SongData (https://songdata.ca/) is a DH project started in 2018 at the University of Ottawa by Jada Watson and Andre Vellino to collect music industry data about country music songs to examine the inner workings of how the genre has developed over time, how it is still changing, and how its growth connects more broadly with cultural frameworks that emerge around it.  The project began with a need to understand inequalities in the radio playtime of women country artists: during a 2015 interview with radio consultant Keith Hill, he recommended that songs by female artists should comprise no more than 13%-15% of a station’s playlist. Hill described female artists as the “tomatoes” of the country music “salad.” SongData’s creators are looking to use this data to highlight gender discrimination in the country music genre by discovering trends and connections that could serve to keep songs by female artists low on the Hot Country Songs chart, and thus not as profitable as songs by male artists and groups. 

There are almost 20,000 songs in the project’s database whose sales led to ranking on Billboard magazine’s Hot Country Songs singles chart, from 1944 to the present.  The data for each single includes the artist, writer, producer, record label, and album, as well as the names and contributions of sound engineers and others who contributed to making the single. The creators note that there is simultaneously a clear bias against artists of color, including Indigenous performers in the genre, and they are working to use the data to help illustrate this discrimination as well. 

SongData uses several music industry databases to collect discographical and biographical information about the singles on the Hot Country Songs chart.  SongData creators use a Python script and RapidMiner to discover and analyze connections among music industry professionals that might have helped shape the genre over the years.  The SongData will be free to anyone interested in doing additional research about the development of the country music genre in the US. To date, they have completed a dataset of discographic and biographic information about charting country music singles from 1987-2017, and they published their first findings in 2019.  

What SongData is able to do is to quickly identify connections among record industry professionals and trends in sales of country music singles.  This data is drawn from several sources, and thus is best and most easily discovered through the use of digital gathering and analysis.  One shortcoming of the project is that, because of licensing agreements, they are not able to publish the Billboard Hot Country Songs charts.  While the analyses of the data can definitely help researchers spot trends and connections, being able to view the actual movement of the singles on the charts themselves would add weight to the project’s findings.  This project definitely has the potential to look at other genres in the music industry and discover more about how music genres are shaped behind the scenes.