Facebook Graph Search: the next big thing, or next big failure?
Sitting in a room full of social media marketers talking about how to best use a social network for marketing isn’t among my favourite tenses to do. However, Facebook graph search is intriguing – I honestly can’t work out if it is an interesting experiment or actually something that could change the way we search.
Which was why I enjoyed the fact that the panel opened by asking the attendees to raise their hands if they had used the tool. About half had. Half of a room of people who had come explicitly to hear people talk about graph search. The second question was even more telling – how many of those that have used it actually found it useful. Less than half again. Not particularly promising….
If you have somehow missed the hubbub about Facebook Graph Search, it’s essentially a tool that allows you to see things you and your friends have shared on Facebook and find things such as restaurants or shops that your friends have liked.
Most people have been on Facebook for almost half a decade. That means Facebook have been storing data about you for five years. That’s an awful lot of data. They probably know more and you and your friends than you do.
Graph search is an attempt to start putting that data to use, for users. But the problem is, it’s probably more useful for brands than for users at the moment. And that does not bode well for its adoption.
Currently the results are based on the number of interactions such as checkins and likes. But over the years the value of a like has been eroded to the point where it’s a lowest common denominator, often given out as a platitude rather than an expression of truly liking something. And just because I check into a place, or my friends do it for me, doesn’t mean that I necessarily like the place, or am endorsing it. I wouldn’t check out of a restaurant if I had a bad meal, so from Facebook’s point of view, I must have enjoyed my time there. To make it meaningful Facebook need to introduce a quality score (as google have for years with page rank), more than just likes. Aside from anything it is just too easy to game likes, purchasing thousands of likes has never been cheaper. But short of adding an unlike button, I am unsure of how this would work.
Arguably Graph search is a massive experiment at the moment. It sounds amazing in theory, but it’s not there yet. And we don’t know if that promise will ever be delivered.
But is there a need for this product anyway? A lot of of our Facebook friends aren’t even our friends anyway. They are acquaintances picked up through life, the people you meet in bars, friends of friends, people who probably don’t know you that well. They aren’t necessarily people whose recommendation you want. And even if they are real friends, as much as I love whole foods I don’t need to know that five of my friends also love whole foods.
We aleady have access to huge amounts of data that we aren’t doing anything with. We’ve had api access to the graph for a ages and no brands seem to be doing anything with it. Search graph isn’t actually anything new for marketeers, it’s more for users. Shouldn’t we be finding ways to use the data we already have rather than getting excited by the theoretical uses of new data?