Revelations about governments’ online snooping have been good news for Gabriel Weinberg, founder and CEO of DuckDuckGo—a search engine that doesn’t track its users. He has degrees in physics and in technology and policy from the Massachusetts Institute of Technology.
Wendy M. Grossman: What made you set up DuckDuckGo?
Gabriel Weinberg: I started it a little over five years ago, just intending to build a better search engine. My initial focus was to reduce spam and prevent irrelevant sites from coming up in links, and also to make better instant answers. A lot of times you want stuff from Wikipedia, so that was the first place I tried to give you an answer from. After I launched, I started getting questions about search privacy. When I investigated, I decided I didn’t want to store data.
WG: Why didn’t you want to track your users?
GW: Google has been pretty transparent about handing over data to law enforcement, to their credit. I thought that would be inevitable if we store data. Also, it’s just kind of creepy for the search engine to know so much about you. You have your most personal relationship on the Internet with the search engine—medical queries, where you’re going, all tied back to one person. That’s the case even more now; infrastructure in tracking people online has exploded in the last five years.
WG: Internet companies make money by selling user data to advertisers. How do you make a profit if you don’t collect data?
GW: It’s a myth that Google needs to track users to make money on Web search. The vast majority—99 percent—of the money in Web search is based on keyword matching. If you type in “car,” you get a car ad. We make money the same way.
WG: Who uses your search engine?
GW: Different people prefer different experiences and user interfaces. Google is trying to appeal to the average user—we are trying to carve out a niche for the serious person who knows what they’re doing and wants their privacy protected and a great result. We have servers around the world, and we can see how much traffic is coming in from which areas, so we know our users are about 50 percent United States, 50 percent international. We also do surveys, which show that about half our users come in after reading about us in the press. It varies what the story was about—privacy, instant answers, or just, “Check this out, it’s cool.” The rest find out about us through word of mouth.
WG: Are you still improving the search engine?
GW: We are focusing more on instant answers. When you do a search, the answer often exists on a particular site. The big ones are obvious: Movies are IMDB, for restaurants you might want Yelp. But when you get into more obscure queries, you really don’t know what that site is. It’s our job to figure that out and give you an answer quickly. That requires two things: classification and plug-ins to give you the data in the right format. We are recruiting open source developers to make those plug-ins.
WG: Have the revelations about the National Security Agency’s monitoring program affected your traffic?
GW: We were close to 2 million queries a day before the NSA story broke. Since then, traffic has passed 3 million. We’ve broken records.
This article originally appeared in New Scientist.
