Our Thoughts

What is a CSI?

26th July 2017

A CSI (customer satisfaction index) is an overall measure of customer satisfaction that is a reliable as a survey measure can be since it uses the same criteria that the customers use to make their satisfaction judgement. Customers base their satisfaction with any organisation on the extent to which it meets their requirements. These requirements must therefore be identified initially through exploratory research, prior to designing a survey based on the customers’ most important requirements.

Having conducted the survey there are three methods of producing an overall index of customer satisfaction.

1) Overall question

2) Average satisfaction

3) Weighted index

1) Overall question

This simply involves asking respondents how satisfied they are overall, and scoring on a scale, which could be verbal or numerical. There are two problems with this approach. Firstly, the more variables you ask people to consider when responding to a question, the less reliable the answer is, and there are a lot of variables in an overall satisfaction question. Secondly, and more seriously, indices are far more reliable than individual questions due to the random error inherent in all survey measures. This can compromise the reliability of tracking in single answer questions but is eliminated in multiple question indices.

The other two approaches are both based on multiple question indices, but are require use of a numerical scale since mean scores must be used.

2) Average satisfaction

It is not necessary to ask an overall satisfaction question if the survey has asked each respondent about all the main things that make them satisfied or dissatisfied. The  second approach is therefore to calculate the overall average of all the satisfaction scores. That would be better than an overall question, but it is not ideal, because some things are more important to customers than others, and their most important requirements influence their satisfaction judgement more than things that are less important to them.

3) Weighted index

An accurate satisfaction index therefore has to work in the same way as the sub-conscious judgement made by the customer. It has to be more strongly influenced by the attributes with the highest importance. In other words it must be a weighted average satisfaction score, which requires a two step process for its calculation. The importance scores are first used to calculate the weighting factors, by totalling all the importance scores then expressing each one as a percentage of the total. The second step is to multiply each satisfaction score by its corresponding weighting factor. The overall weighted average is determined by adding up all the weighted scores. It is normal to convert that score into a percentage.

Updating the CSI

The satisfaction index has to updateable. It has to provide you with a comparable measure of satisfaction that can be monitored in the years ahead even if the questions on the questionnaire have to change as customers’ requirements change. Basically, the CSI answers this question:

“How successful is an organisation at satisfying its customers according to the 20 things that are most important to them?” (Assuming 20 attributes on the questionnaire.)

If the questionnaire has to change in the future because customers’ priorities have changed, the satisfaction index remains a measure of exactly the same thing.

“How successful is the organisation at satisfying its customers according to the 20 things that are most important to them?”

That comparability also applies to organisations with different customer groups who need to be asked different questions as their requirements are not the same. Provided that exploratory research has been undertaken, the CSIs from surveys of two or more customer groups, asking different questions, are directly comparable.

Click to download this article in PDF format

01484 517575
Taylor Hill Mill, Huddersfield HD4 6JA
Twitter LinkedIn