Taylor Hohulin “What Music Research Can’t Do For You”

Winning With Low-Testing Songs: What Music Research Can’t Do For You

2024 was the year that Life 107.1 made a decision to play more low-testing songs and fewer high-testing songs. It’s also the year we recorded the highest cume in station history.

I’m sure more went into our growth than just the average test score of our music library, but I’m trying to grab your attention by overstating my point here. Bear with me.

On one hand, I’m a big fan of weekly music research. We have an active, passionate music panel here at Life 107.1, and I do believe we’re a better station for it. On the other hand, I’ve learned there are things research can’t accomplish. Here are some of those lessons I’ve learned:

Research can’t tell you what to do

Using music research to determine how our songs rotate seemed like a simple process: test everything in your library and then some, then break down the categories from the highest testing songs down to the lowest testing songs. You play your best songs the most and your worst songs the least.

The only problem? Sorting our playlist by research score led to a sound balance that leaned pretty heavily into one type of song. I assumed that sound was our core sound, since it tended to do the best…until a larger study of our whole market showed our listeners wanted more variety. So how do you reconcile consistent weekly research that shows a heavy preference for one particular sound and tempo against a large one-time study that shows a desire for more variety? That brings us to our next point…

Research can’t tell you what you really need to know

Without realizing it, I’d been expecting my weekly research to tell me how listeners wanted my playlist organized, when they were telling me something much different. Panelists generally aren’t thinking about whether they’d like to hear one song more often if it comes at the expense of another. They are telling me, on a scale of one to five, how much they like one song. And they’re giving that feedback in a vacuum.

I love data, but hard numbers will never truly measure subjective things like how good a song is or how popular a genre of music is. They can only measure indicators. We can look at which songs are charting on Mediabase, who’s getting good numbers on Troy Research, and all sorts of consumption data – all things that I think are truly valuable – but ultimately, what makes a song good isn’t how many spins, streams, and “play it more” clicks it gets. As much as we’d all like there to be, there just isn’t a data point for emotional connection.

Research can’t tell you the future

The thing about research is that it always measures something that already happened. Sure, you can map out trends by looking at what’s happened over the past weeks, months, and years, but research can never definitively tell you how things will change. There was a time Brandon Lake had not released a high testing song. Once upon a time, lots of us (myself included) knew that songs like Forrest Frank’s Good Day were unlikely to be embraced by our core audience.

If culture is shifting, then it only makes sense that some of the best things about our stations are things that never would have flown years ago. The key is knowing when to try out a potentially bad idea.

So what?

Music research is a valuable tool. The thing about tools, though, is they’re much more effective when you use them than when they use you. In order to get it right, you need a clear sense of mission and identity. If your goal is just to play songs that test well, you may miss some cool opportunities. But when your first filter is based on mission and market need, then you start asking research to do the things it does best.

Taylor Hohulin co-host The Breakfast Blend with Kristy and Taylor at Northwestern Media’s Life 107.1 in Des Moines, where he also serves as Program Director.

You can contact him to chat radio or weird science fiction at Taylor@Life1071.Com