The Lost Potential For Social Media Sites

Facebook can be a lot of fun. It’s stimulating, you’re reading and thinking up things to say, there’s pictures, videos, games and apps. The timeline feature really works with the linear way people think about their life – from beginning to now. The whole thing is well put together to massage a number of human tendencies that keep people logging in. The site learns from your actions and presents you with stuff that may be interesting to you. If their goal is to make the user experience better, they have succeeded. Some of the comments my friends friends make are funny, “liked” and over time more and more of their stuff appears.

What I’m not sure about though is just how much of the actual information they have that they are giving us. With so many users, there have got to be 100000’s of patterns of behaviour emerging. Liking someones comment does indicate a certain level of interest. Then viewing their profile and  clicking on some of their links will indicate a higher level of interest.

But what does looking at their profile 4 or 5 times in a month mean? What level of interest does that indicate? Then they become friends, and a few weeks later they are tagged together in a photo with someone who works for a particular bank. The bank person “likes” the photo. 2 weeks later, the wife’s profile of the bank person is checked by one of the 2 people who were tagged in the photo.

What will the outcome be in 6 months?

There isn’t enough information there for us to guess, but there is enough data in FB to make a prediction about who will be “friends” and be “liking” the stuff in 6 months. I’d wager some money that one of their data miners would be able to tell us when two people are about to break-up based on their FB actions. And I think the miner would know there was a problem with their relationship well before either one of them knew.

As interesting as it would be to have a dialogue pop up and say “looks like you are about to experience some relationship trouble – click here to view report or click cancel to keep your head in the sand” it would END FB because being stripped naked by the obviousness of your actions is chilling. We’d like to think that we are holding our cards close and out of sight, but our actions reveal our cards. FB remembers all of your actions so it will be able to identify when you start to repeat a pattern; any pattern.

It will also be able to compare your pattern of actions of other patterns and be able to make predictions about the future outcome. The message box saying “there’s a lot you need to know about how you are going to interact with your new friend” or “we noticed that you changed your relationship status to say that you were in a relationship with someone who isn’t on the site. based on some of your new friends, “likes” and profiling viewing habits we’d like to ask you to reconsider this recent update” or “you are about to tag yourself in a recent picture with your old boy friend from 3 years ago. based on some of your new friends, profiling viewing habits and you taking a similar action in the past, we’d like to ask you to reconsider this action.”

It can go on, but given that people tend not to change much overtime, the site could become more and more useful in helping people identify and avoid the unworkable situations they keep putting themselves into. Kind of like an older and wiser friend who tells you what they see but never tells you what to do about it.

I don’t think FB will take the leap and actually start churning out the information they are mining. There is little doubt that they know they are sitting on a reserve of sociological theory and practical knowledge that will push our understanding of human interaction forward dramatically. But since this information is base on millions of individual case studies, the individuals are not likely going to agree that they want to see it. It’s hard having someone read your mind and tell you want you are doing BEFORE you know why you are doing it or even before you know you are doing something.