THE IMPACT OF DATA-DRIVEN WEB FRAMEWORKS ON PERFORMANCE AND SCALABILITY OF U.S. ENTERPRISE APPLICATIONS

Abstract

This systematic review examines how data-driven web frameworks shape performance and scalability outcomes in U.S. enterprise applications by integrating evidence across rendering strategies, data-access paradigms, runtime and placement topologies, and delivery protocols. Using a PRISMA-guided methodology and a registered protocol, we searched academic databases and high-rigor practitioner venues for studies published between 2014 and 2024, applied dual screening with adjudication, and retained 115 studies that reported transparent metrics under realistic workloads. Synthesized findings show consistent user-perceived gains when organizations adopt server-first rendering with streaming and disciplined hydration, especially when paired with cache-aware API contracts and explicit delivery priorities. For composite, authenticated views, GraphQL yields tail-latency and payload benefits when guarded by persisted queries, batching, and cost control, while REST remains superior for flat, cacheable reads that amplify edge hit ratios. Proximity emerges as a first-order lever: edge execution and multi-region placement reduce p95 and p99 only when data gravity follows compute, and targeted warm-path strategies are required for serverless estates to avoid cold-start penalties. Across the corpus, the most cost-effective improvements arise from cache-key normalization, deterministic revalidation, and precise prioritization that make early bytes visible to the network scheduler. We advocate a decision map that sequences investments from cache semantics and priorities to server-first plus streaming, to workload-appropriate API design, to selective edge and serverless adoption, all measured with percentile-aware telemetry and distributed tracing. This alignment consistently delivers low double-digit reductions in LCP and high-percentile latencies while protecting error budgets in enterprise conditions.