Google Senior Exec Alan Eustace on Innovation Strategy and the Technology of the Next Decade

If ever a company defied hyperlocal coverage, it’s Google.

Whether you live in Boston or Bosnia, Seattle or Shanghai, Google is a big deal—and it’s getting bigger every day. Whether it’s attracting or gobbling up the best companies and talent in New England or Southern California, or competing with giants like Microsoft (Bing) and Amazon in the Northwest, that all takes a back seat to the bigger picture coming into focus: that of a truly global company delivering some pretty astounding and pervasive Internet technologies, but also facing some of the greatest challenges of its corporate life.

That bigger picture is exactly what computer scientist and Google senior executive Alan Eustace works on every day, and yesterday he gave me a unique glimpse of the inner workings of the company’s strategy in engineering and technology.

Eustace joined Google in 2002. Previously he had spent 15 years at the Western Research Laboratory (the last three as director), originally run by Digital and acquired by Compaq and then Hewlett-Packard. He is now a senior vice president of engineering and research at Google, where he helps oversee an annual R&D budget of some $2.8 billion (according to 2008 figures, that represents 13 percent of Google’s revenues) and numerous engineering centers around the world, including in the Seattle and Boston areas. (In October, Eustace spoke about the local innovation community at the company’s Kirkland, WA, offices.)

I got a chance to speak with Eustace yesterday about everything from Google’s engineering management strategy and what he worries about most, to its $750 million purchase of AdMob last month and the broader significance of mobile to the company’s business. We also touched on some farther-out implications of advances in areas like speech interfaces, machine translation, and even quantum computing. (Eustace had no comment on the widely rumored “Google phone” coming out next year, though.)

Here are some highlights from our conversation, edited for length and clarity:

Alan Eustace (image courtesy of Google)

Xconomy: How do you allocate Google’s engineering resources across core areas like search versus other areas like cloud computing and mobile? What’s the overall strategy?

Alan Eustace: We have a 70-20-10 rule. We spend 70 percent on core products, 20 percent on emerging areas, and 10 percent on “wild and crazy” ideas. Those are things we may not have a business model for, but may be important in the long run. Things can move between core and emerging. For example, mobile used to be something of an emerging area—the business models weren’t completely clear. Now with smart phones, mobile search is happening with increasing frequency. Internet access has gone up substantially in the last three years. Now people will do searches on either phones or the Web. People are using YouTube [on phones]. Something that used to be a tiny part of all our searches, and was quite different from desktop searches, has evolved into a superset of the desktop experience.

X: What’s your approach to coordinating which engineering center works on which particular products, and managing all those different efforts?

AE: It depends on the expertise in the area. I look at the various areas inside the company. I want to support those areas based on the talent of those [geographic] areas, the quality of the talent and leadership. It’s important that we don’t fragment, that we don’t have 27 places working on the same thing. It’s also important to have critical mass, so communication overhead doesn’t dominate. What I’m looking for is larger groups of people in each of the sites. Where there’s ownership and partnership between what’s going on there and in other sites. It’s not true that one site only works on only one thing. We have diversity.

Just as an example, Brian [Bershad, Google Seattle’s site director] has incredible systems skills. We hired some fantastic systems people. People who can build large-scale distributed systems are rare, they’re in high demand. We think about what’s the place where they’ll have the most impact?

X: It may be a bit cliché, but we’ve heard that Google worries most about how to keep its entrepreneurial and innovative edge as it continues to grow. Can you talk about that?

AE: It’s absolutely true. Technology is at the heart of our company. Technology is an interesting area in that it moves really quickly. People inside the company have to move as fast, and possibly faster, to exploit technology [to serve our users]. People have to stay nimble. We have smaller teams and more ownership inside the teams than most companies. We have bottom-up and some top-down ideas. It’s important that the people working in an area have freedom to try some different ideas. It’s not about having 10,000 ideas—that’s not the important thing. It’s having the right ideas, and enough time to spend on those ideas.

X: So how do you balance “organic” innovation with strategic acquisitions, for example?

AE: Almost all our innovation comes from inside the company. I think acquisitions help. Let’s take the AdMob case—we had a very healthy and active group inside the company. What that group discovered was that [mobile advertising] is a really, really important area. If we didn’t have people working on it, we wouldn’t have known. It’s the fact we’re working on it inside that makes us understand the potential of the idea. In many cases, someone else will help us. With YouTube—we had Google Video, but YouTube was winning, and they had a particular set of people, talent, and a viewpoint. We could try to copy them, but [in the end] what YouTube did was accelerate our effort in that area.

X: Which areas would you say entrepreneurs should actively pursue—not necessarily in core areas of search, but around the edges of what Google does?

AE: I’d encourage them to do it in search too! I was in the group that started AltaVista in ‘96. This whole industry is 15 years old. To say it’s done is way premature. There’s a lot of good ideas. Search is interesting—there’s no switching cost [for customers to choose a different search engine]. There’s lots of areas in search, mobile. We will be helpful to other companies, in terms of maps and other things we’re providing to help build other services. In mobile, we’re seeing incredible creativity around Android. This is a fantastic time to be an entrepreneur. There are so many changes in the way people access information.

X: Do you remember the first time you used Google? And how did you originally meet up with the Google guys?

AE: I used AltaVista first. Then I asked a friend who left AltaVista, and he pointed me to [Google in 2000-ish]. I ran the Western Research Laboratory, which was bought by Compaq and then HP. People who worked for me, or with me, all left and went to Google. It was probably 20 people. When the HP acquisition happened, I’d always said no [to joining Google]. But one friend said, ‘Now that the HP thing happened, I’m not going to take no for an answer.’ He set up a breakfast with me and Larry [Page], two blocks from where I worked. I didn’t have to go very far. Larry spelled out his vision, and it was much broader and more expansive [than what Google was at the time]. It must have gone well. I interviewed soon after. But I wondered about this little company—they weren’t making a lot of money.

X: Looking further out, over the next decade, what are we going to see across the main fields that Google competes in?

AE: The way it’s going to play out is very exciting. The next 10 years are going to be unbelievably powerful in terms of a technology point of view. Things I’m almost certain are going to happen: Machine translation will become ubiquitous and as good as human translation, so the language barrier will be gone. All mobile devices will have speech input. Having all local information—maps, directions, and so forth—will be commonplace. There will be big differences in what people are willing to share, but there will be lots of location-based [applications]. Families will be brought together more tightly; they’ll be able to see where you are, see all your pictures.

There’ll be great things. Education will change dramatically. Communication will be through video, not [just] voice—you’ll be able to participate in conversations, participate in a trip. There’s applications now, I can share real-time video with my mom. In search, we’ll be able to do document understanding, and synthesize information from two Web pages, so we can answer questions [from users]. You should be able to gain insights from data that’s in the world and on the Web, rather than simply finding an expert on the subject. You should be able to do research on the Web.

X: And quantum computing? (Just last week, Google revealed a three-year-old research project to use quantum algorithms to help recognize objects in photos.)

AE: I love looking at the quantum stuff. It’s potentially exciting, but it’s still pretty far off. We’re in a good position. If that research actually yields fruit, we’ll jump on that.

Gregory T. Huang is Xconomy's Deputy Editor, National IT Editor, and Editor of Xconomy Boston. E-mail him at gthuang [at] Follow @gthuang

Trending on Xconomy

By posting a comment, you agree to our terms and conditions.

  • Monica Colangelo

    This gentleman evidently has no idea whatsoever what translation is about. Not now, not in a trillion years will machine translation be half as good as human translation done by a professional. The fact that most words have exact counterparts in other languages under no circumstance implies that you can just use a dictionary, so to speak, and translate anything into another language. As a matter of fact, even common phrases are expressed in an entirely manner in different languages. Knowing what the source text means and how a native speaker of the target language would put it is something no machine will ever be able to figure out.