Optimizing The Customer Service Performance With Using Chatbot By Utilizing Bot Accuracy, Journey Completion Rate, And Customer Satisfaction Score (CSAT) (Case Study: PT YAI)

Abstract

In the world of Customer Service/Customer Experience with Chatbot technology, the use of metrics is required for industry to measure if the technology used is improved performance or regressive. In this thesis, we use 3 key metrics, Customer Satisfaction Score, Bot Accuracy, and Journey Completion Rate, to measure the proposal flows. Researchers have 3 flows to be measured which include Frequent Questions Asked (FAQs) for CSAT, Live Agent for Journey Completion Rate, and Ticketing Flow for Bot Accuracy. Each flow requires a different approach and method as solutions which encompass the Natural Language Processing (NLP) and categorization for FAQs, the use of memory and database for Live Agent, and the automatic fetching user personal information data to the ticket details and the use of memory in the database for Ticketing flow. As the result, we found that 4 commerce industries show progressive results which happen to achieve the minimum to maximum's percentage respectively as follows 71.81% and 96.74% for the CSAT, 83.48% and 98.98% for the Bot Accuracy, 79.91% and 97.71% for the Journey Completion Rate. Hence, these methods are concluded to be successful after having shown the improvement from the existing percentages and evaluated by the 3 metrics, CSAT, Bot Accuracy, and Journey Completion Rate.