A Methodology for Consolidating Decades-Old Enterprise Software Portfolios into a Unified Web Platform: Discovery, Data Model Unification, Architecture, and Migration Approach

Abstract

Large enterprises accumulate operational software over decades, ending with portfolios that overlap in capability, contradict in data definitions, and impose substantial cognitive load on users who must traverse them. This paper presents a methodology for consolidating such portfolios into a single unified web-based platform, synthesizing patterns observed across enterprise consolidation programs and validated through a reference deployment that retired more than forty legacy systems serving portfolio, project, and asset management workflows. The methodology comprises three components: a discovery framework that produces a system catalog, a capability map, and a data inventory; a data-model unification approach that resolves cross-system contradictions through user-authored canonical definitions; and a migration discipline that combines phased cutover, coexistence, and reconciliation. The reference deployment under study used containerized microservices distributed across two public clouds, an API-first surface, and a medallion data platform to support continued evolution after consolidation. We report observed values from the reference deployment that show a reduction from forty-three to one production system, a 71 percent decrease in monthly license and support cost, a defect density reduced from 4.8 to 1.2 defects per thousand lines of code, mean time to recovery improved from 6.4 to 1.8 hours, and a 58 percent reduction in user training hours. The findings inform research and practice on portfolio rationalization in long-lived enterprise environments.