Customer sentiment analysis: When CSAT surveys let you down

It can be hard to gauge your customer sentiment, and CSAT surveys, though helpful, can present some major limitations.

When it comes to measuring customer sentiment, why is CSAT not enough?

CSAT surveys take a one-size-fits-all approach to customer sentiment—your CSAT score can easily be distorted by simple issues with easy solutions. Plus, you can’t be sure that your customers are rating you on the right thing—are they rating the quality of the service interaction they just had, or are they scoring the quality of a product they just purchased from you?

CSAT surveys are a good indicator of how happy your customers are, but they really are only an indicator

The fact of the matter is, customers don’t always fill out CSAT surveys. In fact, response rates linger around 5-30%. This presents two problems: 

1. Not enough data: everyone knows that the more data you have, the more accurate your conclusions will be. When you have a small percentage of returned CSAT surveys, you are likely to be working with a less-than-accurate representation of your customers’ true feelings.

2. The 5% rule: Remember that your 90% CSAT may actually lie between 85% and 95%. And if you have a low percentage of submitted surveys, this 5% deviation can be even higher.

So, is your CSAT score just a vanity metric?

Not necessarily, but:

Businesses that exclusively use customer satisfaction scores have a churn rate of 60-80%, including customers who indicated they were “satisfied” or even “very satisfied” in their last CSAT survey. — Fred Reichheld

CSAT alone should never be used to measure your customer sentiment, so it’s often coupled with NPS (Net Promoter Score) or CES (Customer Effort Score) to try and give a more accurate reading. This combination gives a much better indication of customer sentiment, but, even so, a major flaw in CSAT remains:

Only 1 in 26 unhappy customers complain… and 91% of unhappy customers simply leave without complaining. 

You need a way of analyzing how every customer feels, especially the ones that aren’t giving you feedback. This is where customer sentiment analysis and sentiment scores enter the picture. 

What are the benefits of customer sentiment analysis?

Introducing customer sentiment analysis to your organization will give you a holistic view of how happy your customers are.

• It tells you how happy the customers who don’t return surveys are.

• It’s no extra effort for customers, so it doesn’t detract from the customer experience.

• It provides data that’s representative of your whole customer base rather than just a vocal portion.

Of course, customer sentiment analysis isn’t perfect, but it can be incredibly useful for bringing you closer to your customers and understanding a) what they are very happy with; and b) where their frustrations are coming from. 

When used in conjunction with survey results and KPIs, customer sentiment scores can be very powerful.   

How does a customer sentiment score work?  

Sentiment analysis essentially detects whether a customer is happy or unhappy after an interaction with your support team. 

Here’s one way to do it:

Goal: score your most recent customer conversations to determine whether customers are happy, unhappy, or neutral. 

You will have to score each word on how positive or negative it is to achieve this. And don’t forget emojis 😃 they can also be a great indicator of a customer’s state of mind. 

Adjectives and verbs are given more weight in the analysis because they are the most important factor in understanding a customer’s true feelings. Emojis are also a good indicator of customer sentiment or customer satisfaction.  Sentiment analysis essentially detects whether a customer is delighted, happy, indifferent, unhappy, or angry after an interaction with your support team. 

Let’s look at two customer messages:

“This is extremely frustrating. The app keeps glitching. I’ve tried for hours to get a ride, while the app says the car should just be minutes away. Worthless service 😕.”

Here, “extremely,” “frustrating,” “worthless,” and the emoji: 😕 would be weighted negatively, resulting in a sentiment score that is well below average. 

Now let’s look at a more cheerful message:

Dixa is a great asset for our whole team. Our agents love it! 😁”

Here, “great,” “love,” and the emoji would receive positive points and the sentiment would be marked as positive. 

Messages are considered positive, negative, or neutral depending on syntax, semantics, and the sentiment score of each word. Neutral messages are usually factual descriptions of the user experience, while the customer perspective is made up of either positive or negative words.

This method allows you to get a view of all your negative customer interactions and break them down by contact reason, team, channel or any other dimension. By doing this, you can pinpoint exactly why your customers are dissatisfied and fix it. 

But it’s worth considering that…

Language is very complex and highly subjective. 

In the US, for example, “that’s sick!” is a positive phrase… The same is true for “painfully good.”

One drawback of sentiment analysis is that it’s hard to differentiate between “sick” being used in a positive way or “sick” being used negatively: “Your customer service makes me sick…” for example.  In longer messages, this won’t be a problem (the rest of the message will reflect a positive score despite “sick” being coded as negative), but shorter messages with similar slang will make scores less accurate. 

The reality is that what makes humans so brilliant – that every person is an individual with their own way of expressing themselves, is also what makes it tricky to gauge a customer’s sentiment correctly every single time. 

This is why we recommend combining sentiment analysis with CSAT, NPS, or CES, for a holistic approach that brings you as close to your customers, and their actual experiences, as possible. 

How to get up and running with customer sentiment analysis

First, you will need some way to process and analyze your data. There are a few ways to do this:

1. DIY: You can compile a list of words that you think express positive or negative sentiment and give them scores based on how powerful they are. You won’t catch every single word, but if you have a strong list, this could prove to be valuable. There are a few open-source libraries that can show you how to score conversations and how to weigh certain words. Full disclosure, this will be an arduous process — but, when done right, it can produce good results. Bear in mind, it might involve developer resources. 

2. Machine learning: This will absolutely require a lot of help from developers! You will need to read and code comments manually (for example, using Python), and assign scores to particular words or phrases that suggest a positive, negative, or neutral sentiment, and then weigh them based on how strong the feeling is. This could very well pay off in the long run, but it requires a lot of set-up work.

3. Find an out-of-the-box customer sentiment analysis solution: This could be a good option if you could benefit from sentiment analysis, but you don’t necessarily have the developer resources to build it yourself.

Once you have the foundations set up, you need to start thinking about how to use these actionable insights. 

3 practical applications of customer sentiment analysis

1. Getting a read on customers who don’t return surveys

Customer service sentiment analysis is an effective thermometer for how your customers are feeling. You can drill down into any dimension: Which channels are customers having a hard time with? What contact reason brings the most pain? Which product is associated with the most negative responses? 

2. Understanding why your customers are unhappy

After you’ve taken their temperature, you can diagnose and prescribe treatment. For example, a particular product is associated with a lower customer sentiment score, so let’s alert the product team to move the improvements up on their roadmap. This can prevent customer churn and result in massive retention gains.

3. Tracking customer happiness

To continue the metaphor – next comes rehabilitation. How are you going to make sure your customer experience is constantly improving? Sentiment is a great indicator of how your customers are responding to changes in your products and services over time. 

Customer sentiment analysis is a team sport

Introducing sentiment analysis to your customer service operations can provide you with valuable insights, but you can make it work even harder for you by combining it with other indicators like CSAT, NPS, or CES. If you’d like to learn more about customer sentiment analysis, schedule a personalized demo with us today.


Francesca Valente

Share this article

Get CX insights, proven strategies, & the latest news in your inbox every month!

You might also enjoy

Customer service
7 min read

5 ways to boost customer retention with AI and measure it

Read more
Dixa life
7 min read

Dixa kicks off 2024 with 45 new badges and awards in G2 winter report

Read more
Dixa life
7 min read

Dixa completes SOC 2 type 1 certification, reinforcing commitment to data security 

Read more