Listen To Your Customers With Text Analytics

Listen closely to what your customers are saying

Customer Experience Management, Voice of Customer, Survey Analysis, Customer Satisfaction are all internet buzzwords which loop back to one central concept – listening, acting, and then showing that you're acting on what you've heard.

How can you hear customers (and then act) when you've got tens of thousands of of them? You can certainly survey them. You can ask them structured questions. You can call them up and chat with them.

Surprise me

There are a ton of different ways that you can reach out and gather information from your customers. In many ways, that's the easiest part. What's harder is collecting it in such a way that you can be surprised. That your customer's innermost desires and wishes can come bubbling up to the top – showing you new products, or fixes for old ones – helping to guide you and your company into places that you didn't really expect.

Conversations. That's what you want. But how can you have a conversation with thousands of customers without thousands of employees talking to them at once?

Text analytics

Text analytics will read your customers musings, tease out their opinions, group their words into some kind of meanings. What sort of meanings?

You need to know what are they talking about – nouns. Nouns are the figurative bedrock of customer listening. People, companies, and places.

Then there are contextual clues. Are they talking about food, or cars? Service or bathrooms? Weather or bandwidth?

Then you need to work out what they are feeling. Are they really psyched or kinda down; mildly positive, or sternly negative.


Put all of these together – what can you understand? You can understand who and what they're talking about, and if it's positive or negative. You can see if they're trashing your competitor, or telling you about a problem with your packaging. You can see if they wanted more curly fries, or simply less salt. Was Jessica doing a good job helping them, or does she need a bit more training?

Once you know, you can react. Modern text analytics includes the ability to extract "calls to action" or determine "intent" - you can see what you're being asked to do, and rapidly do something about it.

And maybe "doing something about it" is just responding to an annoyed customer – but maybe it's something more substantial. You have to make the decision about what's worth doing, but with text analytics, you can see what you need to do from anyone who is trying to get you to listen.

  • Seth Redmore is VP Product Management & Marketing for Lexalytics. He has over 10 years of experience in the text analytics and natural language processing industry, previously working for Cisco Systems and Brandmail Solutions.