Finding determinants of e-commerce success: a PLS-SEM approach

Purpose - With the growth in internet technology, the competition between the e-commerce firms has become fierce. This has forced these online vendors to consider customer-centric strategies apart from the retail strategies for generating revenue. The purpose of this paper is to propose a structural model to study the interrelationship between a formative endogenous latent variable (LV), i.e. e-commerce success, and other LVs representing the customer-oriented success factors such as website service quality (WSQ), customer support system (CSS), personalization and electronic word-of-mouth (EWOM). Design/methodology/approach - The authors develop a structural model to validate the determinants of electronic commerce success. Survey-based data obtained from 265 respondents are used to validate the model hypotheses. To analyze the data, partial least squares-structural equation modeling (PLS-SEM) is performed using smartPLS 2.0 software. The results on the interrelationships between the constructs are determined using path analysis. Findings - The data analysis results infer the impact of these success factors on e-commerce success systems. After applying bootstrapping technique, the authors observed that WSQ, CSS, personalization and EWOM have a significant positive effect on e-commerce system success (ESS). Further studies can be performed considering mobile commerce and their critical success factors. Practical implications - The findings of the study emphasize on the importance of these four exogenous variables for e-businesses in order to achieve a competitive edge over their competitors. The study provides a detailed managerial implication with respect to each causal relationship, so as to suggest some remedial measures to gain customer market. Originality/value - Here, the authors consider an empirical model which shows the determinants of ESS. To the best of the authors’ knowledge, a distinctive system success framework is modeled considering a unique combination of exogenous variables not considered by previous studies.