Be A Data Maestro

How I would go about doing a music industry data analysis

Pie and Donut Analytics
Away From Towards Data Science
3 min readSep 15, 2021

--

Ah, the 90s. That was a great time for music!

After months of writing articles about data analysis in such varied industries as healthcare, soccer, and even animal beauty products, I decided it was high time to write about my lifelong passion — music! It’s been my religion ever since I was a wee child listening to my mom’s Red and Blue albums (please refer to the caption of the database photo further down in this post if you are not sure which band I am referring to).

So before I get into the nitty-gritty wrt data analysis, I shall digress and put a disclaimer out there that this article represents MY views only. I don’t have any knowledge about Spotify and Pandora trade secrets. I have never worked at either of those companies or any music-related enterprise for that matter. I once managed to land an interview with Capitol Records in Hollywood, which this fangirl was super over the moon about! They were the record label of the first artists I ever saw in concert — School of Fish opening for Crowded House.

Unfortunately, I did not get the job because I didn’t have Excel skills yet (back then, it was not ubiquitous on both workplace and home computers as it is now. Chicken v. egg problem *sigh*). I did notice however that on my interviewer’s desk was a stack of cassette demo tapes (Millennial and Gen Z readers: are you thinking that this article is so cute and quaint yet?). So, I didn’t leave empty-handed after all — I scored for myself a totally rad consolation prize: the debut album of Supergrass! Boomer readers: you may know them for their mainstream hit ‘Alright.’ Silent Generation readers: I was wearing my Supergrass t-shirt one day, and I earned a disapproving stink eye from one of you. I promise I was not then, nor currently am, a stoner (not that there’s anything wrong with that). Don’t be so literal!

Analyzing Music Preferences Across Generations, Continents, and Other Slices (keep reading, you’ll see what I did there)

I had an idea for a purely-for-fun data analysis project. So I put out a tweet as well as Facebook post in an attempt to gather data. While the number of responses I got was not enough to warrant a social graph, it allowed me to begin putting together an initial database that I could try to glean some insights from.

My mom and I both love The Beatles. Oh darn, I notice I misspelt Alicia Keys. Gotta correct that NOW. Normalized data is a MUST for any data science project!

You know what, I should end this article soon and make it Part One of a series (my first!) here on Medium. After all, as I proved in an earlier article, the shorter it is, the higher the engagement. I could do an analysis using the whopping 7 Person IDs I have at the time of this writing, but it would be more interesting if I had more data. For example, I only have one Millennial so far, no Generation Z, and only Whites, Asians, North Americans, and one European. There is still time to contribute to my project if you click on the hyperlinks to my social media posts in the previous paragraph (they are set to Public so you should be able to view and interact even if we are not connected). Hope to hear from some more music fans!

--

--