The rise of far-right parties has been extensively studied, yet less attention has been given to the specific activities these parties organize to expand their popularity and mobilize voters. This study examines the case of Golden Dawn in Greece and adopts a novel approach to understand how localized activities influenced electoral outcomes. Relying on computational methods, we extract activity-related articles from the party’s official website, parse the documents to classify activities by type (e.g., commemorative events, demonstrations), geolocate the activities, and map them to the municipalities where they occurred. These data are combined with municipal-level election results to assess the impact of different activity types on changes in Golden Dawn’s electoral vote share during four elections between 2012 and 2015. Socio-demographic factors, such as immigrant population percentages and unemployment rates, are incorporated to control for their influence on vote share. By focusing on nonviolent activities, this study addresses inconsistencies in the classification of far-right mobilization strategies and explores how specific localized engagement patterns contribute to electoral gains at the municipal level.