Podcasting On The Wrong Side Of The (Music) Bed

I've launched a podcast about the Detroit arts and entertainment scene with Becky Scarcello and Jag called The D Brief. In this column, I am sharing the lessons I learn from it for other radio broadcasters who want to do the same.

Subscribe: Apple Podcasts • Google Play • iHeartRadio • Website

In my recent columns, I've talked about the method I use to gather feedback on our podcast, and I've shared with you some of the actual feedback that we received. As you can see, some of the feedback is incredibly insightful, and some is better off ignored. But when you begin to hear the same thing over and over again, it's best to pay attention.

Over the course of our first ten episodes, Mike and I solicited feedback via Amazon's Mechanical Turk. Over and over again, we heard the same complaint: people didn't like the music beds we were using in the podcast.

Most podcasts don't use music beds -- looped background music that the hosts talk over. You will hear it very infrequently in discussion podcasts like ours. For example, Pod Save America uses music at the end of episodes as they hosts are wrapping up. Storytelling journalism podcasts from the Ira Glass school might uses snippets to transition from one segment to another, or to underscore a particular piece of audio.

But Mike and I were using music beds throughout the entire show. We'd use a different looped beat for each of our show segments. These music beds ranged in style, from hip hop to rock to country to reggae. At no point, however, did we ever talk without music behind us.

I come from commercial radio, and for as long as I can remember, I've preferred the sound of disc jockeys talking over beds or over the songs they are introducing. To me, it gives the show a sense of momentum. Of course, not all radio programmers agree with me; there are plenty of radio stations where the disc jockeys talk over silence. But I wanted to bring a commercial radio aesthetic to our podcast, and the music beds were a defining part of that.

Unfortunately, the listeners loathed the beds.

We saw more complaints about the music beds than anything else in our feedback from Mechanical Turk. At first, we stuck by the beds, hoping the complaints were just a fluke. They weren't. Every batch of feedback revealed the same issue. But we liked the energy the beds gave the show, and we didn't want to abandon them.

We tried adjusting the levels of the beds. By reducing the volume, perhaps we could make the beds less irritating to listeners. I never listen back to the podcast on expensive, high quality sound systems. I like to road test the podcast under the same real-world conditions that most of our listeners will experience the show under: on a car stereo while driving, and through cheap earbuds while at the gym. Unfortunately, there was a discrepancy between the way the music beds sound under these two different sets of circumstances. In the car, the rumble of the road noise drowned out the music beds, so we would raise the beds' levels in the mix. But in headphones, the music beds would now overpower our voices, so we'd lower the levels. After several attempts, we finally settled on the lower levels, so you often can't hear the beds at all if you listen to the podcast in your car.

Still, the complaints about the music beds continued.

Mike and I recorded an episode without music beds. At the end of it, we both had the same reaction: we felt naked without the beds, as if the show had somehow lost its energy. Finally, we decided to compromise: we would include music beds on half of the segments. These would be the intro, the end credits, and the "list" segments where we rattle off upcoming concerts, comedy shows, movies, sports, or plays. We would try to keep these segments short. For all of the other more conversational segments, we would include a music bed that would fade out after 20 seconds or so.

Six months into our weekly podcast, that's still where things stand with the music beds. Yet I'm still not sure that we arrived at the correct answer. I love the movement that the beds provide, but sometimes the "listy" segments are long, and the looped beds can get repetitive. As I listen back to the show, I often feel that 20 seconds of a trailing bed on the non-list segments is too long as well. It's something that I'll continue to fret about as the show goes on.

Yet if I hadn't used Mechanical Turk to solicit feedback, I never would have known the music beds were an issue in the first place. The lesson? Never underestimate the value of feedback.

LISTEN: Hear the latest episode of The D Brief podcast.

More Glimpses Behind the Scenes:

1. I Paid Strangers for Feedback on My Podcast. Here's What They Said.
2. How to Get Actionable Feedback on Your Podcast
3. Why You Should Create a Teaser Before You Launch Your Podcast
4. Creating Benchmark Features for Your Podcast
5. Every Good Podcast Needs a Central Narrative