And where does radio come in? That is, can radio "make" a hit, even if the song isn't "good" (i.e., inherently likable by audiences)?
Watts experimented on different groups of people with the same roster of songs. Some songs became "hits" every time, but mostly the hits were different in each group (i.e., chance).
Respondents were able to rate each song (and see the rating) and a distinct "follow the leader" impact was noted. That is, folks tend to like what other folks like, largely because other folks like it.
Interestingly, when Watts faked the rating scores so the high-scoring songs appeared low and the low-scoring songs appeared high, some of the "follow the leader" impact persisted, but ultimately some of the "hits" began to percolate back to the top. The broader effect, however, was that less music was consumed in this upside-down topsy-turvy world (lie about the merits of your product, and you may turn folks off to your category altogether).
So, concluded Watts, about half of a song’s movement could be attributed to intrinsic appeal. The rest was luck. As WIRED writes, "Rerun history, it seems, and Madonna could be working as a waitress."
It seems to me that radio's role in all this is profound.
Radio enables the "good" songs to rise to the top quickly.
Radio gives "chance" hits the best chance of success by exposing them to the most ears at once.
Radio facilitates a "follow the leader" effect by being the leaders – and picking the leaders – listeners are likely to follow.
Put that in your pipe and smoke it, music labels.