A Data-driven project categorization process for portfolio selection

Abstract

Purpose

Categorizing projects allows for better alignment of a portfolio with the organizational strategy and goals. An appropriate project categorization helps understand portfolio’s structure and enables proper project portfolio selection (PPS). In practice, project categorization is, however, conducted in intuitive approaches. Furthermore, little attention has been given to project categorization methods in the project management literature. The purpose of this paper is to provide researchers and practitioners with a data-driven project categorization process designed for PPS.

Design/methodology/approach

The suggested process was modeled considering the main characteristics of project categorization systems revealed from the literature. The clustering analysis is used as the core-computing technology, allowing for an empirically based categorization. This study also presents a real-world case study in the automotive industry to illustrate the proposed approach.

Findings

This study confirmed the potential of clustering analysis for a consistent project categorization. The most important attributes that influenced the project grouping have been identified including strategic and intrinsic features. The proposed approach helps increase the visibility of the portfolio’s structure and the comparability of its components.

Originality/value

There is a lack of research regarding project categorization methods, particularly for the purpose of PPS. A novel data-driven process is proposed to help mitigate the issues raised by prior researchers including the inconsistencies, ambiguities and multiple interpretations related to the taken-for-granted categories. The suggested approach is also expected to facilitate projects evaluation and prioritization within appropriate categories and contribute in PPS effectiveness.

Citation

El bok, G. and Berrado, A. (2022), "A Data-driven project categorization process for portfolio selection", Journal of Modelling in Management, Vol. 17 No. 2, pp. 764-787. https://doi.org/10.1108/JM2-10-2020-0257