Most customer satisfaction measurement is conducted using a fairly standard 4 or 5 point scales from Very Satisfied, Satisfied, (Neither Satisfied nor Dissatisfied), Dissatisfied, Very Dissatisfied. Typically satisfaction is reported as the percentage of customers rating you as either Satisfied or Very Satisfied. Unfortunately, this tends to be quite a crude measurement. Most companies will be scoring around 75-85%. Even poorly performing local government can still easily reach 60%. The maximum we have seen is 92%.
The difficulty is that within such customer research studies is that year-on-year improvements are very hard to spot. The accuracy of the study often means that changes of 1 or 2% points are within the statistical tolerance of the design and show no real change, yet most companies would struggle to see changes beyond 1-2%. You then have the second question, which is if you are at 80%, how do you get to 82%? Some clever statistical techniques might show that improving delivery from 60% to 65% say, would improve satisfaction, but then how do you measure and improve delivery?
A second major factor here is that satisfaction questionnaires are really about measuring and controlling for dissatisfaction. If you have a satisfied customer then the only thing you need to understand is that they are happy. Many client satisfaction surveys fail customers because they ask happy customers lots of irrelevant questions. In a good questionnaire, we measure which standard the service reaches and whether the customer is happy with that standard. If they are not, we then ask for and analyze the reasons for dissatisfaction. Consequently happy customers have a shorter questionnaire and a better questionnaire experience.