Customer sentiment is quickly becoming one of the most important metrics for any contact center to track. It helps contact centers better understand the opinions and emotions of their customers so that they can improve their processes and provide a better customer experience.
It can also help uncover major issues outside of the contact center so that they can be fixed, taking pressure off of the contact center and its agents.
In a recent discussion with a contact center expert about the most important metrics for any call center, customer sentiment was identified as the single most important quality KPI for contact centers to track, measure, and coach for.
Thomas Laird, founder and CEO of the 500-seat hybrid contact center Expivia Interaction Marketing Group, believes that “customer sentiment is the greatest call center invention in the last couple of years.” That’s high praise considering all the technological advancements that call centers have seen in recent years.
So, we decided to unpack customer sentiment to better understand what it is, how to track it, and what makes it so valuable for any contact center.
What is customer sentiment?
In order to understand customer sentiment, you first need to understand the technology behind it.
Customer sentiment is made possible by sentiment analysis technology, which uses things like algorithms, machine learning, and Natural Language Processing (NLP) to “listen” to conversations and determine the “mood” or sentiment of the speaker.
This goes beyond just dictation and incorporates tone, feeling, and emotion to produce a “score” of positive, negative, or neutral.
There are many different sentiment analysis software available for contact centers to choose from. Big Tech names like Google, IBM, and Microsoft make their own sentiment analysis tools. Some smaller companies have also made their name by specializing in sentiment analysis software.
For contact centers looking to track and measure customer sentiment scores, they’ll need to find the provider that’s best suited for their needs. Some will come complete and ready to use with APIs and integrations, others can be built up internally using open-source software.
The software behind sentiment analysis can complement telephony systems, traditional analytics technologies, and reporting tools so that it automatically monitors incoming and outgoing communications and appears in dashboards alongside other omnichannel and workforce related data.
Basically, conversations between agents and customers – including audio, chat, social media, and email – are put through a computer so that artificial intelligence can determine how the customer was feeling during the conversation. This has huge implications on quality within customer support and can help evolve the contact center from a cost center to a profit center.
What makes it so valuable?
Customer sentiment analysis gives insights into the quality of conversations at a level that contact centers have never seen before. Before sentiment analysis, the methods for understanding how customers felt about a brand or its products and services were limited. Contact centers couldn’t do much more to gather customer feedback than listen to hours of calls, read through countless memos, or conduct surveys that received limited responses.
Now, technology allows contact centers to get a real-time gauge of how a customer feels about a brand, its products, a specific phone conversation, and even the agent that they’re speaking to. It gives leaders a measure of not only which agents are providing quality services, but also how often their agents are able to impact customer satisfaction by turning customer sentiment from negative to positive.
As Laird points out:
“In the past, the call center always got blamed for everything. If you had a customer who called in and they were mad, it’s because they had a bad experience with the rep. That’s what upper management would always think. It all rolled down at the call center. Well now with analytics and with looking at first call resolution and the issues that customers are having, we can see trending keywords. So, we can [see] this isn’t a call center problem, this is a problem with, maybe there are a couple of dead routes in the IVR [or] maybe there’s a problem on the website.”
The ability to examine every conversation and track it over time, down to the exact words used, gives contact centers a much greater ability to impact overall customer satisfaction. With customer sentiment “We’re not just using it and trying to solve one customer’s problem, but we’re seeing if there’s an issue over a bunch of customers with our analytics and being able to solve it,” says Laird.
This provides a huge competitive advantage and allows brands to stay on top of customer opinion or intervene where possible.
Going beyond the call
Customer sentiment, and sentiment analysis as a whole, goes beyond just gauging the emotions of a caller. It can also help uncover hidden issues that would otherwise go unnoticed.
For example, sentiment analysis can identify trending keywords that are occurring frequently in calls. If the term “too expensive” is appearing more often after a price increase, the contact center can notify other business segments about the change in customer sentiment.
This can make the contact center an asset when it comes to things like marketing campaigns and new product testing. The data science of sentiment analysis and the actionable insights it provides makes it a valuable market research tool.
The real-time nature of sentiment analysis can also help take some pressure off of the contact center by identifying sentiment throughout the duration of a conversation. If 75% of all customer interactions are starting with negative sentiment toward a brand, product, or service, it shows that something outside the control of the contact center is to blame for a customer’s negative experience.
In addition, it can track how often agents are able to reverse customers from negative to positive sentiment, which can provide helpful evidence of a contact center’s value to the customer journey.
Better than NPS or CSAT?
In our conversation with Tom Laird, he shared how he feels about the more traditional quality KPIs that are commonly measured within contact centers. “I’m not a huge fan of NPS. I’m not a huge fan of CSAT either,” he said.
We found this a bit surprising since Net Promoter Score (NPS) and Customer Satisfaction (CSAT) are industry standbys and have been staple metrics among call centers for decades.
Laird went on to clarify exactly why he believes customer sentiment scores are better indicators of customer service levels. He explained that, because NPS involves conducting a voluntary survey where customers are asked to score an interaction on a scale of one to 10, the value of the responses is limited.
“The problem is that you really only have two groups of people that are responding to those surveys. One group is those that were super happy with the call. [With the second group,] they hated the call or they didn’t get what they wanted,” says Laird. That means that NPS is only reflecting customers at the extremes at each end of the spectrum.
While NPS can be a good indicator of some customers’ feelings about their customer journey, it does very little to show how well an agent is performing.
CSAT presents many of the same problems. It relies on voluntary participation and fails to provide a complete picture of customer satisfaction. While the questions asked in a CSAT survey are more detailed and provide more information than NPS, CSAT scores tend to be more reflective of an individual interaction rather than the whole customer journey.
Customer sentiment on the other hand automatically evaluates every interaction, every word used in the interaction, the mood that those words reflect, and how the mood changed over time. This more holistic approach to gathering customer data helps to improve customer interactions, strengthen customer relationships, reduce negative reviews, and build greater customer retention.
Although Laird went on to make it very clear that he sees value in both NPS and CSAT, he also spoke passionately about customer sentiment as a far more valuable metric to focus on going forward.
An added benefit of sentiment analysis technology is that it can track sentiment on both sides of the call. This allows contact centers to track the mood and word usage of their agents.
While it may seem a bit Orwellian, recording phone calls as a way to measure an agent’s performance is nothing new. With sentiment analysis, this can now be done automatically with the help of speech analytics. Agent sentiment becomes a much more powerful, data-backed indicator of how well an agent or group of agents are handling conversations than the subjective practice of listening in to calls.
As Laird told us, “[it] gives a full summary of how our QA is going for the day as well because, if we have positive agent sentiment, I know that we’re using proper tone, proper words, and we’re doing a pretty good job.”
Sentiment analysis has been a game-changer for contact centers. It elevates their ability to understand how customers view their brand and delivers new insights into the customer journey. As the technology that enables sentiment analysis continues to evolve, so will the information that it can provide. One thing is for sure, it’s already become the most important quality metric for contact centers.
Ryan Plank is a content marketer with a degree in Journalism and a background in technology. He lives in Orlando, Florida, and is an avid golfer.