I Know How You Feel! Sentiment Analysis in the Contact Center

This is a guest post from Paul Stockford

The advent of speech analytics in the contact center more than a decade ago opened a whole new world of operational insights in terms of really understanding what customers were saying. Well-trained analysts and data scientists were able to turn key words and phrases into usable customer and business intelligence by testing hypotheses and following the trails that were the result of this contact center detective work.

Speech analytics also opened the door to data mining for hidden customer service treasures in the mountains of recorded voice customer conversations typically held in storage for quality assurance and compliance purposes. Speech analytics was, for all intents and purposes, the contact center’s first foray into the world of Big Data. Understanding of customer content increased exponentially in the decade following the general availability of speech analytics.

With access to all this new data, it seemed that the customer experience (CX) code was about to be cracked, but there was still a piece of the puzzle missing. Solutions such as speech analytics work to tell us what customers are talking about, but they don’t tell us how the customers feel. Speech analytics provides business intelligence that can lead analysts and data scientists toward greater customer insights. Sentiment analysis guides users toward what to do with that information.

Sentiment analysis has its roots in the analysis of online reviews and posts going back to the early 2000s. As online reviews gained in popularity, companies whose businesses or products were being reviewed were interested in more than just the words left in the review. They wanted to know more about the emotions behind the posted text.

With advancements in natural language processing (NLP), combined with innovations in speech analytics, the next logical step for sentiment analysis was the spoken word, as in customer contacts and conversations. Research shows that contact centers are embracing speech analytics as a tool for mining for data in recorded conversations. According to the 2019 Saddletree Research survey of contact center professionals, 34 percent of the North American contact center industry is currently using speech analytics. More importantly, an additional 33 percent of respondents named speech analytics as a solution they will evaluate for purchase in 2019 and eight percent have already funded speech analytics for purchase this year.

Sentiment analysis provides even greater insights into customer calls that can be used by supervisors, managers and others throughout the enterprise.  While speech analysis requires the skills of an analyst or data scientist, sentiment analysis is an automated process that doesn’t require human intervention to the extent that speech analytics, or even quality management, does. Sentiment analysis determines sentiment by utterance, and looks at those utterances in the context of the words around them.

Along with providing business intelligence regarding the customer experience, sentiment analysis can also be deployed to improve the employee experience and monitor performance. As it does with customer communications, sentiment analysis can also provide automated sentiment scores for agents, teams, groups, and so on. Employee sentiment scores can be correlated with other key performance indicators (KPIs), guiding management toward both underperforming and highly performing agents.

With its focus on both the customer experience and the employee experience, sentiment analysis is the classic case of a solution being in the right place at the right time. Sentiment analysis is poised to join an elite class of contact center solutions that holds universal appeal to the market.

To learn more about Sentiment Analysis–check out A Guide to Call Center Sentiment Analysis.

A 28 year industry veteran, Paul Stockford has authored articles on telecommunications, contact centers, and customer service for several major industry journals in the U.S. and Europe.  He was a monthly columnist in a variety of industry magazines including Voice Processing, Enterprise Communications, Voice+ (Great Britain), Customer Interface, CRM Magazine and Call Center magazine.  He currently authors the monthly “View from the Saddle” column in Contact Center Pipeline magazine.  He is widely quoted in business and general interest periodicals, and has held positions on the Editorial Advisory Boards of several magazines in the U.S. and Europe.

Guest Contributor