Removing the barriers to effective customer experience analysis

With a direct link clearly established between customer satisfaction and revenue growth, you would think that customer experience scores would be on the rise. Unfortunately, that’s not the case. According to Forrester’s US 2017 Customer Experience Index (CX Index™), CX quality worsened between 2016 and 2017. And this is in the face of significant investment in CX technology.

Despite the obvious promise of text analytics, many executives remain frustrated and confused.  It turns that more data doesn’t necessarily equal more value. Nor are profit returns guaranteed. In fact, the profit story is rather dismal. But it doesn’t have to be that way. There are organizations generating positive returns using text analytics. What sets them apart from the rest?

At Decooda, we believe the customer experience approach needs to be emotionally engaging to be successful. That means your CX platform needs to be able to detect customer emotions with superior precision and recall. To accurately detect how customers think, feel and act about a product or brand, we employ an approach called Impact Analysis. Using an artificial intelligence algorithm, we look at topics and then assign weights associated with emotions, or the persona changes, we see for a particular customer.

Just because a customer is talking frequently about a topic doesn’t necessarily mean it’s important. By linking topics with emotions and cognitive states, we can tell you what specific aspects of the product or experience most significantly influenced customer behavior. Focusing on low impact conversations is a common mistake organizations make, which complicates the analysis of all the investment they are making to enhance the customer experience.

Overcoming analysis imperfections

One problem with traditional sentiment analysis techniques is that they create significant blind spots in the results. At its simplest, sentiment analysis is about finding single words that have sentiment associated with them. Words like “love”, “hate”, “great”, or “terrible” immediately convey a particular sentiment or polarity.  But is sentiment analysis really just a matter of finding prior polarity terms? Of course not. It’s much deeper than that.

Consider this example:

“I did not like that shopping mall last night.” The word “like, as a verb, conveys positive sentiment. But everyone would agree that this sentence is negative in regards to the shopping experience.

Or what about this sentence:

“That dining experience could not have been any better.”

Even though one comment is clearly positive and the other negative, a classic sentiment analysis approach would likely score these sentences identically. As these two examples show, prior polarity terms by themselves aren’t enough. We need more analysis to accurately determine sentiment.

Driving positive returns

Another thing we’ve done at Decooda is create a more simplified approach to open-ended questions. Instead of asking traditional survey questions, which often produce boredom and disinterest, we capture information from one or two “imaginative” questions.

Though the difference is subtle, research shows this more simplified approach yields significantly more sentiment, emotional and cognitive state detection and in far greater attribution. We then use our cognitive insight engine to analyze the data and interpret the results. All of this gives you a better understanding of your customer, enabling more actionable insights and smarter decisions.

Comments are closed.

Enjoy this blog? Please spread the word :)