Sentiment Analytics – The Gold Mine, which you didn't Mine!

It’s really a no brainer that Customer Satisfaction matters. Every IT or Business unit  I’ve known, considers Customer Satisfatction, a very high priority and does strive really hard to ensure they engage very closely with their customers. To ensure every customer query or concern is serviced well within time. If one does go by the assumption that servicing customers in a timely manner will keep them really happy and satisfied, one would tend to focus more on Service Level Agreements.


About Saama Executives

Saama Executives is an exclusive group of thinkers, leaders, mentors and innovators within the company. The members of this group come together from time-to-time to pen their thoughts on topics that would matter the most for the industry. Over the last few years, the group has written some brilliant pieces of Insurance, Life-Sciences, Healthcare and CPG industry.

Related Posts

Nekzad Shroff says:

Just like any other paradigm shift, it takes a leap of faith to start believing in and then relying on the techniques that try to extract meaningful information from inherently unstructured data. To shift from our view of analytics as a 100% accurate distilling of a hard data set to the view that valuable and usable analytics could come from “fuzzy” extraction of sentiment from reams of chatter.

But the shift is happening – inexorably. As more and more business is interlinked with the “social” web, there’s no more ignoring the treasure trove of information that’s locked away in these massive amounts of data.

Welcome to social analytics. “Anything you say may be mined”

Meta Brown says:


Fascinating description – I take it this is something you are doing for your clients today.

Sentiment analysis is also an important part of my work, and I am always interested in ways that people can make a clear path to measurable business value from sentiment analysis. I have some questions for you:

The examples you give have very clear, strongly expressed sentiment, but often sentiment is not so clear in text. What’s your experience with accuracy in sentiment analysis and how does that affect applications like this?

How does the type of language – whether it is specialized industry language, slang, or a language other than English – come into play in applications like this?

Seth Grimes says:

Ghananeel, I find it interesting and ironic that you list USEFUL comment categories such as “I will wait”, “Not using currently” and ”Next Release?” — “No Business Sense”, “Angry”, “Frustrated”, “Time and Again”, “Too complicated to use” — but seem to think they need to be reduced to simplistic positive/negative/?? bins. Or do you do — or are you interested in — sentiment resolved to business-aligned categories such as the ones you list? Many leading tools can handle the latter, and not just sentiment valence.


Leave a Reply

Your email address will not be published. Required fields are marked *