Identifying the Constructive Elements of “Value Proposition” and their Impact on Customers’ Satisfaction using Sentiment Analysis based on Text Mining

Because of the expansion of digital businesses and the customers’ involvement in this environment in recent years, it is possible to access the data related to the interests and values ​​expected by customers. A review of the previous local and international studies in the field of "value proposition" using "sentiment analysis" based on text mining showed that no independent research has been conducted in this field. Although the study of value proposition plays an important role, this concept is still poorly understood and there is a gap accordignly. Finally, it can be said that using the potential capabilities of analyzing the data produced by the customers on the Internet can be an important step in advancing the goals of businesses in the present competitive market. The purpose of this study is to properly use the large volume of online users’ feedback in order to discover, analyze value dimensions, and examine their impact on customer satisfaction. In this study, the product of "smart TV" has been selected as a case study in order to identify and investigate the features and characteristics of a product and its effect on customer satisfaction. Therefore, the main questions of this research are: 1) what are the components of value proposition? and 2) what is the role of each value element in the feeling of satisfaction or dissatisfaction of the buyers?  

This research is applied in terms of purpose, descriptive-survey in nature, and is conducted based on qualitative-quantitative (mixed) method using an inductive approach. For this purpose, the smart TV product was selected as the case study and 3865 comments by the buyers were collected from the Digikala online store website and were then analyzed using the sentiment analysis method. "Machine learning" method was used in order to rank and classify the positive and negative comments; the "Drikle hidden allocation" method was also used to extract the value constructs.  

Based on the research findings, 30 elements were identified for the value proposition in relation to smart TV products. The research results showed that 15 elements or product features had a positive effect on customer satisfaction and the other 15 elements lead to customer dissatisfaction. Positive constructs of value proposition included: positive feeling toward the product, screen size, product brand, reasonable price, excitement about the product, features and options, overall product quality, how to buy and when to receive the product, relative economic value, sound quality, overall product performance, other customrs' opinions, model and the appearance, installation services, and easy adjustments. Negative constructs of value proposition included: low economic value, low product efficiency, poor appearance, poor support for accessories, dissatisfaction with the price, low quality of the product, display problems, dissatisfaction with installation and operation, improper packaging, warranty problems, no guarantee for returning the goods, lack of accessories on the product, useless and additional options, improper remote control, as well as late and improper delivery of the product.  

The results of the study showed that the analysis of the customers’ feedback (i.e. "content produced by users") is a practical way to identify and examine customers’ attitudes towards the products. It is also an effective tool for the businesses to propose a successful product tailored at the consumers’ characteristics. The implementation of machine learning can lead to great steps toward facilitating, increasing the accuracy and speed in identifying customers’ needs, and creating shared value.