Sentiment Analysis

Understanding when your company or brand is being talked about is important. Understanding whether or not your company or brand is being talked about positively or negatively may be more important.

In the last few years the research behind sentiment analysis has ranged from thumbs upthumbs down and emoticon faces to more advanced methods.

A few companies mentioned in the article above are trying to sort out the signal from the noise like Tweetfeel, twendz, and twitrratr and others like Trendrr (can I buy a vowel please alex). Each has their own way of parsing data, mainly Twitter data, to mine the conversations for sentiment.

In the larger universe of analytical tools you have companies like BuzzLogic, TechRigy, Radian6, Visible Technologies, Viral Heat, and a bunch of new startups all trying to solve this marketers puzzle.

At the low end these services cost $10.00-$20.00 per month and on the high end thousands. Depending on what you are tracking, how far back you want to go, and how hands on of a solution you need, someone is building a sentiment analysis tracker for your.

The goal here is to track beyond mentions. For example, if your brand is mentioned 2,654 times – how do you know what the sentiment of mostsomeall of those messages are? Where are conversations happening about your brand in the first place? How can you respond back to solve problems, answer questions, and apologize where necessary?

In thinking about the sentiment analysis market, I am trying to frame thinking 5 years out when brands will themselves have more conduits to their customers. Will systems still exist that brands do not know about and are not tracking? Or will the convergence of connectivity mean a more aware corporate culture that does not need to use yet another tool in their arsenal. Perhaps 5 years is too short of a timeline, but certainly they will have a grasp on this in 10-20 years.

I am curious to know what tools I should be looking at and how people are solving brand tracking probelms today. Let me know if you have ideas in the comments below.

Reblog this post [with Zemanta]