An Empirical Comparison of Selenium and Puppeteer for End-to-End Web Testing: The SauceDemo Case Study

Abstract

This study presents an empirical comparison between two leading web automation frameworks, Selenium and Puppeteer, based on end-to-end testing of the SauceDemo web application. The research aims to measure and evaluate both frameworks in terms of performance, including execution time, CPU utilization, and memory  and effectiveness, which covers code complexity, ease of implementation, and maintainability. Identical functional test scenarios such as login validation, product filtering, cart management, and checkout completion were developed in Selenium (Python) and Puppeteer (Node.js) and executed under the same environmental conditions. The quantitative measurements revealed Puppeteer's average execution time of 4.7 seconds, which was approximately 45% faster than Selenium's 8.4 seconds. In addition, Puppeteer consumed 35% less CPU and 34% less memory on average, demonstrating higher efficiency in resource utilization. Qualitative evaluation revealed that Puppeteer required fewer lines of code and a simpler setup process, resulting in better maintainability, whereas Selenium remained superior in terms of cross-browser compatibility and integration flexibility. These empirical findings confirm that architectural design significantly influences automation testing performance. The results provide data-driven insight for quality assurance professionals and developers in selecting the appropriate framework according to testing requirements, scalability, and environmental constraints.