Engineering Speed at Scale: A Comprehensive Study of Frontend Performance Optimization Using the PERFORM Framework

Abstract

Frontend performance has become a critical factor influencing user experience, engagement, and business outcomes in modern web applications. As frontend architectures grow increasingly complex, performance bottlenecks related to resource loading, rendering behavior, and interaction latency become more prevalent. This paper presents a comprehensive study of frontend performance optimization conducted during a summer web development internship. A systematic optimization methodology, the PERFORM framework (Profile, Evaluate, Recommend, Fix, Observe, and Report), is used to analyze and improve performance across multiple production-grade web applications. The study focuses on optimizing the critical rendering path, improving Core Web Vitals, and applying practical optimization techniques, such as image optimization, lazy loading, and reducing render-blocking resources. Experimental results demonstrate significant improvements in page load performance, responsiveness, and visual stability. The findings highlight the effectiveness of structured performance engineering practices in real-world frontend development scenarios.