Decision-making is a process inherent to everyday life and momentous situations, which is often complicated, especially in group contexts, due to the diversity of opinions and constraints in the available information. A novel multi-granular group decision-making method based on individual similarity is presented in this context. This method is also a consensus model that secures flexibility by allowing individuals to provide information in numerical format and based on their perspective, avoiding the need to fill in the reciprocal preference relationships with information they do not know. Furthermore, the system assigns weights to individuals based on the quality and quantity of their contributions, recognising and rewarding those who provide relevant and valuable information. Ultimately, this approach promotes a more reliable and accurate decision-making process, tailored to the knowledge and experience of the group of individuals involved, thus improving the quality of decisions made as a team.