Identifying the industry-specific quantitative indicators for cloud migration strategy outcomes

This study’s object is the cloud migration process of information systems (ISs). This paper aims to resolve the task of devising a quantitatively grounded classification of migration strategies while previous approaches relied on conceptual models without empirical validation of industry‐specific performance metrics. Unlike existing categorizations, the proposed approach employs empirical data from 275 successful cloud migration cases, considering cost reduction, performance improvement, migration duration, as well as the number of cloud services used. Missing values are handled by multiple imputations via chained equations (MICE); outliers were removed using the interquartile range criterion, thereby enhancing result reliability. A taxonomy of three strategies – Lift-and-Shift, Re-platforming, and Reengineering – was established. Quantitative results indicate that Lift-and-Shift was applied in 39.64% of cases with an average cycle of 5.94 months and cost reduction of 40.06%; Re-platforming in 38.55% of cases with 6.10 months and 38.12% cost savings; Reengineering in 21.82% with 6.28 months, 42% cost savings, and 141.66% performance gain. Further analysis revealed an industry dependence in strategy selection: Lift-and-Shift predominated in regulated sectors, whereas Re-platforming and Reengineering were preferred in high tech industries. The findings could underpin automated decision support systems for planning cloud migration of IS at medium and large enterprises. The quantitative models enable forecasting of temporal and financial indicators based on system scale, technological landscape, and regulatory requirements. Implementation requires acquisition of performance and cost metrics and integration of MICE and outlier detection into pre-migration audits