Big data is changing social media marketing. There are no two ways about it. The reason is simple: Big data is giving us a look under the hood of social media marketing. And that, in turn, is giving us an ever deeper and better understanding of what’s going on, what works and what is – in fact – counter-productive.
So how does it do that?
Social media = big data
The first thing that needs to be understood is that social media is a huge source of big data. A great deal of our understanding about how people behave and why they choose the things they do is based on how they behave on social media and what they do there.
There are examples of people running experiments on social media in order to see if they can influence people and change their behavior. The results in a nut shell? They can.
What’s more, it’s becoming obvious that social media can now be used to predict people’s behavior. For example, it turns out that measuring engagement in different states with the candidates of the US election was better at predicting who was going to win the election than the polls were.
And that’s just two examples of how powerful social media has become in terms of influencing and predicting behavior.
Even more exciting (or frightening, depending on where you’re coming from) is what social media means for educating algorithms. When you’re trying to build an algorithm to work out why people do something or what they’ll do, you need a lot of data.
And that’s where social media is coming in. For example, they’ve designed an algorithm that can work out, based on your social media posts, what kind of personality you have.
From there it isn’t a big step at all to find out buying habits.
Ever deeper layers
The really interesting thing about algorithms is that they can make connections that we would never have come up with ourselves. There are often countless ways that in which variables can hang together that we ourselves would not have come up with. This is because we don’t have the kind of computational power that computers can throw at variables.
In fact, our brains are rather limited in terms of computational power. For that reason, for us to figure out a connection there has to be a reason for it to exist. Algorithms don’t have that restriction. They don’t care about reasons. All they care about is mathematics and logical relationships.
If the color of the t-shirt you wear in your profile picture correctly predicts whether you like sport cars or Humvees, well then that algorithm will factor that in and run with it.
Unlimited attention span
Similarly, while we might be able to accept that A & B together influence C, our limited attention span won’t let us go much beyond that. We simply don’t have the working memory to put in many more variables.
For that reason, we rarely work with variables that only have small effects – like how handedness affects life expectancy – and tend to ignore them.
Algorithms don’t have that problem. They are willing to incorporate as many variables as it takes in order to find out an effect. Something has less than a tenth of a percent effect on buying habits? That doesn’t matter, because along with these 2000 other variables the algorithm has collected, we’ve still got good predictive value.
The changes being wrought
The question now becomes, are we learning the same things on the different social networks (all people are, after all, deep down pretty similar)? Or is it instead that there still an underlying cultural differences between the different networks that will create different results based on the users that the algorithms study?
And what will the future hold?
That’s hard to figure out. Many companies that are actually collecting the data – such as Facebook and Google – haven’t yet revealed what they can do with it.
Even if we can’t do that much yet, it’s the potential that matters. Sure, at this point the data is still largely unstructured. That won’t always be that way, however. You can be sure that numerous companies are crunching data by the terabytes in order to get a better understanding of human behavior and an edge on their competition.
What’s more, more and more people are jumping in. How can they not? To not take advantage of big data while you’re competition is crunching all the information that they can get their hands on is the same as playing Russian roulette with yourself.
You’re going to lose. You might not know when, but it’s going to happen as your competition gleans some kind of strategy or insight that you were completely unaware of.
What does this mean for social media marketing?
It means that we’re moving from a cluster bomb strategy to a laser like precision with our marketing strategies. We are already seeing strategies shift and change as a result of big data. That is only going to accelerate in the years to come.
That, along with the changes that are taking place to make it easier and easier to analyze big data, mean that everybody and your 90 year old grandmother are going to pile into big data and use what they’re learning to inform their strategy.
The only viable strategy is to beat them to it. Know which way the wind is blowing and get the jump on them while you still can. If you still can, really. Perhaps your competition has already beaten you to it. Perhaps even now you’re running behind.
About the author: Rick Riddle is a marketing consultant, head content manager at Smart Paper Help and an up-and-coming blogger whose articles aim to help people with digital marketing, blogging, entrepreneurship, and career. Feel free to follow Rick on twitter and LinkedIn.