Although the evolution of listening patterns resulting from the advent of streaming platforms calls for a finer and renewed understanding of the drivers of song popularity, few studies have investigated the relationship between song popularity and the typicality of their lyrics in relation to a given canon. Using the set of songs that reached the top five spots of the UK Official Singles Chart Top 100 between 1999 and 2013 as our canon, we show that the more typical a song’s lyrics are as regards to that canon, the more weeks it will spend in the top five spots of the chart. However, this typicality has no significant effect on the highest spot the song will reach, nor on the speed at which it will reach that spot, nor on its trajectory within the chart. In this sense, we replicate and extend the works of North et al. (2020). We reached these results by using an innovative methodology to measure the typicality of song lyrics on several dimensions we identified, using dynamic and time-proof measures stemming from natural language processing research.

Danilo C. Dantas is a full professor of marketing at HEC Montréal, where he is in charge of the graduate diploma in Management of Cultural Organizations. He is Vice-president of the Board of Jeunesses Musicales Canada and a member of the scientific committee of the OICRM (Interdisciplinary Observatory of Creation and Research in Music) of the Faculty of Music of the University of Montreal. His research interests include music marketing and digital marketing.