The rise of Artificial Intelligence and Internet of Things has spurred voice assistant technologies like smart speakers, offering voice-output recommendations that affect consumers’ shopping decisions. This study bridges the research gap between voice-based and text-based recommendations through an experimental webpage mimicking real online shopping scenarios. While no notable differences were found in the number of unique wishlist items and final product selections between the two recommendation methods, voice-based recommendations significantly reduced shopping time. Clickstream analysis showed fewer list page visits and more direct product page explorations in the voice-based environment. Regression analysis confirmed the significant impact of voice-based recommendations on the number of visits to recommendation links and the recommendation conversion rate, providing insightful implications for businesses aiming to leverage voice-based recommendations effectively.