No other company in the world holds as much personal data as Facebook, which puts it in an ideal position to cull data from your friends and friends-of-friends, and even those you don't know, to bring you answers to your deepest questions. Such as, "What sushi restaurants have my friends gone to in New York in the last six months and Liked?"
Ok, maybe that's not your deepest questions, but it's the example that Facebook CEO Mark Zuckerberg used during an interview at the TechCrunch Disrupt SF conference earlier this month when he was talking about the intersection of Facebook and search.
"Search engines are evolving" to "giving you a set of answers," Zuckerberg said. "Facebook is pretty uniquely positioned to answer a lot of questions people have."
You could call Facebook's notion of search "friend mining," extracting specific answers to question by mining the immensely data-rich social graph. Zuckerberg was coy about when Facebook would launch a more capable search engine, but allowed, "At some point we will do it. We have a team working on search." His remarks certainly got the attention of Facebook watchers and Wall Street analysts, who have been speculating about how Facebook, with its nearly one billion users, could harness all that personal data and generate Google-like revenue from search.
"People have been waiting for years for Facebook to jump in to search," said Jed Williams, a senior analyst with interactive media research firm BIA Kelsey. "This is such a big opportunity." Search is an $18 billion business in the U.S. alone, according to research firm eMarketer.
At this point, people don't associate Facebook with search, despite the fact that the company says it processes 800 million queries a day. The vast majority of those queries is people searching for other people, rather than the kinds of search queries common to Google and Microsoft Bing.
You can search the broader Web from within Facebook, but Facebook doesn't make that task easy. Type 'sushi' into the Facebook search bar, for instance, and you'll get a list of Facebook pages about sushi, based on popularity and which ones your friends 'like.' Then, lower on the page, a little box lets you search outside Facebook's walls via Bing, with which Facebook has a partnership.
According to August 2012 ComScore search data, Google had 66.4 share of the U.S. market and Bing 15.9 percent. Yahoo, which also uses Bing search, has 12.8 percent. Microsoft wouldn't disclose how much traffic it gets via Facebook.
While Bing has Facebook integrated into its search engine -- so that what your friends like on Facebook shows up alongside your Bing search results, and you can easily post to Facebook to ask a question -- presumably Zuckerberg's search team is working on something that would derive a set of answers from Facebook first and wouldn't require a visit to Bing.
Like Wikipedia, Quora's value comes from those who decide to participate. But Facebook wouldn't need people to weigh in before it gives you useful answers to all sorts of questions -- because, in effect, they already have. They've 'liked' restaurants, movies, bands and perhaps a page for the Toyota Prius. They ranted and raved about services. They've told Facebook where they work, lived, and on and on and on. And they've been doing it for years, with no sign of slowing.
Facebook's job is to determine what data points matter for what type of question. That's no small task. Eventually, Zuckerberg envisions a question-answer-search service that not only answers basic questions about which restaurants your friends have 'liked,' but one that can offer up meaningful responses for less straightforward questions, such as this example Zuckerberg gave: "Which of my friends or friends of friends work in a company that I might like to work at?" That's not a particularly tricky question when you think about what data Facebook has and how it could mine it to deliver useful answers. To answer the question Facebook would need the right set of data, data you would see on another social networking site, the business-focused LinkedIn.
Yelp also offers clues about what Facebook's possibly cooking up on the search front. Michael Pachter, a Wedbush Securities Internet analyst, even suggests that Facebook could take on Yelp big time. While Facebook currently doesn't have Yelp-like reviews, Pachter points out that it could amass them quickly.
"If I write on Facebook that I had a phenomenal meal, Facebook could have a text box pop up that would ask me if I would like to rate it, give it a thumbs up," said Pachter. "It could be a full-fledged Yelp." And that would give Facebook more data points, including from its bazillions of images, for its friend-mining version of search, providing a big opportunity for snagging local ad dollars.
The importance of mobile is why Danny Sullivan, editor SearchEngineLand.com and a CNET contributor, says he expects Facebook search to resemble Foursquare's 'explore' button. That lets users search for, say, nearby restaurants or stores based on your location and other criteria, including where you've checked in previously. Foursquare also surfaces recommendations based on users reviews, and it shows where your friends have checked-in, what reviews they have written.
Searching for revenue
The crowd that's super excited by Facebook and search is Wall Street. Facebook's beaten up stock, still roughly 40 percent off its IPO price of $38, has been on the rise ever since Zuckerberg's interview. It closed Friday at $22.86 a share, up about 30 percent from its low of $17.55. One reason cited is Zuckerberg's enthusiasm for search, which, rightly or wrongly, Wall Street equates with Google and fat profits. Google made $2.79 billion in the last quarter, dwarfing Facebook's earnings.
For sure, Facebook social search offers a big money opportunity on several fronts. The market for local mobile ads alone -- think Facebook adding Yelp and Foursquare-like features -- is expected to hit $1.3 billion this year alone and quadruple over the next few years, according to research firm BIA Kelsey.
More broadly, Facebook could start selling ads -- or its so-called Sponsored Stories -- against queries. One possibility, according to Pachter, is that Facebook could let advertiser drop deals into search results based on keywords. If you're planning a trip to Hawaii, for example, you could ask the question, "Where's a good place to visit in Hawaii?" Facebook could then mine your friends' status updates and photos to bring you their past vacation haunts. Facebook could then insert Hawaiian hotel packages in these results.
Whatever Facebook comes up with as it builds out its search products, it needs to tread carefully when it comes to new ways of using user data. Ultimately, it's all about making a search that becomes a seamless part of the Facebook user experience. Because, if it's not, Facebook will have some unhappy users -- and then it will have fewer friends to mine.