Case Study: Customer Satisfaction Survey
Case Study: Customer satisfaction surveySatisfaction surveys are an important measurement tool in fields such as market research or human resources management. Serious studies consist of numerous questions and contain answers from large population samples. Aggregation on both sides, the questions asked as well as the answers received, turns the multidimensional problem into a complex system of interleaved hierarchies. Traditional ways of presenting the results are limited to one-dimensional charts and cross-tables.
In this case study, a public transport network performs yearly customer satisfaction surveys. The satisfaction is measured on a scale of 1 to 10. Here is a typical question:
The questionnaire consists of 80 questions. The answers to the 80 questions are then topically grouped and aggregated into 23 quality dimensions. The quality dimensions are further aggregated into 3 satisfaction indices.
Each year between 2000 and 5000 questionnaires are collected. The responses are clustered into 72 analysis groups such as
- demographic (e.g. age)
- organizational (e.g. service providers)
- geographical (e.g. urban/agglomeration)
Traditionally, the results of the survey were presented on paper or with PowerPoint. The presentations consisted of static charts and tables. The diagrams were predefined and did not allow for ad-hoc questions.
The results of the survey form a balanced tree of uniform depth. The big challenge is that the node attributes are high-dimensional, namely the 72 analysis groups (for each question there is one result per analysis group). Standard tree representations however don‘t work well to show more than one or two node attributes at a time. To solve this problem, a new technique was developed, called a Parallel Coordinate Tree. The following animation shows how this representation is constructed:
The Parallel Coordinate Treee combines the familliar tree layout with multi-dimensional analysis. This representation was embedded in an interactive visualization application that provided easy access to all the results of the studies. It received enthusiastic feedback from the statisticians that design and analyze the studies. For normal business users however, who do not have the same analytic focus, while the visualization is attractive, it does not add that much value to their daily business. The lesson learned is that it is critical to provide the right tools to the right users.
This case study is further elaborated in the exercise that follows in this section. There, you will be able to read a longer paper about this tool, but more importantly to actually dowload and try it out.
Last modified: Monday, 5 February 2007, 1:44 AM