Purpose – This paper aims to categorize users of voice assistants and analyze decision-making conflicts to predict intention to adopt voicecommerce (v-commerce).Design/methodology/approach – This exploratory study used expert survey-based data collection founded on data saturation.Findings – This study identifies three forms of voice systems based on senses aroused (screen first, voice only and voice first) and four profiles ofvoice users (passive resistant, hedonistic adopter, utilitarian adopter and active resistant), each with a different appraisal of the benefits and costs ofv-commerce adoption and the experiences (positive or negative) felt during the shopping experience. This study proposes a conceptual model topredict intention to adopt v-commerce depending on voice-system and -user characteristics.Practical implications – Learning from this study can help improve the marketing strategies and actions put in place by voice-assistant brands andadvertisers by providing insights for adapting product recommendation algorithms to meet the needs of the identified profiles.Originality/value – This paper provides an answer to the limits of classical approaches based on “one-size-fits-all” strategy by showing how voiceassistant users have different profiles that span a gradient of advance in technology adoption