Sentimental Journey

“Powerade Zero tastes Sick!”

In only 26 characters, some Twittermeister just put it on the line. They emphatically announced what they think of your new sports drink product. They have a zillion followers. They are hyper-influential. They can make or break your reputation. So what are you going to do about it?

Might I suggest some serious segmentation?

If the tweeter has an IP address in Southern California and also tweets about things being “phat” and “all that and a bag a chips,” then direct message the individual immediately and offer a free case and a load of thanks!

On the other hand, if that tweeter hails from an IP address in Southern Florida and also tweets about the smooth taste of Ensure or how the Rite Aid One-Day-At-A-Time is better than the Ezy Dose Four-a-Day Weekly Maxi-Pharmadose, then you might want to reconsider promoting your goods to the over-the-hill set.

Language is tricky. Sometimes it’s hard to tell how others are feeling when you take away vocal inflection, facial expression, and body language. Great writers are great because they can communicate a massive amount of information in a tiny space. For example, here is what is often acknowledged as the greatest and shortest short story ever written:

“For sale: baby shoes, never worn”

In just 32 characters, Ernest Hemingway conveys pain, loss, grief, emptiness, and then some. Clear communication takes talent. Few people can live up to @ehemingway.

I have publicly doubted that textual analysis will ever give us as clear a signal about how people feel as the behavioral analysis of Web analytics. I’m starting to doubt myself on that. Publicly.

And so, I’m about to take a journey into sentiment analysis.

I’m quite comfortable with Web analytics and how good it tastes with customer satisfaction results. Now it’s time for me to dig into all the sentiment analysis systems – starting with classic textual analysis that companies like SAS provide, and moving all the way up to the more esoteric and fascinating NextStage Sentiment Analysis.

In between are an overwhelming variety of social media-specific tools that will require the best of my skills as an interpreter of promotional literature to discern the useful from pure marketecture.

I know that man does not live by bread alone and one data set by itself will deliver nowhere near the value that combined data sets can. But before we can even experiment with alternative recipes, it really helps to understand the attributes of the various ingredients.

It will be a challenging journey, but I’m keen that it not be protracted. Therefore, I’m going to ask for your help.

Are you using textual analysis tools and systems? Sentiment analysis systems? Ever played with the Gibbs Universal Industries Sentimeter? If so, and you feel you’re getting valuable business insights and can make serious business decisions based on the output, then I want to hear from you.

Leave a comment below, tweet me @jimsterne, or e-mail me and let me know whether you think sentiment analysis is Cugat or Homer…

Related reading

Flat business devices communication with cloud services isolated on the light blue background.