Purpose Several barriers have a catastrophic effect on the adoption of digitization in supply chain (SC). The purpose of this paper is to develop a structural model of the barriers and assess the influence of the barriers on the overall adoption of digitization. Design/methodology/approach The ten most significant barriers to adoption of digitization in SC are categorized as – social, economic and technological. An integrated – partial least squares structure equation modeling (PLS-SEM) technique is used for hypothesis testing and machine learning (ML)-based Bayesian network (BN) is used for predicting the influence of the barriers on the adoption of digitization in SC. Furthermore, an integrated decision-prediction framework for digitization success and sustainability impact prediction is proposed. Findings The model developed using PLS-SEM is found reliable resulting in significant impact of economic as well as technological barriers on the social barriers. Technological barriers are also found to impact the economic barriers. The BN results reveal that the following barriers have higher influence on the adoption of digitization – the difficulty in integration of cyber-physical systems from varied platforms (B8), large investment in digital infrastructure (B6) and management of existing data on common platform (B9). Research limitations/implications Adoption of digitization in SC is getting significant momentum across the geographies, but the rate is very low, creating a need for categorical investigation of factors with their underlying relation and significance. This study bridges a prominent gap through quantitative statistical analysis of the barriers influencing the adoption, which will help the organizations in making data-based decisions during digitization endeavors of their SCs. Practical implications This will help the practitioners to benchmark their readiness toward adoption of digitization in SCs and develop the strategies keeping SDGs in mind. Originality/value This research uses a unique two-phased methodology of – PLS-SEM and ML-based BN for hypothesis testing and predictive modeling, respectively, resulting in a novel approach. The proposed decision – prediction framework for digitization success and sustainability impact prediction is unique.
Abhyankar et al. (Thu,) studied this question.