Abstract
In business, customer satisfaction with a product or service is essential. It is especially effective in campaigns to analyze customer sentiment and satisfaction with the brand or measure customer service quality. Nowadays, users can efficiently perform transactions such as shopping, ordering food and drink online, and then leave feedback on the company's e-commerce websites. Businesses want to analyze customers' opinions and feelings to determine users' sentiment and views towards a specific product or service. This study proposes a customer satisfaction analysis method based on sentiment analysis and net promoter score (NPS). First, a dataset consisting of 48,471 online reviews in Vietnamese on websites in the online food ordering service sector was collected. Next, the pre-processed data is put into experimental machine learning models to evaluate and select the best model. Experimental results show that the proposed method has an accuracy of up to 90%. Finally, NPS is calculated based on customer rating. The result is visualized on dashboards with critical information dimensions.
TopIntroduction
Providing the best experience for customers through products and services provided by businesses is a crucial and critical task during each company's operation. This means that the products and services that companies offer satisfy and meet customers' needs: the right products and services, at the right place, and at the right time. The first step is listening to customers through the information conveyed in customer feedback on the business's sales channels and finding ways to improve and overcome the extraordinary things necessary steps of firms in the digital age. Today, with the strong development of technology and the intense application of interactive technologies on e-commerce sites, a large amount of data is collected from customer comments and feedback. The presence of technologies built on artificial intelligence and machine learning methods, and data analysis tools make it possible to analyze and extract meaningful data from textual data like comments or customer feedback is more accessible and more practical. This approach can be considered as the cutomer analytics method that is to gain customer insight necessary as well as customer satisfaction that are anticipated, timely, and relevant.
Besides, the digital transformation has helped businesses monitor their processes, including branding, promotion, advertising, production, channel distribution, based on collected data and interaction. Business managers analyze customer experience and can make more accurate and data-driven decisions (Marda, 2018; Ludbrook, 2019). In the transition and adaptation to the digital economy, it is necessary to have a new approach to analyzing user experience and emotions to predict and take advantage of disruptive technologies effectively. Advances in information technology have changed how communication makes it easier for customers to access information and exchange ideas about products and services on a large scale in real-time (Ghani et al., 2019). The advent of social networks and online review websites allows customers to give their opinions through reviews of products and services. From an e-commerce point of view, detecting the right user sentiment will help us display better advertising content. (Sarkar & Palit, 2020).
Customer analytics is critical for extracting insights from massive data in order to improve service innovation, product development, personalisation, and management decision-making (Hossain et al., 2022). Businesses utilize this data in particular for very direct marketing, location selection, and customer relationship management in a subtle way. There are many different definitions of customer satisfaction, and much debate is about this definition. Many researchers believe that satisfaction is between customer expectations and feeling practically received. According to Philip Kotler, customer satisfaction is the level of sensory status of a person derived from comparing the results obtained from the consumption of products or services with their expectations. The level of satisfaction depends on the difference between the results received and the expectation. If the actual results are lower than the expectation, the customer is not satisfied; if the actual results are commensurate with the expectation will be satisfied; if the actual results are higher than expectations, customers are delighted. Customer expectations are formed from the shopping experience, friends, colleagues, and the information of sellers and competitors. In order to improve customer satisfaction, businesses need additional investments and at least invest in more marketing programs.
This research aims to review the opinion mining research and propose a method exploiting customers' reviews in natural languages based on the machine learning method. This research applies the knowledge mining method from data collected by automatic programs, including reviews from customers on online ordering services, and eating places review channels. Then, data preprocessing will be conducted and machine learning methods will be applied to find the best model and predict sentiment scores for the rest of the corpus (Mishra & Tiruwa, 2017; Rao & Kakkar, 2017). In addition, the study is also going to calculate and analyze the Net Promoter Score (NPS) (Reichheld, 2003; Mandal, 2014) from customer rating scores and used data visualization tech NPS on an overview dashboard.
Key Terms in this Chapter
Net Promoter Score: A widely used market research metric that typically takes the form of a single survey question asking respondents to rate the likelihood that they would recommend a company, product, or a service to a friend or colleague.
Customer Analytics: Is a process by which data from customer behavior is used to help make key business decisions via market segmentation and predictive analytics.
Text Mining: Also referred to as text data mining, similar to text analytics, derives high-quality information from text.
Machine Learning: The study of computer algorithms that can improve automatically through experience and by the use of data. It is seen as a part of artificial intelligence.
Latent Dirichlet Allocation: A three-level hierarchical Bayesian model, in which each item of a collection is modeled as a finite mixture over an underlying set of topics.
Customer Satisfaction: A measurement that determines how happy customers are with a company's products, services, and capabilities.
Customer Behavior: The study of individuals, groups, or organizations and all the activities associated with the purchase, use and disposal of goods and services.
Sentiment Analysis: The use of natural language processing, text analysis, computational linguistics, and biometrics to systematically identify, extract, quantify, and study affective states and subjective information.