APPLYING AN INTEGRATED MULTI-CRITERIA AHP–TOPSIS METHODOLOGY FOR PROJECT SELECTION IN IT OUTSOURCING COMPANIES

Abstract

This paper presents an integrated AHP–TOPSIS method for project selection in IT outsourcing, where each engagement both generates revenue and shapes capabilities. We operationalize five pillars—financial attractiveness, resource feasibility, strategic alignment, risk profile, and competence synergy—on a common scale, combining expert judgments with quantitative data. Financials include expected NPV and target margin adjusted for tail risk via CVaR; resources reflect total effort and the load on critical roles; strategic alignment covers roadmap support and access to new markets; risk spans technical, contractual, and regulatory exposures; competence synergy quantifies reinforcement or extension of core skills. Weights are derived via AHP with consistency checks and then applied in TOPSIS to compute closeness to positive and negative ideal solutions. A realistic simulated portfolio (USD 100–300k contracts; market billing rates) evaluates four projects on ten subcriteria. Results prioritize options with moderate NPV, low CVaR, and strong competence synergy, while resource-intensive projects with weaker strategic fit are penalized despite higher gross returns. Sensitivity tests show that top ranks remain stable under plausible weight variations, indicating robustness for group decision making. Managerially, the framework speeds evidence-based portfolio decisions, reduces ad-hoc bias, limits portfolio drift, and aligns intake with long-term capability building. A decideXpert implementation enables parallel expert input and automatic consistency feedback, improving reproducibility and governance. The contribution is a practical adaptation of established MCDM techniques to the service context with explicit treatment of tail risk and competence effects. Limitations concern the scale of the illustration and static inputs; future research should extend datasets across firms and add fuzzy and learning-based weight updates to capture uncertainty and dynamics. We also detail implementation parameters—normalization choice, benefit/cost handling, aggregation rules, and reporting templates—to facilitate replication in corporate settings and auditing of decisions.