Constraints of Social Media Analytics

Originally, people used to spend more time online in bulletin board systems where, I’m sure you’ve seen websites like this, where there’ll be a series of threads, and a thread might have a topic like, it could be anything for the things I study. It might be, I just got diagnosed with cancer, help, does anyone know what I should know along my journey? And if you click on that thread, within it, there’ll be lots and lots of posts. So that type of system, that’s what was more common in the past. And people still utilize these types of systems very frequently. But now it’s more common, are used very large social media platforms.

So things like Facebook and Twitter, where it’s not as much of a, there are these one-to-one interactions, but there’s also these very broad, reaching networks. So you might be receiving information not just from your best friends, but information that has been spread through dozens of links along the chain. So it’s similar that people are still interacting with one another. But the ways in which people are interconnected indirectly in very far reaching ways is very different now.

I think that In the very near future, and also into the far future, that more and more of social media analytics is going to be automated. So we need really smart analysts to set up some of these processes. But a lot of platforms, let’s say like Twitter or Facebook, when a company wants to know what customers are thinking. They wanna know that all the time. So a lot of companies are setting up processes where they’re continuously collecting this data and analyzing it, and they need smart people to set that up and then interpret and make sense of it. So I think a lot of automation is part of the future.

I think it’s also that every organization or company that you can think of that wants to become competitive in the future market places, they all need to utilize social media data. Whether it’s for their own customers, looking at competitors, or even just trying to understand how the public reacts or thinks about issues that they need to consider in their decision making, it’s all very vital.

People who are active online in social media platforms can be very different from the general population. In general, you may not be surprised that people who utilize social media are often younger than the rest of the population, and certain platforms attract certain demographics. So you could say it’s common that on Twitter more men use Twitter than women. And Facebook has much broader representation of people from different genders, and ages, and groups. But then you’ll also find that specific online communities that cater to a particular interest, whether it’s people that are interested in building or restoring old cars at home or websites where people go to get tips on knitting. They attract very different groups of people depending on interest.

So it’s important to be aware of the demographics of people who are using particular platforms, and also based on the particular topic or interest that’s on that website.

Even though the information that people post online is public and I think, in general, most people realize this is public. When you’re on your own Facebook page or when you’re on some sort of bulletin board system in your posting, it can still feel a little bit private at times or can feel at least like it’s a group of people that you might recognize and know. And people can sometimes forget how very public this information really is. And as social media analysts, we need to be aware of the fact that even if we’re collecting information that’s easily accessible and very public, the people who created the content, they may not feel that way.

It’s also important to be aware of issues that can pop up when you aggregate data. When you collect data from multiple different sites online and bring it together, because there are techniques that let you match up that information with specific individuals. Now that can be very useful and it can be not at all harmful to the people who are in your data set. But you have to be careful that you’re treating the data carefully and recognizing that it really is human beings that you’re studying.