Facebook CEO Mark Zuckerberg has made a point of stressing that the social network’s Graph Search is not a Web search engine along the lines of Google. Rather than scan the whole Internet, Zuckerberg has positioned Graph Search as a tool for searching for people, connections and friend recommendations.
Despite that differentiating tactic, any Facebook search engine would be—and has been—considered a potential rival to Google’s market-dominant product.
During a South by Southwest session on Sunday, Guy Kawaski—former Apple evangelist and newly hired consultant to Google-owned Motorola—asked Google’s svp of search Amit Singhal what he thought of Graph Search.
“Time will tell if people really need that kind of search,” Singhal responded.
Singhal wasn’t outright dismissing Graph Search, but echoing questions many have voiced as to whether Facebook’s search engine will gain adoption as the company rolls it out to more users, and eventually mobile—and whether Graph Search can truly prove itself as a utility. However, Singhal did acknowledge that Facebook’s interest-oriented search, built on the social network’s user data as opposed to open Web data, is the right foundation for the social networking behemoth.
Facebook’s Graph Search wasn’t the only competition Singhal touched on during Sunday's interview. He also made a point to note that unlike Facebook, Twitter, Pinterest and Microsoft, Google was not a victim of a recent hacking spree, all without mentioning any of these companies' actual names. And Kawasaki queried Singhal’s thoughts on Microsoft’s Pepsi Challenge-like "Bing It On" campaign that asks users to take blind search tests of Google’s and Bing’s search engines.
“We focus on our users and building good products for them. … Others should focus on building good products too,” Singhal said.
Earlier in the conversation, he described Google’s long-term vision for what it’s building. “At Google we believe a perfect search engine should know exactly what you need and give you exactly what you want. … Our dream is for search to become the Star Trek computer,” Singhal said. In a nod to Saturday’s keynote speaker SpaceX CEO Elon Musk, who’s working on essentially enabling intergalactic travel, Singhal said that “between [Musk and Google] we’ll build you Star Trek.”
Singhal said the last time he checked, Google crawls over 30 trillion Web addresses, or pages. Kawasaki tried to pry open what Google’s crawlers are looking for in those pages to determine how to display them on the search results page, but Google has been infamously tight-lipped on it search algorithm. Singhal was consistent. To Kawasaki’s point-blank question on how many lines of code the algorithm contains, he said, “Many … more than I can read or write today … a big number.”
(The nonanswer made it hard to read how serious Singhal was when he later related that when he adds a new employee to his team, the whole unit goes out for drinks and asks about the newbie’s past projects. If he talks, “we fire him,” he said.)
Google’s Star Trek ambitions extend beyond the current ability to index the Internet and are spanning into spinning a story from that index. “Having all of humanity’s knowledge on the Web is not enough. We need to understand that knowledge,” Singhal said.
Part of that push is Google’s Knowledge Graph, while another is Google+, typically perceived as Google’s rival to Facebook despite having a fraction of the latter company’s active user base. Singhal said that Google+ aids the company in factoring “the knowledge only accessible to you” into search results. Of course for that to really be the case, people will need to use Google+ as aggressively if not more aggressively than they do Facebook, converting the social signals into personalization parameters.