Facebook’s Predictive Analytics – A Love Story

Did you know that Facebook can look at its user’s posting patterns and moods in order to predict their future romantic relationships?

As a posting on the blog FierceBigData put it on February 19, 2014: “‘During the 100 days before the relationship starts, we observe a slow but steady increase in the number of timeline posts shared between the future couple,’ writes Facebook data scientist Carlos Diuk in his ‘The Formation of Love’ post.

‘When the relationship starts (day 0), posts begin to decrease. We observe a peak of 1.67 posts per day 12 days before the relationship begins, and a lowest point of 1.53 posts per day 85 days into the relationship. Presumably, couples decide to spend more time together, courtship is off and online interactions give way to more interactions in the physical world.’ “

Facebook as Cupid?

To put it another way, Facebook knows before you do when you are going to become a couple. It certainly knows long before you announce it on your relationship status.

What’s more, Facebook can tell when the physical part of your relationship begins because your online activity decreases. While Facebook diplomatically calls this phase “courtship” in the posts, but in reality the courtship actually occurred during the exchanges Facebook initially traced to accurately predict the relationship.

What value does tracking the love life of its users have for Facebook? Plenty! This information can be used to promote placement of marketing for everything from flowers to chocolates to condoms to lubricants and any other company that can profit from your being involved in a romantic and physical relationship.

Predicting Breakups

Facebook also can predict when your relationship is going to end – often before you do.

In a Facebook posting called “When Love Goes Awry,” which was published on the site on the day after Valentine’s Day, 2014, Facebook data scientist Adrien Figgeri wrote: “To conclude this week of celebrating love and looking at how couples blossom on Facebook, we felt it was important not to forget that unfortunately sometimes relationships go south and people take different paths in life. In this context, we were interested in understanding the extent to which Facebook provides a platform for support from loved ones after a breakup.”

“To that end, we studied a group of people who were on the receiving end of a separation, i.e. who had been in a relationship for at least four weeks with someone who then switched their relationship status to Single.”

“For every person in this group, we tracked a combination of the number of messages they sent and received, the number of posts from others on their timeline, and the number of comments from others on their own content, during a period starting a month before the separation to a month after.

“We observed a steady regime around the baseline before the day of the relationship status changes, followed by a discontinuity on that day with a +225% increase of the average volume of interactions which then gradually stabilize over the course of a week to levels higher to those observed pre-breakup.”

To translate, Facebook can accurately predict when your relationship is going to end long before you suspect anything is wrong. And you can be sure Facebook is using similar analysis to predict other intimate details about your life as well.

Facebook and the Possibilities of Big Data

Most people have no idea about Facebook’s date collection activities and data usage practices. Even if they have heard about it, they generally don’t care or see it as little more than a harmless annoyance.

They mistakenly believe that as long as they don’t post anything of a highly personal or sensitive nature on their wall, Facebook’s practices can’t cause any harm to them.

Facebook isn’t alone in conducting wide scale data harvesting. In fact, most corporations do it to some degree or another these days. While some of these efforts are indeed harmless, some are not. What’s worse, at this point nobody knows how this data will be utilized in the future, or to whom it will be sold and for what purposes it is being purchased.

Most people are legitimately concerned about the potential dangers of the government collecting large amounts of data on them, but they are blind to the dangers of private corporations doing the same thing.

This is a perfect example of how limited perception based on an individual’s personal experience can cause ambiguity in the discussion of big data within and organization.

But Facebook’s use of data offers an excellent opportunity to illustrate what is possible with big data, as well as what some of the potential dangers to privacy are. It can broaden the thinking of people at every level within your organization.