Text analytics and Market Research:

Text analytics has been around for a while now and its use in market research widely reported. According to a study conducted in May 2011 and reported in Business Week, the business potential has reached $1 Billion and it is easy to see why. Text analytics is extremely useful in measuring sentiment towards a product and many companies are now using it in product and service development. It is also an extremely useful tool within CRM. We all write so much more than ever before and social media inevitably has had a huge impact on the growth of this technology. On the rise of social media and growth of text analytics, Tom Anderson of Anderson Analytics says that “The rise of social media during this time certainly helped propel this field and offer even greater opportunities. Expanding beyond survey research to working with data from bulletin boards, Linked In, Facebook and Twitter has certainly kept it exciting.”  The main use for marketing people is to derive value from all the text based comments out there.  As a tool it works by being able to identify major themes without the need for reading responses word for word, you can also distinguish between positive and negative comments. One of the challenges in analysing social media text however is the use of unstructured language on the various sites. Grammar and spelling has generally gone out of the window so the software has to be taught this new language (lol). Also looking at the nuances of language, a human would understand a sentence such as ‘I know the gym is working because my muscles kill me’, although kill would be considered a negative word in this instance it is obviously positive therefore the analytics tools have to be taught that the word because in the sentence changes the meaning. It is very sophisticated. Another challenge is in the amount of stuff written out there on the social media sites about a particular product, some which is really usable and some which is just white noise. One example I read on the Next Gen website talked about Coke “I am kicking back on the sofa with a Coke and a DVD”  versus a really usable sound bite “I really want to buy a Coke because I need a caffeine fix right now”.  According to Anderson Analytics the main customers they focus on at the moment are the ones that spend the most on advertising as they have the most interest in the subject. However PR professionals and insight managers are also using the technology. The growth in interest in this area has really been since 2005 but mainly in 2010 when people started realising how important measurement of social media really is.

I have had experience of using the tag cloud software available on the internet via Wordle which is a version of text analytics software where you type in various words to produce a visualisation of the words (I used all of our testimonials). The words that are repeated most frequently show up as larger text and different words are colour coded. You can see the results on our testimonials page.

Do you use text analytics?  Let us know on Twitter @carrotpharma