In conducting an exploratory data analysis to study the hit mechanisms of manga and anime works with appealing supporting characters, the author began by assessing the volume and content of Twitter posts on Demon Slayer and observed that SNS not only facilitated communication between devoted fans, but also served to fulfill the desire to own merchandise and to provide a place for publication of derivative works. Furthermore, in an analysis using the Character Quantitative Survey unaided recall data, the top unaided recall works were classified into four types based on supporting character popularity patterns. The results obtained by creating a composite variable of unaided recall weighted by simple total based on the number of recall of the work's name or character and assessing the mean values and correlations of each variable measured for each work showed that the composite variable of unaided recall had a strong correlation with the indices of tie-in merchandise purchase and purchase intention, and with the indices of ‘high spirit and fun’ and ‘attention and impersonation’ in the field of offered experience.