The role of artificial intelligence personalisation in e-commerce: Customer purchase decisions in the retail sector

Abstract

Background: Online shopping in South Africa grew significantly during the pandemic, highlighting its potential across income groups. Leveraging artificial intelligence (AI) is crucial for e-commerce businesses to enhance customer experiences and stay competitive in the digital age.

Objectives: This study aimed to explore how AI-driven personalisation in e-commerce influences customer purchase decisions in the retail sector in South Africa.

Method: A quantitative approach within a post-positivist paradigm assessed how factors such as perceived usefulness (PU), perceived ease of use (PEU), relative advantage (RA), and voluntariness of use (VOU) impact purchase intention (PI), repeat purchase intention (RPI), and loyalty (LO). Data were collected through a 5-point Likert scale survey distributed to 175 South African e-commerce consumers using convenience sampling, with structural equation modelling (SEM) used for analysis.

Results: The study showed mixed results. Perceived usefulness had a negative but insignificant impact on purchase decisions, possibly because of unmet expectations or price sensitivity. Perceived ease of use also showed a negative but insignificant effect, suggesting that ease alone is insufficient if personalisation quality is lacking. Conversely, relative advantage and voluntariness of use positively influenced purchase decisions, with customers valuing time savings and control over their data.

Conclusion: This study highlights the importance of AI-driven personalisation in South African e-commerce, emphasising the balance between functional and emotional factors to build customer loyalty and influence purchase decisions in a growing market.

Contribution: The study offers practical implications for African businesses and suggests further research into generative AI and cross-country comparisons.