HECON: Weight assessment of the product loyalty criteria considering the customer decision's halo effect using the convolutional neural networks

Abstract

The economic pressures and increasing competition in markets have led to the CEOs of companies being forced to make the right strategic decisions in the development of products for selling the right products to the right customers. To achieve this goal, companies need to know which criteria of their products lead to customer loyalty to that product. In the past, various methods have been introduced to obtain the importance (weight) of criteria that use the opinions of experts or customers. There is a halo effect in human decisions that leads to biases in evaluating the criteria by influencing human emotions. This study introduces a new method for weight assessment of the product loyalty criteria by considering the customer's decisions halo effect using the convolutional neural network (CNN) called the halo effect using the convolutional neural networks (HECON) method. In the HECON method to consider the halo effect of the customer decisions, a CNN model is proposed as the deep learning pipeline to obtain more accurate weights of the criteria. The HECON method to obtain the weight of the criteria and identify criteria that lead to product loyalty needs to collect the feedback of a large number of customers based on the net promoter score (NPS) scale. The innovation of the HECON method is to obtain the effect level of each product criterion on selection and loyalty to the product through the feedback of a large number of customers by considering the halo effect on the customers' thinking. To date, the analyzing methods have often not been able to identify the halo effect in evaluating the reasons for customer loyalty to the product. The halo effect indicates sometimes some of the product criteria secretly affect the customers' opinions that require deep neural networks to analyze them. By using the deep CNN model of the HECON method to evaluate product criteria for understanding customer behavior, companies will be able to identify customers' behavior and develop their products exactly following the customer's desires. To evaluate the performance and demonstrate the applicability of the HECON method, presented two case studies. It is presented that there are challenging differences between the results of the HECON method with the other methods because the HECON method considers the halo effect on the customer decisions and demonstrates better performance.