My Podcast’s Audience Was Tuning Out. AI Gave Me the Time to Fix It.
Better tools make faster work to have the time to make stuff that matters
I’ve been in the content game for a long time. Long enough to know that you can’t fake engagement. And for a podcaster, there’s one metric that tells the brutal, honest truth: listen-through rate.
On a B2B podcast I was running, mine was stuck. Often below 50%.
That’s a gut punch. It means more than half the people who hit “play” were bailing after just a few minutes. Something had to change. But like every other content person on the planet, I didn’t have any extra time to spend on that core creative task. I couldn’t just magically fix things because I had a show to get out every week, without fail. And when push comes to shove, you don’t miss publishing the episode. You play the hand you’re dealt and get ready for the next episode.
But the story of how it fixed things, doesn’t start with actual editing though. I started using AI tools to fix a really simple—but tedious and time consuming problem—transcriptions.
The “Toy” Phase
When I first started using AI-powered audio editors, I’ll be honest, I thought they were just toys. Not “real” audio editors. I was using Descript, but really just for one thing—better transcripts. It was faster to edit the text along with the audio instead of listening to the audio while trying to keep track and edit in Word.
It wasn’t a huge time saver, but it helped a bit in the final stages of production. I didn’t change my core workflow though; I was still spending hours manually editing audio, cutting “ums,” and fixing levels. I thought any web-based, AI audio editor was a “toy.” It was neat, but it couldn’t be used for serious, professional work.
Yeah I was wrong and I would find out how wrong I was when I had to work with a truly terrible piece of audio.
The “Magic Trick” (Nope, That Wasn’t It)
One week, I had a terrible piece of audio from a guest. Lots of background noise, laptop mic, no headphones, single track recording. If there was a textbook example of an interview where you’d be tempted to say, “forget it, this interview is beyond saving,” this was it. Of course I couldn’t just dump an interview, I had to do the best with what I had to work with. After hours and hours of trying to slice a single track into two tracks to fix the audio issues, I decided I had nothing left to lose and I loaded it into Descript and clicked the “Studio Sound” button.
It was amazing. It sounded like they were in a studio. It felt like magic. I was a wizard. No, I was an audio miracle worker! The episode was saved!
But not the show. Not really.
My listen-through rates didn’t budge. Yeah everything sounded better, but was the interview better? All I was doing was just fixing bad audio faster. The real problem wasn’t technical; it was strategic.
The real problem was taking interviews with good bones and making them into interviews people really wanted to listen to. And to solve that problem you need one crucial thing—time.
The Real Problem (And the Real Breakthrough)
Listeners weren’t dropping off because of a little background hiss or audio warbles. They were dropping off because the content wasn’t engaging enough. There wasn’t a story to follow. Things got lost in tangents that didn’t go with the “meat” of the interview.
We needed to make a change. The problem is, when you’re trying to turn things around, it takes a lot more time. Time to think. Time to be a creative editor, not just a technical one. Time to craft a narrative, experiment with formats, and find the thread in a 45-minute conversation and pull out 20 minutes of gold.
But that takes time and when sometimes I had less than a week to pull it all together, time wasn’t something I had a lot of. This is where the real breakthrough came. It wasn’t the magic “Studio Sound” button. It was the simple ability to edit audio like it was a Word doc. Descript probably wasn’t the first editor to do this, but I think they set the standard for how to do it right. But this isn’t about a specific tool it’s how using better, AI-powered tools gave me the time and breathing room to change how I approached episodes.
I could see the entire interview as text. I could read through and find good stuff. I could search for a word or phrase instead of trying to listen for it (and remember where in the interview it was).
- Want to create a different version of the episode? Duplicate and make the “alternate” take.
- Want to move the last five minutes of the interview to the front? Copy, paste. Done.
- Want to see what the episode sounds like without that 10-minute rambling tangent? Highlight, delete.
- Want to blend sections together to test a whole new story structure?
I could do it in minutes, not hours. I could produce three or four versions of an episode in the time it took me to edit one. Shifting my workflow from editing waveforms to editing text, cut my mechanical editing time by 30%.
Reinvesting the “AI Dividend”
Here’s the most important part. I didn’t take that 30% and just enjoy the extra breathing room. This wasn’t about doing less work; it was about reinvesting that time to do better work.
All the hours I got back from the mechanical, soul-crushing part of editing, I poured directly into crafting stories and working on strategy.
- I had time to listen to the interviews and find the real narrative arc.
- I could experiment with formats, trimming the fat and getting to the good stuff faster.
- I focused on the one thing that actually keeps listeners hooked: a good story.
The icing on the cake was using AI to automate the rest of the rote work. It drafted show notes, episode titles, and chapters. (It wasn’t great at finding social snippets, but templates helped speed that up too.) This freed up even more mental space for high-level creative work.
And the result?
A 40% jump in audience engagement. Our average listen-through rate climbed from less than 50% to over 70%. AI create better episodes. AI gave me the time to create better episodes.
AI Isn’t Magic, It’s Leverage, It’s Your Unfair Advantage
This is the lesson I keep coming back to when I talk about using AI. We get distracted by the “magic tricks”—the cool audio repair, the flashy image generation. But that’s not the real power. The real power of AI is as a lever to move the needle faster and faster than we could do alone.
AI tools—the good ones—automate the mechanical tasks to free up our human time. It gives us back the mental capacity to do the work that actually matters: the strategy, the creativity, and the empathy. So, stop just “testing” AI. Stop playing with the toys. Find a real, tedious, time-sucking part of your job. Hand it over to the machine. Then take the time you get back and go do the human work that no AI will ever be able to do.
Originally published at https://trishusseywriting.substack.com/p/my-podcasts-audience-was-tuning-out
