Support Center > Knowledge Base> Article: Sentiment Analysis

Sentiment Analysis

Article ID: 61405 Email Print
Question
What is sentiment analysis?

Answer

Sentiment analysis uses a selection of words and phrases that will add either a positive or negative value on the chat. If we detect any negative words or phrases said by your customer during the chat it will automatically flag it as a potential negative chat. This sentiment will be determined upon the chat ending, in which it will analyze only the customers messages, provided they sent at least 50 characters total for all messages. At this time only English language messages will be analyzed. Our current logic and list of words and phrases cannot be disclosed at this time, however in the future we will allow enterprise level clients to edit the logic themselves by including or excluding custom words or phrases. The final analysis will only be shared with the account admins or any email added within the account settings, the customer will have no knowledge of this analysis.

 

A few of the biggest advantages we have seen with the Sentiment analysis so far are: 

  • It gives you the opportunity to prevent future negative chats. What can be said to make the customer feel like we have helped them to the best of our ability?  

  • Determine best course of action for agent improvement

  • Instant Damage Control

  • Review what types of chats are getting the most negative responses from customers.  Determine how these types of chats can be avoided.

  • Not having to search through all chat transcripts for chat outcomes. Searching through Chat Transcripts for every possible negative chat is monotonous, the Sentiment Analysis picks them out for you.

 


The Sentiment Analysis provides the upper hand.  Perfect your service by continuing best practices and escape the possible negative chats by proactively monitoring the Sentiments. Get a grasp on what is and isn’t providing positive chat outcomes. 


Related articles

Article Details
Views: 16364 Created on: Feb 05, 2015
Date updated: Feb 05, 2015

Poor Outstanding