UseDesk меню

CSAT, CES, NPS: WHICH METRICS TO USE TO MEASURE CUSTOMER SATISFACTION

Alexandra Shiryaeva
Chief Customer Officer at UseDesk
Jokes aside, it's time to get serious about the metrics. You have a customer support department, and it seems to be working fine. Someone answers the customers' questions and resolves the issues but you are not able to reflect this in your reports or reports of the company because you do not have any statistics and exact data. To measure or not to measure the performance of the customer support department is not the question really. You need to measure it! Otherwise, you will never know what needs to be improved, and what you can be proud of. Well, you may become aware of this at some point accidentally (and, most likely, it will be an unpleasant surprise).

There are several metrics that represent the state of the department and its productivity, and it is up to a manager to select the ones to use and prioritize them. The list is long: the response time, the resolution time, the length of the chain of those involved in the process of solving the problem, the effectiveness of the tools used internally by the support team, and so on. The most important of them is the score of your customers' satisfaction. If they are satisfied, then it is less critical whether your response time correlates with the standards commonly applied in customer service. It is important that the customers feel comfortable with your response time (for example, if your operators answer within 30 minutes instead of 10 but the client receives a full, precise, friendly and thoughtful answer, and the customer provides a positive feedback, then, making the response time a priority is simply unreasonable).

Once you have decided that a satisfied customer is your number one priority, a new task arises: now you have to decide which metric to use and how to use it because, even with the customers' satisfaction, it is not that straightforward.

The three basic metrics are constantly replace each other on the top of the charts of the metrics recommended for use in the customer service. All three of them are aimed at exploring how the customer lives with your product and support but they have nuances.

CSAT (Customer Satisfaction Score)

It is a basic and, probably, the most popular metric. You just ask the customer – 'how do you ...?' The question can be related to anything, and its purpose is to get an accurate evaluation of a particular case. For example, you ask the customer – 'How do you feel about the conversation with the support operator?', and give the option to select the answer: excellent, ok, very bad.

Some experts believe that this does not reflect customer loyalty and questions are too specific, however, this is a very good metric if you are only interested in the quality of the support.
SIDE-NOTE
CSAT metrics are pre-built into the ticketing systems. The screenshot above is an example form Usedesk. How it works: you set up the evaluation form in just a few clicks, and then the customer can evaluate the operator of your customer support team right after the conversation is over. Easy as one-two-three, and incompetent answers or dissatisfied customers will not be able to hide from you anymore.

CES (Customer Effort Score)

The bottom line is to find out how easy it was for the customer to resolve the issue/problem on his own, and how easy it was to contact the support. The developers of the metric had as their main goal getting the results that would allow to evaluate customers' comfort and satisfaction in general rather than evaluate their communication experience because, ideally, there should be no communication at all (a good support is a support which customers do not use because they have all the tools to overcome the difficulties).

There are two versions of the CES available:
Example of CES survey
1. The question about the effort. How much effort did you personally have to put forth to handle your request? (I would change it to "How easy was it to solve the problem?"), and it is difficult to translate the question to other languages ('effort' is not always translated from English in a right way). If we used a literal interpretation of the word, the question would be – "How much effort you made", and this does not sound polite. The answer is to be given on a 5-point scale, and it has caused some complications because it is not obvious whether 5 is the highest grade or the lowest grade. Thus, the second CES version was introduced.
Example of CES 2.0 survey
2. The organization made it easy for me to handle my issue (the meaning: <organization> made everything to solve the problem easy). The customer answers either agreeing or disagreeing with the statement. The chance of confusion is minimal, and, thus, the second version has become more popular.

NPS (Net Promoter Score)

Example of NPS survey
At the peak of interest to the customer service, everyone was talking about NPS – use NPS and, as a result, you will not need any other metrics, your wife will get back to you, and the old scars will disappear. However, there was not much talk about what NPS actually is.

The main question that NPS asks: how likely would it be for you to recommend a <brand> to friends or colleagues? Usually, the customer is given a scale from 1 to 10 to rate the degree of loyalty to the company.

This is an ideal metric to understand the relationship between the company and the customers in general. This is the best metric to use, if you are an analyst or a product specialist. This is an important metric to use, if you are not afraid to review the work you've done and correct the mistakes. Actually, the metric has little to do with the customer service because the question here does not relate to the competence of your operators or their accessibility, rather it helps to investigate the level of satisfaction with the product.
Which one to choose?
Since there is no common rule that you need to follow, it is best to try every metric and decide which ones are helpful.

I would say you should use two metrics: CSAT + NPS. Using the first one, you can have a closer look at what is the performance of the support department, and, using the second one, you can have insight about the customers' satisfaction with the product and learn whether they will recommend it or not. CSAT often contains the product reviews, and as they are not relevant to the support performance, they should be excluded from the assessment. NPS should be interesting not to you only buy to your product placement specialist also. It can be very entertaining to track the relationship between these two metrics: for example, CSAT grows, while NPS falls, or vice versa.
Hidden under a rock
We will discuss the ways to organize and prepare the surveys for the customers later. Now, let's cover some details that are not obvious.

When to ask?
The feedbacks on the support performance (CSAT) should be reviewed quickly, while the emotions are fresh and there is a chance to deal with a dissatisfied customer. The request to provide a feedback should be sent no later than in 8 hours after the issue has been resolved and the customer has confirmed he had no questions left (within 1 working day), and even better – immediately. If the support is provided through a live chat, it is better to send a survey through a live chat also, and not via email.

Customers evaluate the product, not the support
That is the reason we distinguish NPS from CSAT. As I mentioned before, do not hesitate to exclude any survey that relates to the product from CSAT.

Feedback
In addition to the option to rate the customer support experience, the customer should be able to leave a feedback, and the comments section should be obvious. Periodically, review the comments and react on them: not only to the claims or negative feedbacks, but also pay attention to the extremely positive comments.

Experiment on the question and answer forms
As soon as you notice that the customers do not participate in the survey actively, do not stumble – change the wording in the question, edit the answer choices. Often, the questions get more attention than the answers, and it is not fair. Even the effective classical form of "emoticon + description" can have number of variations. For example, replace "Average" with "Ok", "Good" with "Excellent", "Bad" with "Terrible", "Excellent" with "Super" and so on. After every change, review the statistics.

What is a good score?
At the end of the month, when you collect the very first of statistics, you may have a question – what do I do now? So, 90% of customers are happy. Is it good? Is it enough? What about 96%? or 89%?

It is hard to say which score is to be desired and encouraged as the companies and the customers are different. No one is able to say what your ideal score is. Obviously, the closer it is to 100%, the better, but it is better to define your own gradation of the scores. For example, the daily score today is 98-99% of satisfied customers, and the bonuses are given to the customer service department once the score of 96% is achieved. If you ask me why, I will not be able to answer. The best way to set the goals is to look at the companies that you respect and like. Find out how happy their customers are, and set your own satisfaction level bar.

Like vs. Dislike
Do not use CSAT with only two options of answer (bad-good, like-dislike). These answers are too broad, and you risk getting the overstated scores because the customers who feels more like "it's ok" will rarely choose "bad" as the answer. Thus, you may miss the fact that some customers have troubles.

Do you have any questions or a topic that you would like to learn more about? Please, feel free to reach out at support@usedesk.com – a unique email address, sending emails to which is very much appreciated.
What do you think about the article?