ABSTRACT
The rapid development of information technology brought major changes to business activities. Traditionally, the transaction process has been done directly. As a result of the internetpresence, online platforms known as e-commerce began to facilitate the transaction. Furthermore, the online platform also facilitates customers in expressing their opinions and complaints, one of which is through online chats with customer service. The emergence of e-commerce shifts the business focus from only goods and services to customer information and an intelligence orientation. E-commerce companies are forced to understand customer perception including identifying customer satisfaction and disappointment issues related to their service. Customer chat data contain the information needed to understand this customer perception. In this study, we analyzed the e-commerce customer experience through customer chat data. We propose two methods to mine the text: sentiment analysis and topic modeling. Sentiment analysis aims to identify the customer satisfaction level, while topic modeling aims to extract the important customer issues within each sentiment class. We use Paperlust.Co, the invitation and card designs e-commerce company based in Australia, as a case study. This research provides an evaluation and valuable information related to customer experience in the e-commerce industry. We discovered that Paperlust.Co customers feel satisfied with the company service. Operational problems such as site failure were the major issues in the customer opinions and complaints. This insight might help e-commerce companies to deliver more value that fits well with their customers’ needs.
