CSAT

How to improve CSAT with customer sentiment analysis

How to improve CSAT with customer sentiment analysis
Content Manager & Customer Service Expert
LinkedIn icon
How to improve CSAT with customer sentiment analysis

Many companies often struggle with low CSAT ratings.

They might want to take steps to address the poor ratings, but often face a major challenge - they do not understand what’s leading to such low CSAT.

As a result, they are unable to take corrective actions to improve their CSAT.

One such company is Glammmup, a leading e-Commerce company dealing with online makeup sales.

The industry average CSAT in the online retail space is 78. Glammmup was struggling with a CSAT of 62 - a dismal show for a company that takes pride in its superior customer service.

Their head of customer service, Emily, wanted to rectify this. She started by analysing what the current process looks like, understanding the gaps in the process, and addressing them with innovative solutions.

Let’s look at Emily’s journey as she undertakes this huge endeavour of improving Glammmup’s CSAT rating from a sorry 62 to over the industry average.

The current process of analysing CSAT

Emily’s first step was to analyse what the company was currently doing. Like most companies, Emily found that Glammmup had a pretty straightforward way of analysing CSAT.

1/ After each interaction with customer service, Glammmup sent out a CSAT survey to their customers via email or SMS

2/ The surveys were simple, consisting of 2 questions - one asking to rate the service and the other asking for further explanations

A typical CSAT survey
A typical CSAT survey

3/ They compiled the data from these surveys and analysed it

4/ The quantitative question was used to measure the aggregate rating.

5/ A few months ago, Glammmup also started manually analysing their qualitative data to find overarching trends and insights

Problems with the current process

While the current process was effective to a certain extent, it failed to provide Glammmup with the kind of insights that would enable them to make impactful changes.

There were several gaps in the process. Let’s look at some of them.

1/ Low completion rate

Glammmup sent out CSAT surveys to all their customers post all interactions, however, hardly 6% of those surveys were ever completed.

Even fewer customers filled out the comments section - a circumstance which wasn’t conducive to producing rich insights for 2 reasons.

Firstly, such a low completion rate meant it wasn’t a true representation of the entire population of their customers.

Secondly, the data sample was not large enough to provide confident qualitative insights.

2/ Unscalable, inconsistent process

Glammmup were manually analysing all their CSAT data. A situation that was far from ideal.

Therefore, the analysis was subjective, based on the analyst’s opinion and without a well-documented SOP, it was inconsistent.

Additionally, with the number of responses increasing, the process was beginning to become unscalable and difficult to manage.

3/ Lack of rich insights

The most significant problem with current CSAT driver analysis was it wasn’t giving granular, detailed insights.

The major reason behind this was CSAT’s simplicity. With very little and often generic insights, Glammmup were struggling to get to the root of topics.

4/ Lack of distinction between product and agent CSAT

While the CSAT rating is considered to be a reflection of customer service quality, customers sometimes do not consider the quality of service at all.

Some types of issues almost always lead to bad CSAT ratings irrespective of how an agent handles the query. Such ratings lead to a biased representation of CSAT.

For example, let's look at Glammmup's CSAT driver analysis. you can notice the positive drivers are all due to good customer service whereas the negative drivers are related to product or operations issues.

CSAT Driver Analysis - Positive and Negative Drivers
CSAT Driver Analysis - Positive and Negative Drivers

We’ll talk about this in more detail in our next article on how to separate agent and product CSAT. 

Now that Emily knew the problems with the process, Emily understood Glammmup needed a better approach.

Improving CSAT with customer sentiment analysis

While hunting for better approaches, Emily realised they already had all the data they needed. They were sitting on a goldmine of customer data that they were not utilising.

We're talking about customer service conversations.

Emily understood that the best way to understand why people were rating Glammmup’s service the way they were was - by combining CSAT driver analysis with sentiment analysis of customer service data.

Customer sentiment analysis is the process of understanding how customers are feeling about your product or service.

And when done using customer service data, it becomes a powerful vehicle to uncover true, granular, and detailed customers insights such as their pain points or problems.

Therefore, when we talk about customer sentiment analysis here- we are referring to the analysis of customer service data.

By analsying their customer service data for sentiments, Glammmup could connect the dots between when a customer complained of something and how that affected the CSAT.

Let’s look at the exact framework they came up with.

The 4-step process to improving CSAT ratings

1/ Capture data from each customer at every touchpoint

While CSAT surveys are sent only after a major touchpoint, customer sentiment analysis of support conversations captures data across the spectrum of a customer lifecycle.

This meant tracking all the good and bad things that customers are sharing through their journey with the company.

As a result, Glammmup had the quality and quantity of data that they needed to conduct deep analysis.

For instance, Emily realised that 60% of the customers who reached out to customer service were likely to give a bad CSAT rating irrespective of the quality of service.

2/ Combine customer sentiment analysis with CSAT driver analysis

CSAT driver analysis is a great tool that gives a generic overview of what’s driving poor CSAT ratings. But it is not complete.

By combining it with a holistic customer sentiment analysis program, Glammmup were able to get deeper into those generic insights and pinpoint exactly what’s causing customers to leave negative ratings or comments.

Additionally, Glammmup were also able to validate hypotheses that they created based on their CSAT data.

For example, the CSAT analysis surfaced insights like the top driver of poor rating was a long resolution time.

By combining this with customer sentiment analysis of support data, Glammmup were able to uncover that long resolution time was typically caused by issues relating to refunds.

As a result, they could make confident decisions based on data-backed insights.

3/ Make customer service changes

While CSAT is sent out as a way to gauge how customer service agents are performing, the responses are often muddled by how other teams are performing, something that is not under the control of the support team.

By using customer service data and segregating it based on different departments, the customer service team were able to identify the issues plaguing their team and fix those at the core.

4/ Make company-wide changes

Related to the previous point, the customer service team at Glammmup were able to send reports to other teams that were causing customers to leave negative ratings.

Improve CSAT
Support Insights Podcast with Valeria Kast, Prinfity

Listen to Valeria as she speaks to us in our latest podcast episode on 'Boosting CSAT scores and reducing first response times at Printify'

Let’s face it, for a customer, the entire company works as a whole and that’s how it should be.

With the Glammmup team having identified critical product and operations issues that drive poor csat ratings, they were able to send those insights to the respective teams.

As a result, the entire customer experience was improved as a whole.

Closing thoughts

By following the 4 step process, Glammmup were able to improve its CSAT score from 63, a score way below the industry average of 78 to 82.

And all this within a year.

If you’re struggling with poor CSAT ratings and wondering what’s going to move the needle for you, customer sentiment analysis is an excellent process to look at.

While CSAT driver analysis is important, combining it with customer sentiment analysis can help you understand what’s truly driving your CSAT scores.

It will enable you to understand your customers at a granular level, make company-wide changes, and improve the overall customer experience that ultimately affects how customers score you.

Join a community of 2139+ customer-focused professionals and receive bi-weekly articles, podcasts, webinars, and more!

Trending articles

CSAT

How to improve CSAT with customer sentiment analysis

Piusha Debnath
Content Manager & Customer Service Expert
In this article
Understand your customer’s problems and get actionable insights
See pricing

Is your AI accurate, or am I getting sold snake oil?

The accuracy of every NLP software depends on the context. Some industries and organisations have very complex issues, some are easier to understand.

Our technology surfaces more granular insights and is very accurate compared to (1) customer service agents, (2) built-in keyword tagging tools, (3) other providers who use more generic AI models or ask you to build a taxonomy yourself.

We build you a customised taxonomy and maintain it continuously with the help of our dedicated data scientists. That means the accuracy of your tags are not dependent on the work you put in.

Either way, we recommend you start a free trial. Included in the trial is historical analysis of your data—more than enough for you to prove it works.

Many companies often struggle with low CSAT ratings.

They might want to take steps to address the poor ratings, but often face a major challenge - they do not understand what’s leading to such low CSAT.

As a result, they are unable to take corrective actions to improve their CSAT.

One such company is Glammmup, a leading e-Commerce company dealing with online makeup sales.

The industry average CSAT in the online retail space is 78. Glammmup was struggling with a CSAT of 62 - a dismal show for a company that takes pride in its superior customer service.

Their head of customer service, Emily, wanted to rectify this. She started by analysing what the current process looks like, understanding the gaps in the process, and addressing them with innovative solutions.

Let’s look at Emily’s journey as she undertakes this huge endeavour of improving Glammmup’s CSAT rating from a sorry 62 to over the industry average.

The current process of analysing CSAT

Emily’s first step was to analyse what the company was currently doing. Like most companies, Emily found that Glammmup had a pretty straightforward way of analysing CSAT.

1/ After each interaction with customer service, Glammmup sent out a CSAT survey to their customers via email or SMS

2/ The surveys were simple, consisting of 2 questions - one asking to rate the service and the other asking for further explanations

A typical CSAT survey
A typical CSAT survey

3/ They compiled the data from these surveys and analysed it

4/ The quantitative question was used to measure the aggregate rating.

5/ A few months ago, Glammmup also started manually analysing their qualitative data to find overarching trends and insights

Problems with the current process

While the current process was effective to a certain extent, it failed to provide Glammmup with the kind of insights that would enable them to make impactful changes.

There were several gaps in the process. Let’s look at some of them.

1/ Low completion rate

Glammmup sent out CSAT surveys to all their customers post all interactions, however, hardly 6% of those surveys were ever completed.

Even fewer customers filled out the comments section - a circumstance which wasn’t conducive to producing rich insights for 2 reasons.

Firstly, such a low completion rate meant it wasn’t a true representation of the entire population of their customers.

Secondly, the data sample was not large enough to provide confident qualitative insights.

2/ Unscalable, inconsistent process

Glammmup were manually analysing all their CSAT data. A situation that was far from ideal.

Therefore, the analysis was subjective, based on the analyst’s opinion and without a well-documented SOP, it was inconsistent.

Additionally, with the number of responses increasing, the process was beginning to become unscalable and difficult to manage.

3/ Lack of rich insights

The most significant problem with current CSAT driver analysis was it wasn’t giving granular, detailed insights.

The major reason behind this was CSAT’s simplicity. With very little and often generic insights, Glammmup were struggling to get to the root of topics.

4/ Lack of distinction between product and agent CSAT

While the CSAT rating is considered to be a reflection of customer service quality, customers sometimes do not consider the quality of service at all.

Some types of issues almost always lead to bad CSAT ratings irrespective of how an agent handles the query. Such ratings lead to a biased representation of CSAT.

For example, let's look at Glammmup's CSAT driver analysis. you can notice the positive drivers are all due to good customer service whereas the negative drivers are related to product or operations issues.

CSAT Driver Analysis - Positive and Negative Drivers
CSAT Driver Analysis - Positive and Negative Drivers

We’ll talk about this in more detail in our next article on how to separate agent and product CSAT. 

Now that Emily knew the problems with the process, Emily understood Glammmup needed a better approach.

Improving CSAT with customer sentiment analysis

While hunting for better approaches, Emily realised they already had all the data they needed. They were sitting on a goldmine of customer data that they were not utilising.

We're talking about customer service conversations.

Emily understood that the best way to understand why people were rating Glammmup’s service the way they were was - by combining CSAT driver analysis with sentiment analysis of customer service data.

Customer sentiment analysis is the process of understanding how customers are feeling about your product or service.

And when done using customer service data, it becomes a powerful vehicle to uncover true, granular, and detailed customers insights such as their pain points or problems.

Therefore, when we talk about customer sentiment analysis here- we are referring to the analysis of customer service data.

By analsying their customer service data for sentiments, Glammmup could connect the dots between when a customer complained of something and how that affected the CSAT.

Let’s look at the exact framework they came up with.

The 4-step process to improving CSAT ratings

1/ Capture data from each customer at every touchpoint

While CSAT surveys are sent only after a major touchpoint, customer sentiment analysis of support conversations captures data across the spectrum of a customer lifecycle.

This meant tracking all the good and bad things that customers are sharing through their journey with the company.

As a result, Glammmup had the quality and quantity of data that they needed to conduct deep analysis.

For instance, Emily realised that 60% of the customers who reached out to customer service were likely to give a bad CSAT rating irrespective of the quality of service.

2/ Combine customer sentiment analysis with CSAT driver analysis

CSAT driver analysis is a great tool that gives a generic overview of what’s driving poor CSAT ratings. But it is not complete.

By combining it with a holistic customer sentiment analysis program, Glammmup were able to get deeper into those generic insights and pinpoint exactly what’s causing customers to leave negative ratings or comments.

Additionally, Glammmup were also able to validate hypotheses that they created based on their CSAT data.

For example, the CSAT analysis surfaced insights like the top driver of poor rating was a long resolution time.

By combining this with customer sentiment analysis of support data, Glammmup were able to uncover that long resolution time was typically caused by issues relating to refunds.

As a result, they could make confident decisions based on data-backed insights.

3/ Make customer service changes

While CSAT is sent out as a way to gauge how customer service agents are performing, the responses are often muddled by how other teams are performing, something that is not under the control of the support team.

By using customer service data and segregating it based on different departments, the customer service team were able to identify the issues plaguing their team and fix those at the core.

4/ Make company-wide changes

Related to the previous point, the customer service team at Glammmup were able to send reports to other teams that were causing customers to leave negative ratings.

Improve CSAT
Support Insights Podcast with Valeria Kast, Prinfity

Listen to Valeria as she speaks to us in our latest podcast episode on 'Boosting CSAT scores and reducing first response times at Printify'

Let’s face it, for a customer, the entire company works as a whole and that’s how it should be.

With the Glammmup team having identified critical product and operations issues that drive poor csat ratings, they were able to send those insights to the respective teams.

As a result, the entire customer experience was improved as a whole.

Closing thoughts

By following the 4 step process, Glammmup were able to improve its CSAT score from 63, a score way below the industry average of 78 to 82.

And all this within a year.

If you’re struggling with poor CSAT ratings and wondering what’s going to move the needle for you, customer sentiment analysis is an excellent process to look at.

While CSAT driver analysis is important, combining it with customer sentiment analysis can help you understand what’s truly driving your CSAT scores.

It will enable you to understand your customers at a granular level, make company-wide changes, and improve the overall customer experience that ultimately affects how customers score you.

Frequently asked questions

Is your AI accurate, or am I getting sold snake oil?

The accuracy of every NLP software depends on the context. Some industries and organisations have very complex issues, some are easier to understand.

Our technology surfaces more granular insights and is very accurate compared to (1) customer service agents, (2) built-in keyword tagging tools, (3) other providers who use more generic AI models or ask you to build a taxonomy yourself.

We build you a customised taxonomy and maintain it continuously with the help of our dedicated data scientists. That means the accuracy of your tags are not dependent on the work you put in.

Either way, we recommend you start a free trial. Included in the trial is historical analysis of your data—more than enough for you to prove it works.

Do you integrate with my systems? How long is that going to take?

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

What size company do you usually work with? Is this valuable for me?

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

What is your term of the contract?

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

How do you keep my data private?

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

Support Insights Community
Join a community of 2200+ customer-focused professionals and receive bi-weekly articles, podcasts, webinars, and more!

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.