I just call to say I love you. How to track and measure emotion?


Share on LinkedIn

SMALL Quote01_01_Megan-Burns,-Forrester

Emotion trumps the product and the price. But how to capture and measure customer experience?

In June, Forrester’s researcher Megan Burns caused a bit of a stir during New York Forrester’s Forum For Customer Experience Professionals proclaiming, that among the three customer experience dimensions – ease, effectiveness, and emotion, the last one plays the biggest role.

Emotional engagement has been touted as the strongest loyalty driver for quite some time now. This has led to the implementation of the Voice of Customer programs, yet resulted in brands often struggling with an unusual challenge – capturing, analysing and measuring emotion.

I just call to say I love you

The metric often adopted to reflect a customer’s perception of a brand is Net Promoter Score. NPS provides a measurement of the overall satisfaction with the company. This, however, is based on surveys that allow for little specification (read 6 things that Net Promoter Score won’t tell you but conversation analytics will) and doesn’t provide information about the events that triggered the customer’s perception of the brand.

“I wonder who it was that defined man as a rational animal. It was the most premature definition ever given. Man is many things, but he is not rational.”

Oscar Wilde, The Picture of Dorian Gray

While it is impossible to ask customers how do they feel at every stage of their journey, there is a largely untapped source that can provide a hefty chunk of that information. Every day, enterprise servers store thousands of minutes of phone calls, during which customers are voicing their opinions, wishes and complaints about the brand, product or service, and sharing their feelings in their purest form. Each spoken word or phrase is emotionally charged.

Capturing, identifying and structuring those emotions along with their triggers; events, interactions and experiences is the task of Conversation Analytics. Voice is a unique source of knowledge about emotions. It is not subjective; it doesn’t rely on statements. It contains honest, unbiased feedback. How convenient, considering that the phone call is still the most frequently used medium for connecting with brands.

How to measure emotion?

Despite the variety of channels available, customers still prefer talking to a live agent. According to Call Centre Helper report, “What Contact Centres are Doing Right Now” social media inquiries constitute only 3% of total amount of interaction, trumped by email (26%) and a surprisingly high 68% for phone calls. Sounds like numbers from 2001? Not quite. This research study was conducted only a year ago. The findings presented in Parature’s 2015 U.K. State of Multichannel Customer Service are even more astounding – 81% of clients contact brands on the phone on regular basis.

Conversation Analytic platforms use a bundle of algorithms called EVS (Emotional Voice Streams) to process thousands of calls simultaneously and detect words, common patterns and triggers associated with negative or positive emotions. EVS decodes the vocabulary and searches for relationships between words and phrases, simultaneously analysing non-verbal audio cues that contain unfalsifiable information about the speaker’s emotional state.

Contrary to popular speech to text or Natural Language Processing, which most speech analytic solutions rely on, Conversation Analytics doesn’t ignore the audio layer but utilises it to comprehend fully what happened during the conversation and deliver clear information about what events triggered a particular reaction.

While your social media is constantly scanned for feedback, phone conversations are a grey area. This leaves most enterprises with thousands of hours of recordings full of unreachable insights – a useless data clutter. Having a system that can capture, understand, map and structure phone conversations, means taking the visibility of customer feedback from 30% to almost 100%, without ever conducting a survey.

Attitudes in their natural habitat

Customer Experience creates a constant thirst for analytical input. Unfortunately, consumers are famous for their blatant disregard for all kinds of questionnaires. What an irony, considering that they like nothing better than giving companies a piece of their mind. The only crux is that customers voice their opinions when and where THEY see fit. Surveys ask clients for a review usually minutes after they have just called and shared their thoughts on the product, service and company. Filling out an additional survey not only creates a sense of investing extra effort (especially in the case of interactions that did not go perfectly well), it also sends a message “You called and spoke, but we didn’t listen”.

Somewhere between the “Hello, how can I help you?” and “Thank you! Goodbye” there are thousands of words and phrases, which contain information about speaker’s feelings. When organised, they constitute the ultimate source of knowledge about the customers; personalised and specific, encompassing and universal.

At the moment, the industry standard is to fish in the desert, rather than cast the nets where the fish live. In other words; get the feedback where it’s already waiting. Start taking advantage of your contact centre.

Andrew Lamrock
Andrew Lamrock is the Director of Enterprise Intelligence at Call Journey, and was also the National Manager of Sales Programs at Telstra and CEO of Medibank Melbourne Ice.


  1. In your post, you state “While it is impossible to ask customers how do they feel at every stage of their journey….” That’s pretty categorical, and it doesn’t recognize extensive research in this field. So, we certainly don’t think generating this insight is impossible; and, in fact, we have successfully been measuring the impact of experience emotion and memory on customer behavior for over a decade.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here