Analysing the effectiveness of chatbots as recommendation systems in fashion online retail: A Brazil and United Kingdom cross-cultural comparison

Chatbots as recommendation systems have become promising for online retail, enhancing the customer experience by offering a virtual shopping assistant with product suggestions. Fashion brands increasingly adopt this technology to offer personalised shopping recommendations, reduce cart abandonment, returns and complaints. This study proposes an enhanced Recommendation System that combines a simple Graphical User Interface (GUI) with the DigAI chatbot, making the customer experience more interactive and presenting more relevant clothes to the customer. The design and evaluation of the chatbot interface followed a three-phase methodology: a systematic literature review, building DigAI, and a cross-cultural fashion consumer study to evaluate the chatbot’s performance. In the study, fashion customers were to select an item from an e-commerce site both with and without the chatbot and then complete a survey questionnaire. The experiments took place in Brazil and England to identify possible cultural differences. Albeit these differences, the chatbot interface was found to provide a more satisfying user experience than the simple GUI. This effect was more pronounced among Brazilian individuals compared to those from the UK, suggesting that cultural or contextual factors may have influenced their perception of the interface improvements, which can eventually be exploited to guide fashion online retail strategies.