Customer satisfaction – dilemma of comparing multiple scale scores

Abstract

Most businesses use customer satisfaction scores. It is difficult to compare scores obtained on one scale with another. For instance, if an organisation is rated 4 on a 5-point scale and another organisation is rated 3 on a 4-point scale, should they be considered equal or different? If the approximation is applied, e.g. considering a score of 8 on a 10-point scale as 80% on a percentage scale, it can be incorrect. This study suggests a mechanism by which customer satisfaction responses from one scale can be converted to an estimated response on another scale. Ten satisfaction scales of varying length are compared by collecting responses from over 400 participants. Data are analysed using correlation and polynomial regression methods. Methodology and matrix to convert scores are prescribed. Merits and demerits of different scales are presented. It is observed that the 10-point ordinal scale is more suitable for the measurement of satisfaction and for inter-scale conversion. A percentage scale is also a good option for this purpose. Scales of less than 7-point length are not proved good from a score conversion perspective. The study will help individuals, organisations, and business analytics companies analyse customer satisfaction data measured on multiple scales.