Nearly all medium and large companies spend hundreds of thousands, if not millions, on customer surveys every year. Customer survey results are used to amend strategies, design new products and services, and focus improvement activities. Gathering customer data is only the first step. The second step involves making best use of the data analyzing it, drawing business relevant conclusions and making important decisions. In this step, practitioners can benefit from Kano analysis, a model for classifying customer responses. The following case study illustrates its use.
A home appliances manufacturing company called on its customer care staff to analyze new satisfaction data and to suggest actions to the management team. All data was gathered using a four-point Likert scale based on whether customers were satisfied enough to buy the product again and a five-point Likert scale for how likely they would be to recommend the product to others (Figure 1).
Some conclusions were drawn immediately:
Calculating confidence intervals for all survey results proved the significance of all changes. The team decided to conduct more detailed analyses to find the culprit for the drop in delivery quality. Looking into the four major drivers for delivery quality revealed that cleanliness and punctuality left room for action (Figure 2).
But would improving cleanliness and punctuality really drive the ultimate goal of repetitive business and recommendations from existing customers? The data seemed to suggest it, but practitioners wanted to be sure about the impact.
Kano analysis is a tool often mentioned in Six Sigma training, but not so often applied in projects that can greatly help to structure customer needs based on feedback given. It divides customer needs into four categories:
Information about these categories of customer perception for a product and service are of enormous value for improving performance and gaining market share. Practitioners can use customer satisfaction data to establish a Kano analysis with help from the Jaccard index of similarity. Paul Jaccard developed an algorithm that enables regression-like comparison of non-continuous data such as that collected through a Likert Scale. Additionally, this algorithm can filter Musts, More The Better and Delighter characteristics out of the data.
In this case study, use of Kano analysis and the Jaccard index enabled the following conclusions:
Customer satisfaction data is not easy to come by. Therefore, it should be used to the fullest. Some basic mistakes can be avoided by appointing Belts with experience in discrete Likert data to complete the analysis. Additional value can be added with tools beyond the standard Six Sigma toolbox, such as the Jaccard index. Instead of relying on customer survey data providers for this analysis, Black Belts should be trained on additional methods to gain flexibility and save costs. Remember: Attaining the data is expensive, analyzing it can be cheap.