Who Do Consumers Trust? : The Role of Algorithmic vs. Human Recommendations and ProductType in E-commerce Consumer Behavior

Sammanfattning

The rapid digitalisation of the retail sector has transformed how consumers interact withproducts and information through AI-driven recommender systems. While algorithmic recommendations offer efficiency and personalization, consumer trust in such systems mayvary depending on the nature of the product and the perceived recommendation source. Thisstudy examines how recommendation source (algorithmic vs. human) influences consumertrust and purchase intention across utilitarian and hedonic product categories in e-commerce environments. The study applies a quantitative experimental survey design using structured questionnairesdistributed through online platforms. Data were collected from online consumers andanalysed using statistical methods including reliability analysis, regression analysis,moderation analysis, and PROCESS Model 7. The findings indicate that recommendation source significantly influences consumer trust. Human recommendations consistently generated higher levels of trust than algorithmic recommendations across both utilitarian and hedonic product conditions. In addition, consumer trust was found to positively influence purchase intention. Although the indirecteffect appeared descriptively stronger for utilitarian products, the moderating effect ofproduct type was not statistically significant within the overall model. The study contributes to existing research on algorithm appreciation and algorithm aversionby demonstrating that consumer trust in recommendation systems is context-dependent. Practical implications suggest that e-commerce platforms should align recommendation strategies with product characteristics in order to strengthen consumer trust and enhancepurchase intentions.