Google’s Cambridge Office Assumes Growing Role Inside Search Giant

If you glanced at the software engineering job listings page for Google Boston, you might think that the company’s Kendall Square office has only two positions open. That would be wrong. Site director Steven Vinter says the office has been hiring aggressively since December.

The growth would have started sooner if it hadn’t been for the recession, which didn’t slow Google’s expansion much, but did seem to engender a kind of constipation in the local software community, with people unwilling to risk leaving their current positions, Vinter says. But conditions have eased and the company is now getting more resumes. In any case, the two positions described on the job page—software engineer and software tester—are just roles, Vinter says. The company is hiring many people to fill each one.

In fact, with more than 100 engineers and 100 business development staff spread across four floors at Five Cambridge Center—space the company occupied in early 2008 after outgrowing its cramped quarters at One Broadway—Google Boston has evolved from a mere outpost of Google’s Bay Area headquarters into a major engineering and sales center. I stopped by a couple of weeks ago to hear the latest about the office’s progress from Vinter, whom I last interviewed in depth way back in November 2007. (At that time, the office had half as many people.)

The main point Vinter made, as you’ll read below, is that Google Boston is now big enough to have what he calls “end-to-end” responsibility for major parts of the Google product lineup, including the Chrome browser and operating system, the YouTube server and client infrastructure, Google Book Search, and the Google Friend Connect social Web service. As Xconomy founder Bob Buderi has argued in his book Engines of Tomorrow and elsewhere, it’s crucial for corporate outposts to have this kind of responsibility and autonomy if they want to avoid becoming marginalized within their own companies. My impression is that Vinter has been working hard to make sure that Google Boston isn’t simply a vehicle for hiring talented New England engineers who don’t want to move to Mountain View, but that it builds teams that have a direct impact on Google’s bottom line and on the problems the company is trying to solve.

Google Boston PosterWith major news about Chrome, Chrome OS, Friend Connect, and other products expected later this year, it’s likely that Google Boston’s profile within the company will keep rising. That may be true within the Kendall Square neighborhood as well: Vinter told me he admires Microsoft’s efforts to open up its New England Research and Development Center for tech-community events, and says he’d like Google to be more active in this area. Here’s a writeup of our conversation.

Xconomy: Other than your big move into the Cambridge Center space, what have been the biggest changes since we had that long talk back in 2007?

Steven Vinter: There are two big things. All throughout 2009 we were looking for more candidates to hire, and the thing we didn’t really understand was why there seemed to be so few people making it into Google. We didn’t understand why we wouldn’t have seen a continuous flow. Looking back, I think there were just a lot of people [who were] really uncomfortable with moving. The economic problems, in the same way that they affected consumer confidence, affected people’s concerns about wanting to go out and try something new. But that just disappeared around the December time frame, and we haven’t seen such an influx of talented people since I arrived here. So the challenge for us now is basically to make sure that the projects we have keep step with the level of the incoming folks.

X: During that lull, were you actually getting fewer resumes, or was it that the quality of the applicants was below what you wanted?

SV: More the former. There were just fewer people in the pipeline. One thing that’s helping is that we are aggressively seeking new college grads here. Obviously there is a huge wealth of talent in Boston, but in previous years I was more concerned about having a senior team here before we supplemented that with a lot of people coming out of college. So we really opened that up, and we’ve got a lot of people who have accepted positions but who haven’t gotten started yet because they’re winding up their academic studies. We’ve also significantly increased the number of interns we’re going to have here this summer, as part of increasing that pipeline of interest and experience for students.

Another factor is that if you look at people who are working here they really come from three different sources. One is new college grads. One is people from outside the area who move to Boston. And third is people from inside the area who worked at other companies. There’s a certain critical mass that you have to reach before an arbitrary engineer in the local industry knows somebody who is at Google, and my view is that we have passed that critical mass of visibility in the area. We’ve tried not to aggressively market ourselves; I wanted this office to be a magnet based on the people coming, and I think that approach takes a while to pay off. Now I’m seeing people coming in from the same companies as other people who have already joined.

X: I’m kind of surprised that you feel Google, of all companies, had a visibility problem.

SV: I think Google is very well known, of course, but it isn’t well known in this area. It may have been a bad strategy on our part not to be more visible. I was here for two years and I still had people coming to me and saying “I didn’t know Google had an office in Cambridge.” Having somebody you know who works here makes you think differently.

X: What’s the other big change?

SV: The other factor is having the critical mass in an organization. A very common thing in the industry is that when you have 15 engineers you have a critical mass for a team to have deep involvement in something. And when you have over 100 engineers you have the ability to have multiple teams of 15. Suddenly you are able to talk with candidates about having a significant impact on projects. If you come in an interview and you want to work on X and we have three people and Mountain View has a hundred, that’s not significant; there is a legitimate concern that a small team is not going to have a big impact. But when you get to teams of 15, from a management standpoint, it feels like you are offering a stable team with lots of opportunity.

So there are a lot of consolidated teams that work here now. The tendency is to think of sites that are not headquarters as being marginalized in terms of the projects they are doing. Sometimes people think of us as being maybe more research-oriented or future looking. But I view impact as our primary goal in the kinds of projects we take on. The team in Mountain View wants to expand here because they believe there’s good talent here. So now it’s much easier to work on something and expand independently. Teams collaborate heavily with Mountain View, but the credibility we’ve established offers enormous opportunities at a time when Google is growing.

For example, we’ve branched out from Chrome into Chrome OS. Now that we are doing Chrome OS, with a specific focus on performance and networking issues, it allows us to have end-to-end responsibility. Likewise, when you play a YouTube video, all of the servers involved in showing that are things that we do here, but what we didn’t have until recently is the client that you actually play the video on—we were responsible for the streaming but not the actual playing.

X: It sounds like you’re saying it’s good for this office, and for the company, if you have more ownership of certain products.

SV: I think instead of looking at it from the standpoint of products you need to look at in terms of problems. For example, one of the concepts that Google has been very prominent about espousing is that the Internet is all about speed. When experiences go from 10 seconds to 1 second you get a very different value proposition, and more people using it. So performance, and especially reducing latency, is a fundamental part of the Web’s success. When you play a YouTube video, there are many reasons that cause that video to take a certain amount of time to start, and it’s our job to ensure that that experience is a smooth one despite the constant growth in demand and bandwidth requirements.

In order to have influence on all aspects of performance, you’ve got to understand and be involved in every aspect of the user experience. So you, we could get information about the client side of the YouTube playback, but actually working on that client to make the experience better is a fundamentally different way of contributing and gives us greater depth. It’s being able to solve a bigger problem by having all the pieces of the problem available to you.

X: So what parts of Chrome and Chrome OS are you working on here?

SV: We have core teams for both Chrome and Chrome OS. It’s a big investment. But at the moment there is not a ton more we can talk about.

X: Okay, what other projects can you talk about?

SV: There is an open source project that we work on called Page Speed. The idea of Page Speed is to enable any arbitrary developer to assess the performance of their site and make recommendations about how to improve the performance of that site. That tool is something that is available to Google internally, but we see Google’s success as being dependent on the Internet’s success, and on the Internet’s speed. So making that tool available to everyone is in our best interest. The same thing applies to browsers; we are interested in making not just the Chrome browser but all browsers as fast as possible.

X: What about Google Book Search—that’s been one of your big projects for a while.

SV: Google Book Search is one example of a number of corpuses of information that we do searches for. We also do magazine search out of this office, and also patent search. Basically, we’ve been involved in taking the Book Search pipeline—the projects of taking scanned input and turning it into search results—and reusing that pipeline for other content. So, part of it is being very knowledgeable about the infrastructure of Book Search, and part of it is becoming involved in content from other sources. That’s an area where we have diversified the set of things we work on, building off that core experience we have.

X: And what’s the latest on Google Friend Connect, which was your first wholly home-grown product in the Cambridge office? I haven’t seen any new announcements about that lately.

SV: Friend Connect is run exclusively out of this office, and there are still people working on it. The basic idea is trying to make all of the Internet more social. Any webmaster can easily add a few lines of code—in the same way that they add a few lines of code to allow AdSense ads to be shown—and they can then add social features to their site. With the release of Google Buzz, there has obviously been a renewed interest in Google’s part in providing a richer social experience. I think what we’re doing here is very much in line with that overall effort.

There are a lot of things in the works right now with Friend Connect, so I can’t say too much about it, but the short version is that the social experience is deepening significantly, as Google adds more and more social features to its products. We want to be able to make those things available to webmasters more broadly.

X: Speaking of the social Web: I know that Google Wave isn’t one of your projects here, but there’s been a lot of buzz about it, no pun intended. And since I’m sitting here at Google, I want to ask you for your impressions of it. Personally, I haven’t quite figured out who it’s aimed at, or what it’s good for.

SV: That’s a project out of Google Sydney. It’s phenomenal as a collaborative tool to work on shared documents, in ways that no other product on the market allows. It captures the combination of real-time exchanges on those documents, sharing of the document, and the ability to play back the edit history of those documents. To see the value in that collaborative experience, you need a critical mass of people who are not physically together working on a shared thing, with a combination of real-time and non-real-time exchange. That, I believe, is the real sweet spot for Wave. But it’s a challenge because you do have to have a critical mass for acceptance and adoption.

We use it in my team, locally. And it’s amazing how once you start using Wave, you discover that you’ve been working around problems that had horrible solutions. For example, you have an instant message exchange, and now you want to incorporate pieces of that into a document. But those are two separate tools, and you end up copying and pasting and re-editing as opposed to doing that right inside the document. If you look at the number of rounds a document goes through in terms of editing and sharing, it makes no sense to have individual IM exchanges decoupled from the document and the overall group. That’s just one example where as soon as you see a different way of doing it, you say, “Oh my god, of course you would do it this way.” But for many people, their first experience with Wave is one of an e-mail system, and that’s not where it really shines.

X: Last question. I’ve heard people in the local tech community, particularly in the last year or so, commenting on what a remarkable job Microsoft has done of opening up the space at their New England R&D Center, and hosting events and doing outreach, whereas Google doesn’t seem to be as active in the community. You don’t have an equivalent facility here at Cambridge Center, so it’s probably harder for you to host big events. But I know you do host some things here, so I’m curious about how you feel about those comments.

SV: I think what’s critical to Google’s success here is having this be a place that has a large, vibrant engineering community. I think the contributions of all of the companies in this area that make that happen are good for everybody. The problem we have is keeping people from leaving this area, not competition among the companies that are here. But yes, I see what you’re saying. I think we actually do a lot of things. I hosted an event here two weeks ago for alumni from UMass, where I got my PhD. That was 80 people from the Boston area. And we have events on a fairly regular basis. We do events for non-profits talking about how they can use Google’s free services—we just did one of those for the Boston Chamber of Commerce. But I think you’re right that there is a real value to having shared space that people can go to. I think there needs to be more of that, and I’d like to see it happen here more as well.

Wade Roush is the producer and host of the podcast Soonish and a contributing editor at Xconomy. Follow @soonishpodcast

Trending on Xconomy

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

  • Good job Wade, thanks. I look forward to seeing the Goog out in the community more. We can surely use their participation.