How Well Do Social Networks Know You?

How Well Do Social Networks Know You?

This article by our Head of Behavioural Science, Patrick Fagan appeared earlier on Liveminds.

Have you ever met a couple and knew it probably wasn’t going to work out? There’s a science behind it.

Psychologist Dr John Gottman interviewed newlyweds for fifteen minutes and videotaped the interactions. He then coded the tapes for certain behaviours, like stonewalling and defensiveness. From these behaviours, Gottman was able to predict whether a couple would get divorced within the next 3 to 6 years. This was done to an accuracy of 83%. (Contempt was the biggest predictor, in case you’re wondering.)

All it takes is a small glimpse into someone’s psyche to predict how they’ll behave. In psychology, this refers to a principle known as “thin slices”: If I gave you just a thin slice of a birthday cake, you could tell me, very well, what the rest of the cake looks like.

In the same way, as humans we make first impressions based on tiny, but accurate, cues. One study found that exposure to just a picture of a person’s face was enough to estimate their personality to around a 70% level of accuracy – and this after just a tenth of a second! What’s more, taking more time did not improve the accuracy of the guess. In other words, first impressions are on-target and long-lasting.

A lot of other research has supported this finding, showing that brief glimpses of people’s faces are useful clues to behaviour – for example, faces can be used to infer sexuality and violent tendencies. However, it’s not just your face that provides clues to your behaviour; among other things, personality can be accurately read from your bedroom, your clothing, and your music tastes.

The key is that our actions are the result of underlying dispositions – usually both biological and neurological in origin. These cause us to act in a particular way. Take extraversion, for example. Extraversion is associated with increased volume in the medial orbitofrontal cortex, a part of the brain which processes rewards. In other words, extraverts are more reward-sensitive. This goes some way to explaining why they are happier, more active, more excitement-seeking, more risk-taking, and so on.

This disposition manifests itself in observable behaviours which can be used to infer that someone is extraverted. For instance, extraverts are more likely to smile, ostensibly because they are more positive; listen to energetic music like dance, because they seek out sensation; and wear faded shoes, because they are more active and outgoing.

Once this personality trait has been inferred, it can then be used to predict and influence other behaviours. For example, if you are a recruiter, and you see that a candidate smiles, has faded shoes, and is wearing a Chemical Brothers t-shirt, you can infer they are extraverted, and therefore will prefer more entrepreneurial and social job roles.

In the digital world, you can customise your messaging to different audiences in the same way: a peer-reviewed study found that Facebook adverts customised to personality result in significantly higher conversion rates.

Fortunately, you don’t need to see someone’s face or inspect their bedroom to know their disposition: the cues they leave online are more than enough. In a landmark study, Cambridge psychologists discovered a relationship between Facebook ‘likes’ and personality (as well as other traits like sexuality; intelligent people, for example, were more likely to ‘like’ The Colbert Report and curly fries).

This kind of digital footprint can be used to understand someone’s deepest psychological drivers and predict highly personal behaviours. If a brand knows, from your purchase history, that you own a dog and like traditional art – and are therefore not open to experience  –  then you’re more likely to vote for conservative politicians.

So how well do social networks know you? Probably better than most of your peer group, to whom you would never reveal these deepest parts of your psyche. In fact, a study found that your Facebook profile is better at estimating your personality than is a work colleague, a friend, a flatmate, or even a family member…the algorithm came a close second only to your spouse.

For brands, the recommendation here is that there is a whole wealth of information about consumers available in their digital footprint. This big data from social networks, and beyond, can allow you to hyper-target your required audience, and see improved performance metrics from your messaging.