Determinants of cloud computing deployment in South African construction organisations using structural equation modelling and machine learning technique

Abstract

Purpose

This paper presents the findings from the assessment of the determinants of cloud computing (CC) deployment by construction organisations. Using the technology-organisation-environment (TOE) framework, the study strives to improve construction organisations' project delivery and digital transformation by adopting beneficial technologies like CC.

Design/methodology/approach

This study adopted a post-positivism philosophical stance using a deductive approach with a questionnaire administered to construction organisations in South Africa. The data gathered were analysed using descriptive and inferential statistics. Also, the fusion of structural equation modelling (SEM) and machine learning (ML) regression models helped to gain a robust understanding of the key determinants of using CC.

Findings

The study found that the use of CC by construction organisations in South Africa is still slow. SEM indicated that this slow usage is influenced by six technology and environmental factors, namely (1) cost-effectiveness, (2) availability, (3) compatibility, (4) client demand, (5) competitors' pressure and (6) trust in cloud service providers. ML models developed affirmed that these variables have high predictive power. However, sensitivity analysis revealed that the availability of CC and CC's ancillary technologies and the pressure from competitors are the most important predictors of CC usage in construction organisations.

Originality/value

The paper offers a theoretical backdrop for future works on CC in construction, particularly in developing countries where such a study has not been explored.

Citation

Aghimien, D., Aigbavboa, C.O., Chan, D.W.M. and Aghimien, E.I. (2024), "Determinants of cloud computing deployment in South African construction organisations using structural equation modelling and machine learning technique", Engineering, Construction and Architectural Management, Vol. 31 No. 3, pp. 1037-1060. https://doi.org/10.1108/ECAM-05-2022-0464