Is a like just a like? Facebook ‘likes’ are such simple social interactions but according to new research, the very act of liking something reveals all sorts of information about us.
Predicting your personal details through Facebook ‘likes’
A bunch of clever people at the University of Cambridge recently designed a machine-learning system to predict Facebook users’ personal details and personality traits based on their ‘likes’.
The results took the boffins by surprise: the system distinguishes between black and white users 95% of the time and separates men and women, gay and straight men and democrats and republicans with uncanny accuracy. It even predicts subtler stuff like whether a person is a drug user or has divorced parents, with slightly less but equally impressive accuracy.
A bit too ‘Big Brother’ or a jolly good thing?
When a marketing campaign taps into our desires, popping up exactly where and when we need it, it’s usually a positive experience because we get what we want. When it’s irrelevant, mis-timed and poorly targeted it’s nothing more than annoying junk. From a marketing perspective the Cambridge team’s findings are good news for consumers and businesses. Unless you’re uneasy about organisations collecting information about you, using it to define marketing campaigns, inform new product development and set prices, in which case you’ll be feeling extra-paranoid right now!
Is it magic?
No matter how many hard facts a business knows about you, it can’t predict your buying behaviour reliably. Humans are too emotional and irrational – inherently unpredictable. Big data isn’t a silver bullet. But it’ll definitely help marketers create better customer and prospect profiles, sending relevant offers to the right people, at the right time, in the right place with the right message. All of which makes sales conversion more likely.
Facebook’s data-gathering future
Data giants Epsilon, Acxiom and Datalogix are collaborating with Facebook to collate the vast amounts of information collected about account holder activity and make marketing sense of it. The network has also launched Graph Search, a semantic search engine combining external data with a information from a billion users to provide natural language search. It’s set to make tapping into all that lovely, rich data even easier.
Who will be using the data?
Insurance companies are interested in mining Facebook’s big data to see if they can separate high risk customers from low risk and tailor their prices accordingly. More will follow, some more controversial than others. Depending on your world view it’ll either improve marketing beyond your wildest dreams or compromise individual privacy to an unacceptable degree. The truth probably lies somewhere in between.
Update Sept 2014: Facebook continues to gather data, and seems to be annoying more and more users along the way…