Research Tool: MusicID

Well, I’ve gotten through a hectic semester end and I’m blissfully avoiding my summer writing goals, so let’s get back to posting.

I recently started doing a little work for Academic Rights Press, writing case studies to help demonstrate the usefulness of their analytic tools, and I’ve been finding some really interesting material.

I’ll tell you about some of the interesting stuff I found in the next post. Today, let me tell you about the tool itself, a record chart database called MusicID, because other people who do pop music research will probably find it really useful. (For full transparency: though I am working for them, they haven’t asked me to write this post and I’m not being compensated for it.)

At our last meeting, my dissertation advisor gave me their MusicID flyer. I didn’t even read it, because I was sold as soon as he told me that MusicID is a centralized database of all the world’s record charts. I’ve been referencing record charts in my dissertation, because they’re a good starting point and provide some measure of empirical data when trying to make a case for how widely known an album might have been. But if you’ve ever worked with charts, you know that 1) most of the official chart company archives are terrible and cumbersome to use, 2) finding reliable data from the ’50s and ’60s, when the charts were still in development and not really centralized yet, is an enormous pain.

You look up books or articles that offer research on best-selling records, and they throw out names and numbers without providing any clue whether they’re talking about worldwide, for their own country, which chart they’re actually referencing (Billboard had separate charts for stereo and mono albums until mid-1963), etc, etc. It becomes a huge headache in a hurry.

And then, when at long last you find what you need, you have to create an arsenal of spreadsheets, cross-referencing every other Hendrix-forsaken database you trawled through. It’s tedious. (Seriously, how is Billboard’s archive so badly designed? You have the money, Billboard. Call me. Let Katie help.)

So, when I heard there’s a database that collects all of the chart data in a single search engine, and makes it easy to sort search results by peak position/date/weeks in chart, I fell to my knees and thanked the Dark Lord for this gift. The search results also automatically generate graphs, so that’s another tedious step out of the way. And they’re working on building in more reference data, so that you don’t have to leave the software to look up discography information and such, which save a lot of tabbing back and forth.

I mean, imagine if instead of using JSTOR, you had to go to each individual journal and search for your topic. Or if your only choice for searching central publications in your field was sitting down with a pile of back issues and combing the tables of contents (which is what you have to do with old issues of Billboard Magazine).  I know that anyone who was in grad school before the internet will say, “uh, yeah, that was how we all did it” but it’s the 21st century now. We have better technology. Let it make life easier.

And for me, when I’m referencing charts, it’s usually just to provide one point of data to support an argument. The point of data itself isn’t central. So why would I waste 2 hours researching a single footnote that doesn’t even make or break my argument? It makes it stronger, but the argument doesn’t collapse without it. To that end, I appreciate being able to look up points quickly to add to my argument, without feeling my life ending one minute at a time.

But more than that, as I’ve dug into charts using the software, I’ve actually found a whole bunch of interesting stuff that I might not have otherwise thought about. So rather than just providing the occasional, incidental bit of data, I’m learning how charts can open up a whole new world of research questions—which are that much easier to discover when you don’t have to spend 2 days compiling your own raw data.

So check it out. There’s a lot of data that’s publicly available (the Revenue and Impact sections), and you can get a free trial to try the rest. It’s in development, so you have the opportunity to give feedback and make sure it grows into something that really serves academic needs. As it is now, it’s basically a central repository of record charts, with some streaming and revenue data, and it lets you export your search results as spreadsheets, so you don’t have to manually copy everything. You can easily take that exported text data and run your own python scripts to analyze it. As it stands, the tool is already well worth the cost of admission, and I’m surprised no one has done it before.

But the ultimate vision is much bigger. The CEO’s dream is to build on this and make MusicID like IMDB for music, where you can view record release info, personnel, tracks, writing credits, commercial history, and so forth, so that you have all the information you need at a click. AllMusic fills some of that niche, currently, but its search engine is woeful. If you don’t already know the name of what you’re looking for, it’s completely useless. It also doesn’t reliably include personnel and writing credits, nor does it include release, catalog, or chart information. I have to go somewhere like Discogs for that. So, if all of these different sources of information can be mashed into one uber-database, I’m down. I cannot abide inefficiency. Mama don’t take no mess. Get your data on the good foot. Hot tub!

(Um, as usual, please pardon my delirium. It’s nearly 90 degrees indoors now with 75-80% humidity. We are only lucky I don’t have full blown Dave’s Syndrome.)