On my journey towards a master’s in Data Analytics and Visualization at the Maryland Institute College of Art, I took a course that asked me to bring data to life through visual storytelling, with an emphasis on iterations, iterations, iterations.
For you curious creatures who enjoy reading articles, context is provided below. For those who are part of the Impatience Club—welcome! Here are the final results:
Audience has always been central to my work, but this was my first time specifically working with and developing user personas, a common practice in UX design. In essence, user personas hone in on a target audience to help create a successful product or outcome.
For example, a vehicle is designed for a general population of humans. But maybe Human X wants to drive fast and furious, so you design a speedster. Human Y, on the other hand, is a certified daredevil and bonafide rebel who doesn’t have time for traffic. Naturally, a monster truck must be designed for them. Or, perhaps, a tank. (You get the point.)
That being said, let’s move on to the next stage—the data!
Tuning into the data
I landed on the final dataset through an amalgamation of past and present, cross-referencing Rolling Stone’s list of “100 Greatest Metal Albums of All Time” with Spotify’s ‘monthly listeners’ counts from January 2020 to August 2021 (scraped by MetalSucks). Spotify’s algorithm scores each song’s popularity between 0 and 100—with 100 being the most popular—based on the total number of plays the song has had and how recent those plays are. Songs that are played a lot currently, generally speaking, will have a higher popularity than songs that were played a lot in the past.
Since this project was aimed towards a performing, modern day metal cover band, I was interested in seeing which of the bands and albums listed by Rolling Stone were streamed the most, condensing those top performers into a current ‘Top 10’ list.
As for the four metrics plotted on the radar charts—valence, energy, acousticness, and danceability—I chose them somewhat arbitrarily from a dataset compiled by Sean Miller, who pulled audio data using Spotify’s Web API. While there were many audio features in Miller’s database, these four felt the most applicable and tangible. It certainly would be interesting to view all of the features as they relate to metal music, but in this case, less was more for a legible radar chart.
With these datasets and the prompt in mind, the persona I ran with was—drum roll, please…
Katie Kruel (She/Her)
Kruel is the badass lead singer of Mugshots, a grungy, metal cover band based in Cleveland, Ohio, home of the metalheads. During the day, Kruel works as an accountant. So, she understands the importance of numbers. However, she also understands the importance of fun and thriving outside the cubicle. What’s more fun than rocking out in front of a massive crowd while they headbang alongside you? (The answer is nothing. Well, except maybe data viz.) Kruel wants to know what it takes to get the head[bang]counts up and rip out the hearts of the crowd.
Rocking the charts
In my mind—and there’s a lot going on in this mind—two basic questions I imagine a cover band would ask are: Who are the most popular bands? and Which songs are the most popular? For the more nuanced questions, like What are their shared audio features?, you just have to get your hands dirty and dig! And visualizations are often the best way to do this. Radar charts seemed like a good way to display the audio features because:
- You can compare multiple variables that connect to form a group in a less cluttered manner.
- They allow for quick comparisons.
- Outliers are noticeable.
- You can use different scales of measurement.
- They look cool.
However, visually representing every song from an album on a radar chart, while insightful, was also a hot mess with no real practical value. Ultimately, I decided to visualize the most popular song (red) and the least (gray).
The record holders
My findings: The majority of the metal bands listed do not diverge from their established audio features. (Slipknot was the biggest culprit of this, maintaining their little mountain.) Van Halen, Black Sabbath, Ozzy Osbourne, and Evanescence were the bands with the most stylistic variation, particularly when it came to ‘acousticness’. These outliers felt significant because they show a band can stray from the “standard” metal sound and still have a strong following, which is pretty darn cool. Consequently, I chose to represent the outliers in the final visualization as outlines. (I was able to present this project to an amazing virtual audience [shoutout to my cohort], but I do recognize that I failed to provide an explanation of these in the key.)
To further break down the popularity of individual songs, I created a bar chart using the Spotify “monthly listeners” data.
Slipknot had the most evenly distributed album; each of their songs are relatively popular, without a supreme favorite. In theory, that would make them a solid choice for a band who wants to cover a sole artist, rather than individual songs. Another area of interest is the spikes in total streams, which generally occur when a band announces a tour or releases a new album. Sadly, the loss of a band member is another reason for these spikes.
A teachable moment
I am not the biggest fan of dark backgrounds because they are incredibly difficult to work with—kudos to people who do—and, for lack of a better explanation, they feel heavy to me. (Fitting for this project.) Generally, when it comes to data visualization, I would consider a dark background if the content was related to technology or had an ominous tone. So, what did I decide to do? Challenge myself and make a dark background! And what screams ‘METAL!’ more than red and black?
I tried to use the red sparingly, since it quite literally competes with the black for attention and often causes eye strain. High color contrast is useful for readability, but pure black text on a pure white background, for example, can cause eye strain because of the intense light levels that overstimulate the eyes. (Science!) Rule of thumb: When using a color picker, try not to have your brightness or saturation set to one extreme or the other (i.e. 0% or 100%). Here is one of my personal combos:
The primary critique I received was the lack of visually communicating my insights, which is absolutely fair. If there is anything I have learned over this past year, it is that personal insights are what make pretty much anything in life more compelling, especially visual storytelling.
That’s all she wrote, folks.
Melissa Malguy is shaping the design process from snout to tail with a suitcase full of wide-ranging experiences and complementary skill sets.