Emily Wright is a recent graduate of Carnegie Mellon with a B.S. in Economics and Statistics. Wright, who was given access to the Next Big Sound API in order to collect relevant data for her thesis, was awarded first place at Carnegie Mellon’s Meeting of the Minds research symposium for “Oral Presentation of an Honors Thesis” for the department of Statistics.
Even in this digital day and age, commanding radio attention is key to becoming the Next Big Sound. Not only does radio reach a large audience, 244.4 million consumers a year, but radio is also the most common way for consumers to discover new music. If we are looking to identify the Next Big Sound, we can do so simply by deciphering the path artists take from their early online career to their first major radio appearance.
In the paper “Artist Music Discovery: The Digital Road to the Top of Radio” I aim to explore this path. First, I created a total view of an artist’s path from the online sphere to the radio sphere using statistical data matching techniques linking artists’ online metric data published by Next Big Sound to artists’ radio airplay data published by Digital Radio Tracker. With this comprehensive dataset in hand, I then performed longitudinal modeling to capture the relationship between artists’ online presence, radio airplay, and time.
The online mediums examined include: Facebook, Twitter, YouTube, Wikipedia, SoundCloud, and Vevo. For each artist the time period examined began once an artist established a presence on all six online metrics and ended with an artist’s first major radio appearance or at the end of 2013 if they were not aired on radio. To ensure the radio appearance was indicative of success, only artists with the highest radio airplay ranked on the weekly top charts of radio were considered successful. In total, 2,933 artists were examined of which 71 were defined as successful.
Generally, we would expect increased attention on any medium to lead to increased chances of being aired on radio. However, the only sources found to be statistically significant were Twitter, Vevo, and SoundCloud. Potential reasons as to why a major network like Facebook fails to predict, could be due to the ubiquitous use of the platform. The relative ease with which any artist can gain Facebook page likes makes them inexpensive. In other words, if every artist gains Facebook page likes at the same rate a successful artist and a non-successful artist appear to be the same. Thus Facebook does not provide useful information to identify successful artists. Youtube may not be predictive because Vevo videos, hosted on Youtube.com, may command more attention than the Youtube videos as viewers tend to prefer the higher production quality of Vevo videos. Finally, Wikipedia may not be predictive because it does not receive enough attention on a regular basis to sufficiently signal future success.
Keep in mind, radio is only one of many ways to define or measure artist popularity. There are many other ways to do so. The question is whether there is a general underlying relationship with increased attention on Twitter, Vevo, and SoundCloud and popularity. To determine if there is consistency further research examining the same relationship, but with different measures of artist popularity, is needed. For now we look to radio.