When Google Plus came out, everyone said that Google was finally taking the fight to Facebook. Now it seems that the shoe is on the other foot with Facebook firing off a salvo, which may end up putting a serious dent in Google’s armour when it comes to online search.
This week, Facebook announced that they were slowly rolling out their own search feature, called Graph Search. For an initial explanation, here’s Facebook themselves telling us all about it.
So in other words, when Graph Search fully rolls out to everyone, we will be able to get personalized search results that Google can only dream about. Why? Because when Google “personalizes” your results, its algorithim is limited. It doesn’t really know you. It doesn’t know your likes and dislikes when it comes to food, movies, music. So in a way, when it tries to deliver personalized results to you, it has to guess to a certain degree.
But Facebook on the other hand, they know everything about you – because you willingly tell them in status updates and in your profile. Facebook knows who your friends are, who the friends of your friends are, what your interests are and the interests of your friends, and friends of friends. It knows what movies and music you liked (based on what pages you “liked”). So Facebook has a more complete picture of who you are as a person and can more accurately send you personalized results via their Graph Search. If you say to Facebook, “tell me a local restaurant that one of my friends like”, they would have that information ready and available. If you asked that same question to Google, well good luck sorting through all the junk that you’d be given.
It’s the same with Amazon and other services such as say Netflix. They can deliver more personalized results because they have your account history to fall back on. They can see what you have liked and not liked. But if you were to type into Google “tell me a new movie that I am going to absolutely love”, Google won’t have a clue. How could they? They have no previous information to consult, other than your search and email history (and that won’t indicate movies you either like or dislike). Amazon and Facebook on the other hand can offer you more relevant results because they can see what you have ordered previously and “liked” previously, respectively.
Personalized search really is the next generation of search and with Graph Search, Facebook has ignited an earthquake under their competitors.
So in lots of ways, Facebook Graph Search is going to give Google a serious run for their money, if we stop howling about privacy violations by Mr Zuckerberg. For Google to then seriously compete, they are going to have to tweak their Google Plus service to collect more information about their users and encourage users to tell Google more about them, so their algorithim can adapt accordingly. They would then have to match up that information with the information of your friends in your Google Plus circles. That’s going to take a while, and instead of Google always being out in front, way ahead of everyone else, instead it will look as if they are playing catch-up for once.
Meanwhile, Graph Search, with the user information already in place, will start generating buzz and user loyalty, especially if Graph Search is rolled out to Facebook’s mobile apps. Imagine being out and about, suddenly hungry, and you decide you want a restaurant recommendation. You could then ask Graph Search “what was that restaurant Bob recommended last month?” and your friend Bob’s status updates would then be searched for that recommendation. Try doing that with Google. When I tried looking for a nearby kebab restaurant with Google+ Local, it told me the nearest one was 20km away – when in fact the nearest one was just around the corner (and yes, I had GPS switched on).
Facebook Graph Search, in its current beta form, is proving a little underwhelming for a lot of critics (although the underlying idea is exciting). But give it a year or so, and by 2014, we could see a serious rival to Google’s search. Just think about how many users Facebook has, and then think about this – would those users want a recommendation from a trusted friend or from a faceless algorithim that is just going to guess? I know which one I would prefer.