Blog Home

What is sentiment analysis? Examples, best practices, & more


The Team at CallMiner

April 30, 2019

Woman cupping ear to hear better
Woman cupping ear to hear better

A Definition of Sentiment Analysis

Sentiment analysis is a method for gauging opinions of individuals or groups, such as a segment of a brand’s audience or an individual customer in communication with a customer support representative. Based on a scoring mechanism, sentiment analysis monitors conversations and evaluates language and voice inflections to quantify attitudes, opinions, and emotions related to a business, product or service, or topic. Sentiment analysis is sometimes also referred to as opinion mining. As part of the overall speech analytics system, sentiment analysis is the integral component that determines a customer’s opinions or attitudes.

How Sentiment Analysis Works

Sentiment analysis is often driven by an algorithm, scoring the words used along with voice inflections that can indicate a person’s underlying feelings about the topic of a discussion. Sentiment analysis allows for a more objective interpretation of factors that are otherwise difficult to measure or typically measured subjectively, such as:

  • How fast the individual is speaking (rate of speech)
  • Changes in the level of stress indicated by the person’s speech (such as in response to a solution provided by a customer support representative)
  • The amount of stress or frustration in a customer’s voice
Product Video
CallMiner for Experience Management - Retail
Discover how CallMiner can help you understand customer sentiment and improve experiences
Right Arrow

In customer service and call center applications, sentiment analysis is a valuable tool for monitoring opinions and emotions among various customer segments, such as customers interacting with a certain group of representatives, during shifts, customers calling regarding a specific issue, product or service lines, and other distinct groups.

Sentiment analysis may be fully automated, based entirely on human analysis, or some combination of the two. In some cases, sentiment analysis is primarily automated with a level of human oversight that fuels machine learning and helps to refine algorithms and processes, particularly in the early stages of implementation.

Examples of Sentiment Analysis

Sentiment analysis is used across a variety of applications and for myriad purposes. For instance, sentiment analysis may be performed on Twitter to determine overall opinion on a particular trending topic. Companies and brands often utilize sentiment analysis to monitor brand reputation across social media platforms or across the web as a whole.

One of the most widely used applications for sentiment analysis is for monitoring call center and omnichannel customer support performance. As companies seek to keep a finger on the pulse of their audiences, sentiment analysis is increasingly utilized for overall brand monitoring purposes.

Sentiment analysis has been used by political candidates and administrations to monitor overall opinions about policy changes and campaign announcements, enabling them to fine-tune their approach and messaging to better relate to voters and constituents. In brand reputation management applications, overall trends in sentiment analysis enables brands to identify peaks and valleys in overall brand sentiment or shifts in attitudes about products or services, thus enabling companies to make improvements perfectly in-tune with customer demands.

The CX Pro's Guide to Speech Analytics
The CX Pro’s Guide to Speech Analytics
Learn how to leverage speech analytics for improved CX
Right Arrow

4 Sentiment Analysis Use Cases

1. Segment Buyer Groups Based on Opinions:

Tracking sentiment allows an organization to see which customers are more opinionated than others. For instance, many believe that 80% of customer issues come from 20% of buyers. If this stat happens to be true, you will be able to segment the qualities of that group and either fix common issues or even avoid those buyers. (Of course, avoiding buyers would have to mean there is little to no ROI based on the level/type of opinions of said group.)

2. Plan Product/Service Improvements:

Analyzing customer opinions is a treasure trove of data, especially when it comes to what you sell. Updating software products, improving the design of physical goods or bettering your services can all come from customer sentiment. At times, this data can even yield new products/services for your business to offer.

3. Plan Process Improvements:

Customer sentiment isn’t always positive. However, negative feedback isn’t necessarily false. These opinions may need sorting out in a systematic way, meaning improving your overall customer service (or other) process.

4. Continuously Track Sentiment Over Time:

Sentiment is a metric worth continually checking. As you improve both your processes and products, opinions will change. Seeing these changes allow for better navigating the tumultuous waters of sentiment.

Benefits of Sentiment Analysis

A relative sentiment analysis score provides insight into the effectiveness of call center agents and customer support representatives and also serves as a useful measurement to gauge the overall opinion on a company’s products or services. When sentiment analysis scores are compared across certain segments, companies can easily identify common pain points, areas for improvement in the delivery of customer support, and overall satisfaction between product lines or services.

By monitoring attitudes and opinions about products, services, or even customer support effectiveness continuously, brands are able to detect subtle shifts in opinions and adapt readily to meet the changing needs of their audience.

Challenges and Best Practices for Sentiment Analysis

Language is complex, and as a process for quantifying and scoring language, sentiment analysis is equally complex. What is relatively easy for humans to gauge subjectively in face-to-face communication, such as whether an individual is happy or sad, excited or angry, about the topic at hand, must be translated into objective, quantifiable scores that account for the many nuances that exist in human language, particularly in the context of a discussion. For instance, a word that otherwise carries a positive connotation used in a sarcastic manner could easily be misinterpreted by an algorithm if both context and tone are not taken into consideration.

Given these challenges, customer analytics software vendors must consider acoustic measurements (the rate of speech, stress in a caller’s voice, and changes in stress signals) in the context of the conversation. Additionally, integrating machine learning into the mix enables sentiment analysis to become more accurate over time, as algorithms learn and adapt to the commonalities in conversations and how the context of conversations can change outcomes.

In call center and customer support applications, sentiment analysis that operates in real-time provides crucial feedback to agents and representatives, allowing them to respond appropriately to impact the outcome of the interaction. When used in conjunction with powerful features like Semantic Building Blocks, companies can easily monitor agent performance by identifying positive patterns such as frustration or dissatisfaction in the early stages of a call followed by positive sentiments in the later portion of the call, representing an agent’s ability to talk-down frustrated customers.

When utilizing the right technology tools and applying it to key business drivers, sentiment analysis is a powerful tool for steering companies and their individual business units to successful outcomes from every customer interaction and business performance improvement.

Expert Sentiment Analysis Tips & Best Practices

1. Understand how to classify sentiment based on the different approaches.

Machine Learning: This approach employs a machine-learning technique and diverse features to construct a classifier that can identify text that expresses sentiment. Nowadays, deep-learning methods are popular because they fit on data learning representations.

“Lexicon-Based: This method uses a variety of words annotated by polarity score, to decide the general assessment score of a given content. The strongest asset of this technique is that it does not require any training data, while its weakest point is that a large number of words and expressions are not included in sentiment lexicons.

“Hybrid: The combination of machine learning and lexicon-based approaches to address Sentiment Analysis is called Hybrid. Though not commonly used, this method usually produces more promising results than the approaches mentioned above.” – Symeon Symeonidis, 5 Things You Need to Know About Sentiment Analysis and Classification, KDNuggets; Twitter: @kdnuggets

2. Ensure your methods are ethical, unlike some examples.

“With technology’s increasing capabilities, sentiment analysis is becoming a more utilized tool for businesses. Social media monitoring tools use it to give their users insights about how the public feels in regard to their business, products, or topics of interest.

“It’s widely used by email services to keep spam out of your inbox and by review websites to recommend new content like films or TV shows.

“However, it has been used in more murky circumstances. Facebook, for example, came under fire when it was discovered they were using sentiment analysis to see if they could manipulate people’s emotions by altering their algorithms to inject negative or positive posts more frequently into their users’ news feeds.

“By using this process of “emotional contagion,” they found that they could decisively influence their users’ emotional output by flooding their news feeds with positive or negative posts. The big problem is that Facebook never informed its users that they were part of an experiment and may have caused emotional distress to them in some cases.

Clearly we can see how this use of sentiment analysis can be problematically unethical.” – Adam Coombs, Understanding Sentiment Analysis in Social Media Monitoring, Unamo; Twitter: @UnamoHQ

3. Analyze competition sentiment, in addition to your own.

“In addition to monitoring your own online mentions, you can also track your competitors’ mentions to see how your business stacks up. Positive sentiments help you pinpoint where your competitors are succeeding. Negative sentiments can reveal opportunities for your business. For example, a groundswell of negative attitudes toward a competitor’s product redesign might reveal an opportunity for your product to fill a void.

“Following your competitors’ mentions can also help you identify areas where you can improve. If a competitor’s marketing campaign is getting higher marks than yours, examine it in detail and identify what tactics are most effective. Use these insights as inspiration for your next campaign.” – Milosz Krasinski, Sentiment Analysis: What Marketers Need to Know, Social Media Examiner; Twitter: @SMExaminer

Contact Center Operations Speech & Conversation Analytics Voice of the Customer North America EMEA APAC