Why Do Some Become Popular?
Why do some cultural products become wildly popular while others fall short? Researchers Jonah Berger and Grant Packard investigate in a paper published in the journal Psychological Science, focusing on popular music. Based on the data examined, they propose that atypical tracks are more likely to become a hit.
How many people had predicted that “Barbie Girl” by Aqua, “Macarena” by Los del Rio, “Wannabe” by Spice Girls, “Smells Like Teen Spirit” by Nirvana, or “Gangnam Style” by Psy would become so popular?
It is nearly impossible to predict future popularity; even experts have difficulty predicting which songs will become hits. There are likely multiple reasons why these songs were so successful (e.g., marketing strategies, viral music videos).
We know that human beings desire stimulation, so they are more likely to be drawn to things that are novel and different. Could one reason for these songs' success be that they were unusual? That is what the authors of the present study decided to find out. The authors performed “bounded” analyses; that is, they evaluated to what extent a song's lyrics were different from the average song in that particular genre.
Using Billboard’s digital download rankings, the researchers sampled ranking data from 2014 to 2016 for seven genres of music: Pop, rock, country, rap, dance, rhythm and blues, and Christian music. The resultant data set included 4,200 song rankings (with 1,879 unique songs).
Berger and Packard then measured the co-occurrence of words in the songs’ lyrics; based on this information, they determined the theme of each song. Ten potential themes emerged: “anger and violence,” “body movement,” “dance moves,” “family,” “girls and cars,” “uncertain love,” “fiery love,” “positivity,” “spiritual,” and “street cred.”
For instance, the words “car,” “girl,” “drive,” and “road” were high-probability words for the topic of “girls and cars.”
The authors also determined each genre’s “lyrical topic composition.” For instance, the genre of country featured more lyrics related to “girls and cars,” followed by “uncertain love,” “family,” etc; while rap music featured more lyrics about “street cred,” followed by “anger and violence,” and so on.
By comparing the lyrical topic composition of each song to the genre’s average composition, the researchers determined which songs were more atypical.
The results of the study indicated that, in general, the more atypical songs (ones that featured less genre-typical content) were indeed ranked more highly. Further analyses also showed that factors related to radio airplay, the artist, language complexity, number of words, linguistic style, etc., did not account for these main results. Nor was the popularity of the high-ranking songs due to the song’s similarity to ones in other genres; that is, the popularity of successful songs was not due to their broader appeal.
Berger and Packard further tested the robustness of their results by comparing the performance of every song that simultaneously charted for more than one genre. As expected, they found that these songs were more successful in those genre charts in which they were more atypical.
Given that lyrical differentiation is an important reason for a song’s success, it follows that in genres such as dance music, which put less emphasis on lyrics (as opposed to the beat), the relationship between atypical lyrics and song performance would be weaker. The relationship would also be weaker in pop music, which, in the words of the authors, is “almost by definition... more about mainstreaming than differentiation.” Both these assumptions were supported by additional analyses.
So what can we conclude from these findings? That the less genre-typical the topics, the more successful a song? Not exactly.
A song needs to be similar enough “to evoke the warm glow of familiarity,” but atypical enough to “feel new and exciting,” the authors write. Case in point: Remixes may become popular in part because they add a new spin to beloved tunes.