Google really wants to be a cloud company. But as the Internet giant best known for search begins a two-day conference in San Francisco dedicated to its cloud computing business, it's faced with the reality that Amazon got there first. Despite running what is likely the world's largest and most advanced computer network---the global network of data centers and machines that underpins every one of the company's myriad online services---Google was rather slow in allowing other businesses to build and run their own software and services atop this online empire. That's called cloud computing, and for Amazon, it's now a $9.6 billion-a-year business.
According to Morgan Stanley, Google's cloud business pulls in closer to $500 million a year, a fraction of the Amazon empire. The potential market is still enormous---tech research firm Forrester predicts the cloud market will exceed $191 billion by 2020---but there's an added rub. Google isn't a company that's geared towards selling stuff to other businesses. Larry Page and Sergey Brin built Google on free services used by the world's consumers and, yes, ads that appear on those free services.
Nonetheless, some at Google believe the cloud could be an even bigger business for the company than online ads. So, earlier this year, hoping to change the culture of its cloud computing group, Google brought in a ringer: Diane Greene.
In 1998, the same year that Page and Brin founded Google, Greene and her husband, Stanford University professor Mendel Rosenblum, founded a company called VMware. With Greene serving as the company's CEO, VMware offered a technology called virtualization, which allowed businesses to run many virtual computers on a single physical machine. This sparked a revolution inside the computer data center: Businesses could get far more out of each machine---not to mention all the electricity needed to power their machines. And eventually, at places like Amazon, virtualization also gave rise to cloud computing. With its cloud service, Amazon is merely offering virtual machines to the world at large, via the Internet.
In 2008, amid ongoing disagreements between Greene and the VMware board, the company fired her. And in the years since, VMware largely missed out on the cloud revolution, focusing more on software that businesses could use inside their own data centers rather than pushing toward public cloud services that businesses could use without setting up their own hardware. It was yet another example of that old Silicon Valley conundrum: the innovator's dilemma. Should VMware venture wholeheartedly into a new, unproven business that could potentially cannibalize its old, proven business? For VMware, the answer was "No." But in 2012, Greene resurfaced on the Google board. And after taking over not only the Google cloud group but the group that runs Google Apps, a set of pre-built online office applications, she's trying to take Google to a place where VMware should have gone.
For Google, the trick will lie in combining the best of Google with the best of VMware, a company that very much knew how to sell stuff to big businesses. Greene assumed her new role at Google only about 90 days ago, but she has already started to transform the company, combining the engineering and the sales staff inside the cloud computing group while working to pump new money and new ideas into the operation. As Bloomberg Businessweek reported yesterday, Greene and company plan to erect 12 new cloud data centers over the next 18 months. And to further expand its cloud, Google will draw on the work of BeBop, a startup that Greene and Rosenblum were building before she took control of the Google cloud group.
In her new role, Greene will work closely with Urs Hölzle, the former University of California, Santa Barbara, professor who joined Google in 1999 as employee number eight and oversaw the creation of the company's massive data center network. In fact, the two have worked together for quite a while in a less official capacity, discussing the future of the Google cloud while walking their dogs on weekends. They share the same enormous vision. Now, we'll see if they can achieve it. Today's conference will serve as a kind of rechristening of Google's cloud business now that Greene has taken the reins. We spoke to Greene prior to the conference as she explained the path to this new Google and described where she hopes it will go in the future. (The conversation has been edited for clarity).
From the outside, this seems like a natural next step for you after what you built at VMware? Do you see it that way?
Well, it's not anything I ever anticipated. It sort of surprised me that I found myself doing this. But it is extremely natural. At VMware, we did the virtualization, which is now used extensively across the cloud. Virtualization layers---and containers---are what facilitated the cloud originally. And it certainly would have been a natural progression at VMware had I stayed there.
After leaving VMware, how did you wind up on the Google board?
I met Larry and Sergey when they were still at Stanford. I live on the Stanford campus. My husband is a professor there, so I knew them and I thought the world of them, even when they were graduate students. (Google board member and Stanford president) John Hennessey I know through Stanford. Paul Otellini, the former CEO of Intel, I worked with while I was still with VMware. I knew Ram Shriram through the tech industry. I basically knew all these people (on the Google board), and they reached out to me. Then, once I joined the Google board, I started getting involved on the cloud side, working with Urs Hölzle---fairly lightly at first.
How did you end up taking control of the cloud group?
I was actually helping them look for someone to run the organization for a few years. Larry Page had suggested that I run it a while ago, and I thought: 'Well, that's a nice gesture. Thank you.' But Urs and I got so that we were talking a lot while walking our dogs together on the weekend, and he kept bringing it up. It's such an exciting space. It's such an enormous opportunity. Google's assets---Google's technology and technologists---are, I think, unparalleled. So I said: 'Okay, I'm in.' I was also committed to working on my startup, and as Urs and I talked, we started seeing a pretty strong strategic fit with Google. We discussed it with Sundar (Pichai, the new Google CEO), and we decided it all made sense.
BeBop was in stealth mode when you joined the Google cloud group and Google acquired the startup. Can you now say what BeBop aims to do and how it plays into Google's ambitions?
Our hand was kind of forced by the acquisition. I hadn't planned to talk about BeBop for a while. It's a pretty ambitious project. But basically, we're developing ways of quickly building very rich enterprise applications, which is what you want to do in the cloud. We've invested very heavily in design---human-centered design, how humans look at this. Eventually, we were able to build a real center for design excellence at BeBop, which I found really exciting. You have to recognize how important it is---and how powerful it is---to a have a good user experience.
Can you get more specific? This is where many companies aim to go. Amazon is adding all sorts of pre-built applications to its cloud. Companies like Workday are redesigning offices applications for the cloud age.
I'm not wild about getting into too many details. I like to under-promise and over-deliver.
Few would deny that Google's technology underpinning the Google cloud is well ahead of the competition. But at the same time, few would argue that Google is ahead of rivals when it comes to actually selling stuff to big businesses. Google is an engineering company. And it's an advertising company. It's geared towards building all sorts of new consumer tech and then driving revenue through advertising. But to find success with the Google cloud, you must do something very different. Do you have to change the culture of the company? Do you have make it more like VMware?
I do---to some extent. VMware had an extraordinarily strong engineering culture. But I used to joke with the engineers: 'We could let all of you go and still pull in revenue for a good two years.' And then I would joke with the salespeople: 'We could let all the engineers go, but then, after two years, you'd be out of a job.' You need both. At VMware, the engineers worked very closely with our customers, closely with the field, and it was exciting for everyone. So, one of things I did here at Google when I arrived was to combine sales, marketing, engineering, and product. This is powerful. Engineers love having an impact on what customers can do, and by bringing (the Google cloud staff) together in this way, we can work towards that, feeding off each other and moving more quickly.
But can't it also be hard to combine two disparate organizations in this way?
It helps to build roadmaps---roadmaps that include new inventions from the engineers, that have features our customers need, that have general horizontal improvements, including security and performance. When you put all that together---and prioritize it---it's a wonderful way of bringing everyone together. People love the clarity of it. It's about getting very high level about what we're trying to do. You have to keep going up a level until you're all aligned. Then you can come back down and implement the details.
Urs has said that the Google cloud can generate more revenue for the company than ads. That's an easy thing to say. But it's an extraordinarily large goal. How will you actually reach that?
We're investing heavily in the business side, so we can go out and support our customers. The cloud is really at the beginning. As big as it already is, it's in the very early stages. Once you get everything in the cloud, what it enables for a company is unbelievable. You can start applying machine learning and intelligence to everything you do. I don't think we even know where this is going to go. It's all about us taking our expertise and our capabilities and then going and understanding what the possibilities are for other companies, and this requires investment. We're very serious about that.
VMware was a very different company from Google. It was focused on business computing from the beginning. But as Google cloud guru Eric Brewer has pointed out, the two companies were cut from at least some of the same cloth. They were created at around the same time, just as the country's great research labs were dying, and many of the top minds at these labs, including those run by DEC, moved into Google and VMware. "At the time of the bubble burst in 2001, when everyone was downsizing, including DEC, the main two high-tech companies that were hiring were Google and VMware," Brewer once told me. "Because of the crazy lopsidedness of that supply and demand, both companies hired many truly great people and both have done well in part because of that factor." Do you see this phenomenon from the inside?
Absolutely. In the early days, we had parties with Google. Then we really started competing with them for all the top people out of the DEC Western Research Lab. We got some of them, and Google got some of them. And then weren't having so many parties together. But we were very similar culturally. It was all about engineering excellence. Google went in a consumer direction, and VMware was system infrastructure. But there are a lot of parallels and similarities.
And now, at least symbolically, you are bringing the two companies back together?
In a certain way. I'm really enjoying it. I almost wish I had done it a few years earlier. But now is a great time.