AI, e-Commerce and Understanding Emotional Response

Emotional response is an often-overlooked aspect of automated communications. Let’s take a look at a common automated customer service response that might be generated to respond to a customer’s comment:

Complaint: I haven’t received my order and I ordered it TEN days ago. It was important that I receive it by last Friday and yet your slow site still says it hasn’t shipped! When will my order arrive? I should probably just cancel it.

Automated Response: Thank you for your email. Our staff typically responds to all email requests within 48 hours.

We all can see what is wrong with that. First, the customer is clearly angry – justifiably so – but the initial response says, “Thank you.” Is that what the customer wants to hear? We can read this and automatically recognize that this individual in a state of anger and frustration – two emotions no retailer wants their customers to feel.

When I look at AI and automated responses, I often see the lack of emotional connection it brings in its communication. It doesn’t gauge anger, happiness, or disgust; everything is treated in neutral context. But what your customers really want is to be treated in a manner that responds appropriately to their emotional state. Let’s rework the response in the basic example above:

Automated Response: I am really sorry that you are having issues. I have notified the appropriate people to look into this right away. We will get back to you shortly.

This response is simple address the emotion state of the customer. You will notice in the first example the customer was complaining about speed, yet the automated response amplified the problem by telling the customer it will be another two days before she’ll receive a response.

The Emotional Response Engine in our products tries to do just that: detect the emotional state of the person it’s communicating with, and respond appropriately based on that state. We classify six emotional states and let our customers add special parameters to each. So let’s take a look at another hypothetical example:


Customer: I want to return this damn item!

Response: I’m sorry it wasn’t what you expected. Please use this free return shipping label to return it to us at no cost.


Customer: I’d like to return this item.

Response: OK, that’s no problem. Please ship it to 123 Test Street, New York, NY.

In the “Anger” example, the emotional state was identified and in the interest of making the customer comfortable certain words were used as well as an offer. “Sorry”, “free”, “no cost” all convey a sense that you want to please the customer and make things right.

Properly understanding your customers can go a long way to improving customer interaction and turning every interaction into a successful transaction.

Article Name
AI, e-commerce and understanding emotional response
Emotional response is often over looked in automated responses. Let us take a look at a common automated customer service response one would get when replying to customers comment.
Publisher Name

Author: Alan Washington

CTO and Chime AI Evangelist