What Is It And Why You Need To Harness It
Sociology 101 And The Fond Recall of Simple Social Network Maps
I will start with a confession: I always wanted to be a sociologist but never could see how to make an interesting, non-academic career out of it. Yes, that dates me. However, I did the next best thing, which was to focus on behavioral economics, moving ultimately into business intelligence. Extracting actionable information from data.
Now, PhD sociologists, of course, are in high demand. Particularly as the *big brands* wrestle with *big data* of which an increasing percentage is of the social signals kind. As I follow the unfolding of this social data era, I think fondly of my first social network maps. Back when I thought it was awesome just to map out how information flowed between individuals, groups or organizations.
The New Era: Big <Data> Relevance In Old Concepts
So just starting with the network map is useful, but there is a lot of other information that can be gleaned from social media data. When I think of social media, I consider at least every place that is now being referred to as ‘earned media.’ This is anywhere regular visitors can post content, including blogs, directories, Twitter, Facebook, Pinterest, LinkedIn, YouTube, SlideShare, Flickr, Instagram, etc. The list goes on. Changes will likely continue at this accelerated pace over the next few years; platforms will rise and will fall following patterns of user adoption of social media. I am particularly interested in where the early information will be posted, so I can see trends unfold and watch ‘mavens, connectors and persuaders’ influence outcomes.
Social media data, then, are simply the data produced by this explosion in user-based content. It is one thing to engage in social media, and quite another to step back to look at the aggregation of all the engagement to derive something that might be of value to you, to your business. There are a few things that are interesting here. One is just how much can really be known. Another is how credible and relevant the data really is to most businesses. And, finally, how actionable the derived information can be to those of us that are not the big brands.
Social Media May Be Deeper Than Your Next Tweet… And What In The Heck Is A Mechanical Turk?
I remember the day that I realized that this social media might be a lot of fun for an analyst. That is the day I first heard the term, “mechanical turk.” This term, in essence, is defined as the use of human beings to aid computers in areas that they stumble. Now THAT is a thought-provoking idea, is it not, in the age of sci-fi artificial intelligence that regularly takes over the world? It has some interesting undertones when it is we humans that now assist machines to accomplish their tasks more effectively. Before I digress too far, I will explain that the need for mechanical turking came from the early attempts to classify ‘sentiments’ in social media, whereas ‘bad’ is not always ‘bad’ and in fact, can be quite good. Understanding the meaning of ‘bad’ depends on the context, which is something computers cannot yet accomplish. So we humans assist them in classifying some ‘bad’ as truly bad and some ‘bad’ as good.
As an aside, check out this marketplace by Amazon: https://www.mturk.com. It is a bazaar for the buying and selling of ‘human intelligence tasks’ where we humans can aid machines. Interesting.
If you are like me, the whole concept of mechanical turking is quite stimulating to think about, and now you too might be off to the races to learn more about these social media data. And hopefully it is clear that there is more to this than meets the eye. So even if you do not personally *like* social media, even if you do not engage in it, it is much deeper than you might imagine. So, whatever you do, don’t dismiss it. It MAY very well be a critical part of the future of your business.
Why Should The Not-So-Big Brands Care About Social Media Analysis
The question is, what can be known? And then, what can you do with that information? Just as a way of introduction, I have already mentioned above that you can know the general sentiment about something, whether it is a company, a brand, a product, a sports team or an election. You can know how a sentiment trends over time, in terms of percentages of positive or negative commentary.
Other things you might know include:
- Volume, and changes in volume in sentiment over time. How many people are talking about this? Are greater or fewer people talking about this?
- Velocity of commentary. How much speed is in the commentary? Is there an increase or decrease in urgency, as reflected in people talking faster?
- Classes of commentary, whether highly emotional or more rational in nature, for instance. Is the nature of the conversation appear to be more toward rational discourse or emotional frenzy?
- Changes in classes of commentary over time. How is the classification of commentary changing over time?
- Trends in offshoot commentary, in volume, velocity or sentiment. So for example, the idea of ‘yoga’ might be growing in the US, and within that, ‘hot yoga’ might be really exploding.
- Physical location of sentiment. Where in the world or where in the US is the greatest positive sentiment for an idea, for instance?
As you can imagine, this list is limited only by your imagination. If you are truly inspired to learn something from ‘the crowd’ through social media data analysis, you can probably find a way to get it. This matters if you are not a big brand because it means that you too can access rich information that can drive many elements of your business strategy. You may no longer be inhibited by the staggering price tag of traditional market research to make important and well-informed business decisions.
The ‘Invisible Hand’ Of Self Interest For Social Commentary
Once you consider this, the first thing that may come to your mind is, “How legitimate is this data?” Or, “Can I trust it?” First the caveat that there is no perfect predictor. But I will quickly follow by saying that use of this type of data is supported by the Law of Large Numbers, a probability theorem that suggests that when you have enough votes, the expected value will closely approximate the actual value. So, assuming cleanliness of the social media data, the greater the volume, the more accurate the results.
The second thing that comes to my mind is John Maynard Keynes, the father of macroeconomics. Anecdotal, but illustrative, a fact I learned about him: he became very wealthy speculating on the stock market. What is interesting is that for all his economic knowledge, he bet on sentiment. He’d watch the nature of the commentary, the volume and the velocity, and then bet on it. And it was extremely successful as an approach.
Of course, there are much more interesting and current examples of social media data predictiveness. We’ve seen accurate predictions of social unrest, outcome of elections, and even movement in the stock market. Check out ThomsonReuters ‘Mood Index’ at http://www.marketpsychdata.com/ for an example of current application, an example of which (the Bubbleometer) is shown below.
In the end, I think social data is more reliable than not because it is grounded in well-proven theory. It also reflects the uncontrived motivation of ‘individual self-interest,’ which we’ve all learned can be the basis of a very stable behavior over time.
Business Application: From New Products To Facilitating The Purchase Process
Now the question is: what can you do with the derived information you will gain? DirectiveGroup approaches our work from a strategic perspective, whether we are developing a website or creating some form of digital marketing. You don’t find us following the ‘ready, shoot, aim’ method of many agencies; instead, our tactics follow our strategy work, and social media analysis informs us greatly in both areas.
You may want to consider these uses as well. Some examples of how you might apply social media data analysis to your day-to-day business decisions might be:
- As you develop your next website architecture, you might want to consider trending topics. Developing pages around trending keywords can be places you can rank faster (from an SEO perspective) and where you might establish thought leadership with your target market.
- How about considering the development of new services or products based on social media signals? How about determining your next location based on where discussion is happening?
- Online and offline marketing could (and actually SHOULD) be developed to speak to the trends, to argue against the trends, or to otherwise stake some claim associated to the signals.
Hopefully these examples will stimulate some ideas about how you might apply this sort of information to your specific situation. In the next part of this series, we will tackle one specific use of social media data, which is to develop relevant content for your target market segment.