But the question for those interested in social media is this...is listening good enough?
In fact, if you are only trying to listen are you even scratching the surface of what is possible in social media? I would argue you are not. And that brings us to the point of this post...
Listening isn't enough...you need to understand what it is saying. And understanding requires that you can pull apart the relationships within the sentences. If you can do that on mass scale then you suddenly can begin to understand what social media is really saying. Why? Because if you can efficiently understand the language of all that content you should theoretically have a much more accurate look at the data which then will enable the most important lynch pin in social media...METRICS.
In my experience on the social media dance floor everyone is trying desperately to tell you they can accurately understand what is being said. But the question is how do they BUILD their data set. Do they use "people who know where to look"? If they do then they are at the mercy of what I call the hit by a bus principle. If anyone who "knows where to look" is hit by a bus while crossing the street then than integrity and accuracy of that data just changed and it is less valid than it was before. Another thing I always see if people stating, "on this website there were 3.2 million tweets and here are sample of what they were saying". How is this helping you understand? Volumes of tweets are interesting because it give some sense of scale between two subjects (the buzz), but it does nothing to help you understand the sentiment (emotion) or better yet the concept of intensity (passion). It merely allows a small comparison between volumes on websites. And in the world of metrics that is only worth a bit but not everything.
In fact, how could you translate the number of tweets into ROI on say an advertising spend? This brings me to my "pitch" (a bit of promotion for my company NetBase). At NetBase, because we have 80-90% accuracy automatically captured in our Index (we have googled the web so to speak), you can now UNDERSTAND the emotion within the content using a repeatable method to collect the data. Because this is the case, each search that is done is apples to apples and because we pull this data from content that is indexed, the principle of comparison is at my disposal.
This brings me to the metrics. It is easy to think about how to apply social media first and THEN build metrics that suit the application. I choose to go a different direction in my conceptualization of applying social media to business. I would rather think about the measures I have at my disposal and THEN find the right applications.
What I have spent much of my time contemplating are the usage of buzz (mentions only), sentiment (emotional mentions) and passion (intensity measure like capturing love, like dislike and hate) to help create metrics that will enable viable applications of social media data. In fact, what I found quite interesting is that in our index for every 100 buzz sound bites, 20 of them will contain sentiment and 2 will contain pure passion (love, like, dislike and hate...while if you include other emotional expressions of passion the number goes up to about 5%). A good analogy would be the following...when you go to the carnival and grab the hammer to ring the bell...the deeper you connect with your consumers with either your products or online messages the more "response" you will get. If you "hit it" 1/3 the way up to the bell, you get buzz. 2/3 the way up you get sentiment. All the way up you get passion.
Another way to say it is this...
If my social media message gets a higher than average number of positive passionate expression on a daily basis without spending a dollar on advertising then I have gone viral and hit the holy grail.
That is really what we are all trying to do on social media. This post for instance (is mostly for me to get my thoughts out), but there is a part of me that wants millions of people to read this and say they love it without every having to tweet it to get people's attention. Why do I care about that? Because I like to think about concepts that create change and if my message has the maximum impact without any effort that I can drive a lot more change efficiently.
Now onto measuring your spend. If we think about those three counts as a means of accurately capturing what is being said online about a brand, I can begin to think about their application to something like advertising. In advertising we know what the strategy for the ad was. We also know what we spent on the ad. But how do we understand the response to it in near real time. I would propose that social media can be that response using buzz, sentiment and passion Why? Because after I start my campaign I can very easily begin to track buzz, sentiment and passion (if they actually ring the bell). Once I know those counts for a promotion, I can then begin to look more closely at the relationship between these measure.
- I could look at the the counts in relation to an attribute beneath the brand like customer service versus store cleanliness.
- I could take a look at the amount of sentiment generated versus some average across all brands (this will help me see if the attribute like customer service elicits a greater sense of emotion versus another attribute)_.
- I could even look at the rate of change of each measure over time and see how the change relates to other events
The point is this...knowing what the buzz is only will never be enough to accurate extract real insight from social media. It is going to be accuracy in the data that only natural language processing can give you. It is going to be the ability to capture more measures than simple buzz. And lastly, it is going to be ALL ABOUT APPLICATION AND METRICS.
I always preach in social media the name of the game is application...and that is my game. Make it yours.