You may not know who Lars Rasmussen is. But chances are, more than once he helped get you where you were going. That’s because Rasmussen, along with his brother, invented what is now Google Maps. He then developed Google Wave—an online collaboration tool meant to replace email that flopped—before leaving Google for Facebook.
Eight years after the launch of Google Maps, Rasmussen is once again out to change how we find things. On Jan. 15, the 44-year-old Dane joined Facebook CEO Mark Zuckerberg to present the company’s latest, and likely most radical tool—the inelegantly named Graph Search that allows users to search the 1 trillion connections between Facebook’s 1 billion users. A few days after the announcement, which sparked excitement, curiosity, and privacy concerns, Newsweek spoke to Rasmussen about how you build a product “the Facebook way,” about what a toothache can teach you, and what not to tell “Zuck.”
How will it change the way we use Facebook?
Mark described Graph Search as a third pillar of [Facebook], where the two existing pillars are News Feed and Timeline. So you can look at it like this: News Feed is the daily answer to the question “what’s going in the world?” Timeline is the answer to the question “what’s up with this one particular person?” And then this new pillar is “ask us any question you want.”
Back in the early days, before search really worked, the Web was still supremely useful. But when companies started making search work, the usefulness just grew dramatically. We’re hoping that, by adding a very useful search mechanism to Facebook, we’ll see the same effect.
One of the demos I did was I searched for dentists my friends like. This is actually a true story: I had a toothache and I needed a dentist because I’m relatively new to the area, and I found a really good one by just looking at what dentists my friends liked in the area. When I had that really nice experience at the dentist, I felt a couple of things: grateful toward the friends who shared that they liked that dentist, and the urge to go and share with my friends what things I like.
So it’s almost contagious?
Yeah, exactly ... another thing I felt was grateful to the dentist that she had given me this experience, and I wanted to help her succeed in her business, so that was another reason I went and clicked that “like” button.
What’s it like to build something like this at Facebook?
A couple of months after I started at Facebook, Mark took me for a walk—Mark likes taking people for a walk—and he explained how important he thought it could be ... if we really made search work on Facebook. As you can tell from the things I’ve worked on in the past, I like working on reasonably ambitious and somewhat risky things. And so I thought it was cool.
He argued that, primarily, the Web is a very large collection of documents, and so keyword searches are appropriate for that. But the data that people share on Facebook has a lot more structure to it. Mark wanted to be able to ask more structured queries, like “What restaurants do my friends like?” or “What Indian restaurants do my Indian friends like?” which is a query that you can’t really express in keywords—you really need some other mechanism than keywords to express it.
You once said that with Google Wave you wanted to recreate the atmosphere of a startup. Did you do that when working on this project?
I pulled back a little bit on that theory. I think we took it a little too far at Google. When I did Google Wave, everyone had to be in Sydney, and a lot people actually traveled there to be part of it. There was a lot of isolation. There were a lot of things we kept secret from the company while working on Wave—just like you would at a startup.
We don’t have any of those things going on here. But I think at Facebook, in general, we like to have—and this is kind of a trend in the Valley—when a company grows, you try to maintain the startup feel for each team.
The problem we encountered with Wave at Google was that we became very isolated from the rest of the company. And in the time that it took us to build Wave, the rest of the company changed direction. I think that has a lot to do with why the product failed. By the time we kind of emerged from hiding, Google had changed its strategy toward Google Plus ... and Wave wasn’t superaligned with that. And I think that happened because we were so isolated; because of this desire to be a startup. There’s much better balance here with my current project.
What does it take to be an innovator?
My favorite analogy is that of an explorer back in the days. Most people—and I’m not trying to be critical here—but most people, before they go on a journey, they look on a map and make sure they know how to get there. They make plans, make sure they have enough gas in the tank or money for the tickets, before they start on a journey. But if you’re an explorer, you have to be willing to let that go because there is no map of uncharted territory. You have to be willing to embark on a journey without really knowing how you’re going to get to the end, or even where the end is, and what you’re going to encounter along the way. And it’s not just that you have to be willing to do that, you have to be more excited about an endeavor like that than you are about just following a path on a map. I know it’s ironic that I have this analogy because I happen to work on maps, but this is what it’s like.
You want to be working on the outskirts of what you can actually do, which is quite outside of what you think you can do. If you take it too far, of course, then you fail. And you have to be willing to live with that risk. For me, Wave was clearly too far away. It was too ambitious; I wasn’t able to pull it off. For whatever promise there was in that idea, and I think there was a lot of promise, I just wasn’t able to make it work.
So what’s next for Lars Rasmussen?
Graph Search is next for me. We’re just getting started. Now the fun begins because the rate at which we can improve the product will go up dramatically now that we can start getting actual feedback and actually look at how people use the product.
How complete is Graph Search at this point?
One percent? We’re really just getting started. Think of the quality of Google Maps when we launched it back in 2005 compared with what it is today. It’s a pretty dramatic difference. This is just like the dipping-the-toe-in-the-water move we’ve done here.
What’s the next big thing in technology, at Facebook or elsewhere? Who’s doing something that excites you?
Oh goodness, the next big thing? If I knew, I’d probably be working there. [But] don’t tell Mark I said that.