Your customers want to communicate with you. Your partners, employees, potential clients, competitors, press and social media influencers want to communicate with you too.

Humans communicate in a wide variety of ways – body language, eye movement, intonation of speech, hand gestures, use of color and clothing. There is no system on earth that can yet capture all of the wide bandwidth of communication that we all use, as we go through our days.

Text Communication

Let's simplify the problem down to text communication. Text is an excellent intermediate point between "doing nothing" and capturing all the rich variety of human interaction with some magical system that only the NSA could afford.

Text is the way humans try to have "structured" communications (even though its highly unstructured data, compared to what computers do.) There's grammar and syntax, you know where a sentence starts and stops, and you know how you're supposed to break it up.

And oh boy, is there a lot of it. In one minute, there will be:

There will also be uncountable millions of:

  • Internal support documents
  • Customer calls
  • Web pages
  • Customer reviews

All of this information, while being the most recent available, is months out of date – and with the possible exception of Wikipedia, all of those numbers are almost certainly higher now than ever before.

Text Mining

Text mining. Text analytics. Text analysis. Unstructured data mining. These are all terms for (basically) the same thing – turning a mass of unstructured conversations into structured information that you can then analyze.

But you don't want to become some mad big data scientist, understanding crazy statistical "maximum entropy" models and how they interact with pronominal co-referencing.

Right. Me neither. The whole point of this note is to show you that listening is important, that there's probably something that concerns you in the text, and then, to give you hope that this isn't some crazy task that only a PhD can survive.

Text mining is a horizontal technology like search. There are tons of SaaS and on-premise players that have integrated state of the art text mining technology into their products for customer listening, media monitoring, survey analysis, business intelligence, customer support, and any other application you can think of.

You can become a text-mining expert, and you will have an edge because of it, but you don't have to. All you need to do is decide what conversations are important for you to listen to (like, say, Tweets or Yelp! Reviews) and then find someone that has integrated that data into a good dashboard for you to review and base your decisions on.

In summary, you need to listen to what the world is trying to tell you, and the premier technology for doing so is "text mining." But, you can lean on others to help you use this daunting technology to extract the right conversations and meanings for you.

  • Seth Redmore has over 10 years of experience in text mining industry. He is currently VP Marketing & Product Development for Lexalytics Inc.; previously working for Brandmail Solutions and Cisco Systems.