Design and Implementation of an ETL Pipeline for Prospective Student Data Analysis in Higher Education Admissions

Abstract

The number of universities in Indonesia continues to grow. This condition certainly makes the flow of new student admissions increasingly competitive between universities, thus encouraging universities to do branding, show quality, and do the right positioning. Therefore, it is important for universities to adopt a data-driven approach that can provide in-depth insights into prospective students and the effectiveness of marketing strategies. The purpose of this study is to design and build an ETL (Extract, Transform, Load) pipeline to collect, process, and analyze prospective student data as part of the business intelligence (BI) system to be built. The proposed ETL architecture design supports automated microservices-based data transformation in data cleaning, normalization, and integration. In addition, it can also be used as a solution to increase the scalability and flexibility of data mobilization in the BI system. This study introduces a novel approach by designing an ETL pipeline within a business intelligence framework aimed at enhancing university marketing efforts. Unlike prior research, which has primarily applied business intelligence tools to evaluate academic activities within learning management systems, this work shifts the focus to marketing analytics. Additionally, while existing studies on higher education marketing often center around digital marketing techniques and the marketing mix, this research fills a gap by proposing a technical infrastructure that supports data-driven marketing through automated ETL processes. The resulting ETL was tested using several methods, namely Source to Target Count Testing, Source to Target Data Testing, Duplicate Data Check Testing, and Data Transformation Testing. The results of each test are valid