Determining customer satisfaction from mobile phones: a neural network approach

Over the last few years, the number of mobile phones in the world has increased at an exponential rate with many developed countries reaching 60% ownership rates. The reasons for this are numerous, however low prices and the availability of new technology mean that even children now own and regularly use mobile phones. Furthermore, the number of mobile phones in the world has already passed the number of fixed land lines and the revenue from mobiles phones will soon exceed that of fixed land lines (ITU 2001). This paper explores the relationships between a number of key input factors and customers' overall satisfaction with their mobile phone, and develops a neural network model to predict the overall level of customer satisfaction derived from mobile phones in the UK. The final model uses eleven input factors, the most important of which are experience of product quality, level of service charges, level of call charges, and level of satisfaction with the service provider. The estimated neural network model predicted very well and appears to be very robust. Finally, the implications of these results for marketers are discussed