Recently, Facebook unveiled Graph Search, its new search engine. The early beta application does not work in the same way that Google works – but Foursquare and Yelp might have cause for concern.
Full disclaimer first: I was not one of the lucky ones who actually got to try Graph Search. When I finally got the Graph Search page to work, after multiple attempts, the button I got at the bottom of the page was for joining the waiting list. Instead, I sifted through a lot of commentary and reviews that demonstrated how the search worked and speculated on its future. I concluded that, in principle at least, Graph Search is neither better nor worse overall than Google; it’s simply different.
So how does it work? Facebook Graph Search taps into the various “likes” and other data that users have entered into the social network, and then returns answers based on what it knows about you and your friends, and what information has been made public. So if you typed in “friends who live in San Francisco,” for instance, you’d get a list of all of your Facebook friends who live in that city. That’s great if you’re planning a visit to the area.
Where the search shines, however, is with all of the modifiers you can attach to it. This reminds me ever so slightly of the search engine Blekko, which uses hash tags to modify its searches. Facebook Graph Search seems to use natural language, however, and seems to be trying to reach a different market.
So how would Graph Search work in a real life situation? I live in the Orlando area, but later this month I’m going to see the musical “Wicked,” which is playing in Tampa. I don’t get out to Tampa very often, but I do have a number of friends on Facebook who live there. I could use Graph Search to look for the restaurants in Tampa that my Tampa friends liked, so I can get a nice dinner before the show. Similarly, in a video posted on TechCrunch, Josh Constine demonstrated that one could use Graph Search to find dentists liked by your friends. That’s genuinely valuable information – much better than you can get from Google.
You can also use Graph Search to look for photos, posted by your friends or publicly available. You don’t have to limit yourself to recent photos, either. Remember that awesome Grand Canyon trip one of your beer buddies took five years ago that he keeps talking about? Search for photos taken at the Grand Canyon before 2008 and you just might find his pictures. Or perhaps you’d like to see pictures that are even older? Believe it or not, you can find those, too. The article I linked to recommends somewhat whimsical searches such as “photos of me taken before 1990” or “photos of my parents between 1970 and 1979” (if you’re wondering what they looked like when they were “cool”).
You can also search based on interests. So if you want to form a local bicycle club, say, you can search for people in your area who like bicycling. Or if you’re a big Star Wars fan, you can search for friends who share that interest. You can even combine interests, to search for friends who enjoy both Star Wars and costuming (to invade your next science fiction convention in group costumes, perhaps?).
And of course, you can search for places. This is why I said that Foursquare should be worried. You can search for places your friends have been. You can search for photos of places, as I’ve noted above. You can even just search for places by city.
But what if your filtered Facebook Graph Search doesn’t yield any results? The social network has negotiated a deal with Bing . Facebook CEO Mark Zuckerberg said that he would love to work with Google, but the search giant wasn’t quite flexible enough to protect Facebook’s members’ privacy. To do that, according to Zuckerberg, you need the infrastructure in place to quickly take down photos and such when users change their privacy settings. Apparently Bing offered that, and Google didn’t. (If any of this sounds more than a little fishy or laughable to you, well, you’re not alone).
Just how well will Graph Search work? Steve Cheney points out one of the social network’s stumbling blocks: dirty “Likes.” How many times have you wanted to compete in a contest or get a free sample from a company, but you could only do it if you “liked” the firm’s Facebook page? I’m pretty conservative about that myself, and I know I’ve done it a number of times. But those aren’t real “likes,” if you know what I mean. What kind of answers will Graph Search return when it’s full of dirty data like this? Not very good ones, I’m sure…which is one reason I said that it’s not a Google killer.
On the other hand, advertisers will no doubt appreciate Graph Search, once Facebook figures out how to monetize it. Right now, though, I’d look at it as one search method among many for finding the kind of information you need. Cheney said it best: “Offline we consult different places, people, and resources, and you will do the same with social networks and web services online.”