Robbie Blom

October 15, 2024

6 minute read

Customer Satisfaction: An Example Of Measuring What You Can't See

Customer satisfaction, customer loyalty, brand perception, employee engagement, perceived quality... some things that are central to the success of a business are also difficult to measure.

Since these things involve human beliefs, feelings, and perceptions, we often take a human approach to understanding them by immersing ourselves in the situations where they occur.

We need this qualitative approach, no doubt. But at scale it can be time-consuming, expensive, and difficult to track over time.

A quantitative approach can help us understand these fuzzy but important concepts in a more systematic way. But how do we identify these things? And how do we put a number on them?

Turns out, we can answer both these questions with a clever combination of survey design and statistical analysis. In this article, we'll see how it works by looking at a specific example for measuring customer satisfaction.

Simple Pulse Checks For Customer Satisfaction

Common customer experience metrics like Net Promoter Score (NPS), Customer Satisfaction Score (CSAT), and Customer Effort Score (CES) suggest a rather straightforward way to put a number on something like customer satisfaction.

They're super simple: you just ask customers to rate their experience and then tally up the responses.

Net Promoter Score (NPS)

NPS relates to how likely a customer is to recommend your product or service to others. It's really a measure of loyalty, but it can serve as a pulse check for overall satisfaction.

To calculate NPS, you simply subtract the percentage of detractors (those who rate 0-6) from the percentage of promoters (those who rate 9-10). The result is a number between -100 and 100. The higher the number, the better.

Customer Satisfaction Score (CSAT)

CSAT is a more direct way of measuring satisfaction: it simply asks customers how satisfied they are.

To calculate CSAT, you just divide the number of "Satisfied" and "Very Satisfied" customers by the total number of respondents. The result is a percentage between 0 and 100.

Customer Effort Score (CES)

CES asks customers how easy it was to do business with you. Like NPS, it's not an exact measure of satisfaction, but it's a decent proxy.

To calculate CES, you divide the number of customers who found the experience "Easy" or "Very Easy" by the total number of respondents. You can also assign each response as a number between 1 and 5 and then calculate the average.

The Bottom Line With These Metrics

These metrics are simple, easy to understand, and easy to calculate. However, they are incomplete measures of customer satisfaction.

What's missing is that customer satisfaction is a multifaceted thing that's difficult to capture in a single question. For example:

• Customers may be satisfied with one aspect of your product or service but dissatisfied with another.
• They may be satisfied with the product they bought but dissatisfied with the service they received.
• Or may be satisfied with the purchase experience but dissatisfied with the price.

To get a more complete picture of customer satisfaction, you need to ask more questions.

A Wholistic Customer Satisfaction Index

One obvious way to get a more complete picture of customer satisfaction is to combine the scores from NPS, CSAT, and CES into a single index. Given that CSAT is a more direct measure, you might allocate 70% weight to CSAT and 15% each to NPS and CES. That way if CSAT is low but the others are high, then the index will give you a more balanced view.

How do you know that the 70-15-15 split is the right one? You don't - they're just guesses.

But this approach is exactly the one we need to get a more complete picture:

1. We carefully design questions that capture different aspects of customer satisfaction. These questions may measure satisfaction with varying degrees of accuracy.
2. We ask customers to answer these questions in the form of a rating.
3. We combine the responses in a way that optimally weights each response based on its contribution to satisfaction that we see across respondents.

We can see this meaurement strategy visually in the diagram below.

Customer satisfaction is a thing that we can't measure directly like we can for things like height, weight, or revenue. But we can measure the responses to our questions, which are related to satisfaction. We can then back out and combine the magnitude of those relationships to form an index.

What questions we ask is a topic of its own. For customer satisfaction, we would probably want to follow an Expectation Confirmation approach.

The way we can combine responses is through a statistical technique called factor analysis. Factor analysis is a way to identify underlying relationships among variables. In this case, the underlying relationship we'll find is customer satisfaction.

In factor analysis, that relationship between the survey responses and customer satisfaction is expressed as a single score per respondent.

The idea is that satisfied and dissatisfied customers will answer the survey in different ways. Certain questions are better at discriminating between the two, and thus have higher weights when calculating the index.

Estimating the weights from the data like this means we're not guessing, and thus it gives us a more accurate and wholistic measure of customer satisfaction.

Why It's Important

Tracking customer satisfaction with simple pulse checks is definitely a good start, and it may be all that you need.

But if you differentiate yourself by creating superior customer experiences or you want to be more intentional about customer considerations in frameworks like Balanced Scorecard, then you'll probably want to go deeper.

Fortunately, there's a well trodden path to do just that.

Conclusion

Some things that are central to the success of a business are a little fuzzy and difficult to measure. Because these things often involve beliefs, feelings, and perceptions, we often take a more qualitative, human approach to understanding them. It turns out, though, that there's a quantitative approach that can be just as useful.

In this article, we saw how that quantitative approach works through an example of measuring customer satisfaction. Not just limited to customer satisfaction, this approach can be used to measure other important constructs such as customer loyalty, brand perception, employee engagement, or perceived quality.

Together with our own observations and experiences, this quantitative approach can be a valuable tool to give us a more complete picture of the things that are important in our business.