At a major event, Facebook recently unveiled Graph Search, a new discovery tool that would take advantage of your friends’ data to provide better recommendations and results from, well, your friends, rather than the anonymous mass that is the average Google inquiry. The tool just opened to early registrants yesterday and is still relatively early, meaning it lacks nearly all the data from third-party apps. But placing all that data in this tool can’t solve the simple fact that Facebook’s new Graph Search is fundamentally useless. Underwhelming is the word I’ve seen across a swath of reviews to describe Facebook’s new toy and it’s one that fits aptly here. It could not be more underwhelming.
What Facebook Is Doing
So if you were so inclined, what would you use Graph Search for? Facebook thinks you’d search for people, places, or things, like your friends’ favorite restaurants, or maybe their favorite movies, or what have you. So let’s try a few examples.
Suppose you were to ask Graph Search what your friends’ favorite movies were. You’d get a lot of results, sorted by which movie has the most Likes. You see, since Facebook has no live wire to when people do many of these things, such as watch a movie or love an album, you’re going to get the accumulate result of everything they’ve ever liked. Do you want to know what their favorite movie was this weekend? Can’t do that. What if your friends don’t Like a particular movie? You’ll simply never know if you’re compatible. Then you have the OCD-affecting issue where someone created a ‘Boondock Saints’ page and someone else made a ‘The Boondock Saints’ page. So what does this inquiry really get you? An answer that isn’t very useful.
What if you ask it who likes Star Trek and currently works at your former employer? I got two results back, both containing people who haven’t worked at that employer for years. But why would you even ask that? Why would you ever ask Facebook who, at your former employer, likes Star Trek? The only possible answer can’t be very useful unless you’re secretly enlisting Graph Search as a dating tool, in which case I won’t judge you.
What about my friends who like Jay-Z? None, because apparently none of them have Jay-Z liked.
What about my friends who like a particular song? Can’t tell, that data’s not available, yet.
Facebook, show me some photos of my friends from June 2012. Oh, that works. That’s pretty cool.
How many of my friends have been to a place called ‘In N Out’ lately? Can’t tell, it has no time data, but it does show me people who have been there recently.
How many places has our very own Kelly been to? According to Facebook, only six, despite a healthy quantity of Foursquare check-ins.
What Facebook Should Be Doing
You’re getting the idea: this is a pretty lazy search of data my friends have available. Facebook punched a hole in the corner of that info to run another chain through so I can get collated responses to uninteresting queries. This is exceptionally boring and mostly useless. Unfortunately, Facebook unveiled this as a major ‘third pillar’ to their News Feed and Timeline features, both of which are actually incredibly useful.
This is not.
Facebook should be following in Google’s tracks here: this should be predictive search like Google Now, not the end result of questions you’ll never ask. What if you had a Facebook tool that made recommendations to you based on what your friends liked and where you were, when you were there? Doesn’t that sound like a tool you’d actually use instead of something that could’ve come out years ago? Imagine the possibilities that a billion users’ worth of data could produce! You could use some kind of Facebook app to plan a trip and all of this data just flushes to the surface, ready to provide you with good decisions.
Maybe this will become more useful when third-party data starts seeping in, but you can keep your horses held in the meantime.