Mapping Customer Relationship Management Data to Customer Experience using Machine Learning: In the Case of Ethio Telecom
No Thumbnail Available
Date
2022-01
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Making a positive customer experience is a strategic priority for a business company. To enable a seamless customer experience, the company used methods like surveys to rate the experience. Survey-based measurement systems of customer experience is expensive and time consuming. Another method the company uses to get customer experience insight is by utilizing new technology such as social media platforms and mobile applications. However, technology such as social media platforms failed to obtain managerial insights. A better approach to a seamless customer experience is to take full advantage of wealth data available in the customer relationship management system.
The main objective of this paper is to make use of classification algorithms widely used in machine learning research and then map CRM data to customer experience. To do this, 1208 individual records were collected from ethio telecom. The collected data is customer sentiment about their experience (a dependent variable) and independent data from the CRM system. And prepare the data set and use a classifier to map CRM data to customer experience and see the impact on classifier performance.
The comparisons will be made between Naive-Bayes, J48, and Random-Forest. And finally, J48 has a better accuracy performance of 82.03 %.
The result implies using CRM system data and mapping to customer experience is possible. However, the data type, data size, and data feature all have a significant impact on classifier performance.
Description
Keywords
Customer Experience, Naïve Bayes, J48