, even today as of SAS applications.
On the second one, Amazon is definitely doing very well, they are off to a good start. They’re even a partner of ours, but we’re the number two player. If you ask me, “Hey, score yourself in the last year, have you bridged the gap?” You do your own survey, maybe in startups not as much but as far as when it comes to the enterprise we are the number two guys and we’ve been growing faster.
In the world of cloud, startups are where you find the most future growth. Dropbox, for example benefitted from Amazon Web Services. Companies that are growing really, really fast are going to Amazon. So as a result of that they (AWS) get to experience a whole different kind of usage behavior (and scale) which perhaps doesn’t translate to Azure. How do you attract the startups to your platform?
I look at it and say, first of all, I think my own first-party diversity from Skype to Yammer to Xbox Live to Bing to Office365 gives me a lot of diverse set of what I would call SaaS application feedback. The other thing is, let’s not forget that all the ISVs out there in the world who have SQL applications, SQL server is the most ubiquitous database [chuckle] in the world and they are moving all to Azure.
In fact, I have meeting ISV after ISV, who built banking applications, insurance applications, healthcare applications; they’re all database applications of the past but now they’re big data applications. They have a huge web front end to it, they have a mobile front end. So they’re the ones that are moving into the cloud.
Even this morning, one of the biggest applications in India happens to be the Rail Reservation System app, and that’s now all built on Azure on the back. So we have a lot of these wins across various countries where we are seeing the third-party usage. You’re right in saying, do we have to do more in the Silicon Valley in particular where we have to get more relevant? Absolutely and that’s right. I was recently meeting with the Cargo guys who are one of the Azure customers, they’re the machine learning company out here. So we have Blinkbox which is like the Netflix of the U.K. They are owned by Tesco now — that’s on Azure. So we are seeing Warner Brothers … replicating exactly what Halo did with game play analysis for their games.
What is your outlook for cloud over the next 24-72 months? How do you see the Cloud evolve both from scale, the size and the scope and explain to me in terms, perhaps that more can understand.
Cloud is perhaps the most secular growth engine out there because it scales with a number of devices, it scales with a number of apps, so it scales with the users and their devices and their apps. So that means all of those things are calling back home, they need more compute, they need more storage, and so therefore, cloud is the one thing you can bet on, is that there is going to be more cloud. Just because you can imagine all of those other things exploring.
Now then the question is, what is that nature of the cloud? That’s where I think, those three things that I mentioned: one is proliferation of SaaS applications and in particular … the question of what’s the most ubiquitously used SaaS application? And the race is on for us and the industry and I think that’s where I think it will be something around communications and collaborations. That’s why Office365 is super important to us.
The second thing that I think is also going to be true is, who are the scale players who can actually satisfy that demand? In the past, in the enterprise, there were actually many enterprise players. There was us, there was Oracle, there was VMWare, there was EMC, there were many guys who in fact all serviced the enterprise. In the new world, who are they? There is Amazon, there’s us, and maybe Google if they are serious about the enterprise infrastructure business.
You don’t think they (Google) are serious?
I don’t know. I don’t run into them. We compete with them, but on Office365 and Google apps, not as much when it comes to Azure. It’s really Amazon and Azure as far as I can tell when it comes to the public cloud. That would be the other thing that I see, who are the ones – and you can’t just say I’m going to enter this business tomorrow, it requires five, six billion dollars of capital expense every year — who have the muscle to be able to get into the public cloud business. If you’re already not in it at this scale, you’re probably never going to be in it. You have to wait for the next big inflection point. And then lastly, it is like, if you go with the notion that the world will still be hybrid, and there is a need for server products that mimic what you use in your own public cloud, that’s where VMWare and others will play and I think that’s still going to be very important.
Right now there’s a lot of applications which are in private hosted environments and internal hosted environments. What type applications do you think are going to actually move to the cloud?
That’s a great one. What I see is sometimes we will be thinking of it and conceptualizing this is just a migration. … Take American Express – they’re doing a lot with Azure but one of the fascinating things I’m noticing is that they are using Azure for a lot of new things that they never did in the past. They are building back-ends for a lot of mobile experiences they are creating. They are automating a lot of things that they never automated in the past when it comes to their merchant experiences. Take Telefonica, which is a big customer of ours, where they basically wanted to partner with us to build a private cloud.
They built that but they are now using the private cloud with Azure, not just for dev/test but also to be able to do geographic expansion, which Telefonica itself is expanding geographically. And the best means for them to do that is to not open up data centers everywhere they go but to be able to use us. Same thing with Disaster Recovery; one of the fascinating things I see is everybody who buys a SQL server, in fact in this release, gets an always-on secondary in Azure. So it even comes through where it’s just part of what you’re doing. So you may not even need migration. So one of the things is, I have an always-on database, where is it running? The always on database is kind of on your private cloud and kind of on Azure and so the distinction even goes away.
Can we switch gears a little bit? So we were talking a lot about the cloud, right? The question I have for you is, rightly or wrongly, Microsoft is known as a company of the past, compared to Facebook or Twitter and any of these new companies that are popping up. How do you attract talent to Microsoft? There’s some serious options for smart people now, it’s not just one or two big companies.
Look, first of all, I joined the company in 1992. It’s not like that serious competition and seriously hot competitors is new. We’ve always had them and it’s just that our memory is such that the present state always looks hotter and cooler than our past. But even if you look at the dot-com era there was lots of companies and things going crazy in the stock market. So therefore, we’ve always faced this.
I do a lot of recruiting from college and even at the senior levels, and in fact I just hired one of the key guys out of Amazon, doing all of their machine learning work on the retail side and he’s working on our side. So we have tremendous amount of success in getting people from different places. There’s a guy who was working on database technologies in startups who just recently joined us, so I would say, the pitch is, “Look, we are the only company that is audacious enough, bold enough, to do a bunch of different things.”
From building the core operating system for the cloud to building the games console. And that level of breadth, you could say if you do it well, can be a fantastic opportunity for anybody, either as a student who is graduating to a senior person in hardware. Take some of the people who we are hiring nowadays are people with a lot of silicon experience and silicon is changing how you build servers, how you build game consoles. So I think that depth and breadth of experiences one could get out of a Microsoft tenure is still unparalleled I would claim.
The not-cool image works against you guys in the sense that, I don’t find (young) programmers who say, “I’m going to go work for Microsoft.” They all say they’ll go work for Facebook and they can perhaps have a bigger impact. Don’t you think that’s a legitimate challenge? Forget all the market forces at work. I think the inability to attract young, fresh, talent can be basically the end of any enterprise.
I track all the graduating class out of CMU or Stanford or Berkeley or anyone Comp Sci (schools) we track and we look at how many people did we make offers to, how many people accepted and the data shows that we win. There’s definitely new people who are showing up on campus than the traditional guys. There’s Google and there’s Facebook and there’s Twitter but as far as people who look at it — if I want to do serious systems work that is relevant in the era of the cloud — where else would you go? I think we get our fair share.
In the world of cloud, I think of five big players. There is you guys, there’s Amazon, there’s Google, there’s Apple, and then there’s maybe Facebook and Facebook is still a pure play software company. The rest of the four are doing devices, cloud and services. How would you rank those four?
Let me take the enterprise lens because that’s the one I live and breathe. If you look at what I described as the framework at least that I evaluate by, which is who has the most used SaaS applications in the enterprise, who has the cloud infrastructure that backs that up and is available for third parties at scale — and there’s no lock in effect which is somebody who can say, “I’ll give you even the infrastructure software to stand up,” — I’d claim that really it’s us and Amazon and Google; those are the three players. If I had to rank in the enterprise, I’ll claim we’re number one, Amazon number two, Google number three. That would be my rank ordering of the new enterprise players.
How would the overall market, consumer enterprise, cloud…?
SN: It’s Google and Microsoft.
You don’t think Amazon and…
SN: Not broad enough.
You don’t have a big device penetration just yet.
Steve (Ballmer) gave me this framework which I think works a lot which is — if you have to predict what the future could look like, then you have to say, “Do you have capability, do you have concept, do you have capability and do you have the culture?” I think those are all three worthwhile things to think about.
The concept of devices and services, in just looking at our share, our share in mobile is low but our share on PC is high. But as long as you say there is going to be all sorts of ubiquitous computing around you, there’s going to be big screens, small screens, things in your pocket, and we’re going to be a player. Do we have share today, and do we have to grow share, that’s I think we will be a player.
The second one, really will come down to the capability that most difficult capability that I think is not easy to build because it requires both scale and immense amount of software skills is the cloud. The cloud infrastructure is perhaps the toughest, if you’re not in it, and so if you take those two things, it’s really us and Google are probably the two players who have the best.
I’d say developers are the third thing. I think while you guys do have a good control of the past operating environments, I think the new world is very different. I think there is multiple choices and Microsoft hasn’t done well in the world of multiple choice, it has always struggled. I think that is the challenge, when I look at the cloud environment — the big four plus Facebook — and the challenge for Microsoft comes down to: you’ve got developers, you’ve got cloud, you’ve got the legacy, the enterprise relationships, those things you got right. You’ve got old legacy products which people like to use however there is a little bit of challenge when it comes to the new world. Look at the world where people are spending money on applications, devices, other way of thinking. We don’t have a Microsoft product in our company now, which is unimaginable five years ago and I think that’s the challenge you have.
SN: I thought you had an Xbox in your lounge?
The Xbox is a different story. [laughter] (Editor’s note: We have an Xbox 360 we forget is there.) And if you really think about the challenge, is that as a company, you don’t get to scale with me. Google gets to scale with me and Dropbox gets to scale with me and to some extent Apple gets to scale with me or maybe not, we don’t know yet. But I think that is the question I was trying to get to — isn’t that a big challenge?
Oh, it is, absolutely. Take your own case, the thing is we have to have things that really make it such that you adopt. That’s why I think the Dropbox was the Skydrive or Yammer was as anything else you were doing for your social, inside the enterprise, which then is just an entry point into Office365. So some of the cycles you’ve gotten into with let’s say Google, you sort of described it well, you said, “The Google ID is critical for me and your communications is critical for me,” I agree and the fact that we lost you as an Office365 is horrible, but luckily, I think three out of four of those customers, we win and that to me is really what is the story of our relevance in the future.
Before I go, two questions. What is the office space, the work space of an employee, when you have no offices and you have completely fluid on demand work force. What does it look like?
SN: In fact, you’re sitting in one such, Microsoft Office. It’s actually all open space and I studied, in fact, the Yammer and also a lot of other Silicon Valley companies, and our developer division is completely open space now. In fact, we did it in such a way that every twelve person team can self assemble at any given point in time and have a pod for themselves and we can keep changing it on a weekly, monthly basis. So we absolutely are believers in saying, “Look, the demographics are shifting, how people work, share the tools that they use at work shift, and how do you design the work space to keep up with it,” is one of the quests we have right now. One of the other things [chuckle] that I’m also aware of is the spread of diseases as I’ve learned [chuckle]. One of the fascinating things is one person gets the flu, everybody gets the flu, so what do you do about those kinds of things, are interesting issues but we definitely have some very interesting learning and experimentation. At this point it’s no longer like our work space is very different from Silicon Valley work spaces.
What do you think is the big challenge for the incoming CEO of Microsoft?
I’d say our current challenge and our current opportunities are the following. You look at, let’s divide it into two parts, on the enterprise side I actually think we’re playing offense. We have a big business, the business I’m associated with closely is 20 plus billion dollars but guess what, there’s two trillion dollars that play in this move to the cloud, so we’re playing offense all the time. In fact, all of today’s announcements are all about, “Hey, the data center of the future, device management of the future, date management, big data, and so on,” so I feel like we have only things to gain and we’re already in there. So it’s not as if you’re talking about the future that’s some time in the distant future but it’s today.
Then on the consumer side, you could say we have devices today and we have a new operating system that we launched, that’s a pretty radical set of bets. New silicon, new UI, new app model, and the fact that we had had tremendous success with the desktop, people look at that and say, “Hey, instantly why is this not at the same scale?” But we’re not confused that it’s touch first and big screens are all important and we’re coming at it and you’ll see us to continue to innovate on our devices front. So the challenge really, and the opportunity is the devices and services and making progress on it.
It’s funny, you didn’t have a single comment about Apple’s Cloud Services. I feel like that’s their biggest shot coming out of the company. I was pretty amused that you didn’t bring that up. Maybe they should use Azure.