Investigating the role of utilitarian and hedonic goals in characterizing customer loyalty in E-marketplaces

Abstract

Despite significant growth in sales in recent years, retaining customers remains a major challenge for the electronic marketplace (e-marketplace) industry worldwide, including Indonesia. The small basket size of Indonesian customers has created a highly price-sensitive market, making it difficult to nurture customer loyalty. This study investigates the factors affecting consumer behavior and loyalty in Indonesian e-marketplaces by employing a newly proposed E-Marketplace Customer Loyalty Model. Since customers' expectations have extended from mostly utilitarian-oriented goals to incorporate hedonic goals, the proposed model includes both utilitarian and hedonic-based constructs by integrating the Expectation-Confirmation Theory, e-service quality, and the Hedonic Information Systems Model. The proposed model was tested using a comparative approach of machine learning classification algorithms and partial least square, aiming to create a more robust model. The PLS analytical results of ECLM hiearchical component model from 678 customers show that the fulfillment of hedonic values via perceived enjoyment has a greater impact on customer satisfaction than utilitarian values of perceived service quality. Perceived enjoyment, personalization, and customer satisfaction positively affect customer loyalty. The classification results provide further evidence for all hypothesized relationships of the ECLM. The Sequential Minimal Optimization (SMO) algorithm has demonstrated superior performance compared with other classifiers in predicting the dependent variable in most cases. Based on findings, this study offers theoretical and practical implications, and recommendations for sustainable loyalty programs in e-marketplaces.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 41 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.