Podcasts

At the end of our show, we often ask you — our listeners — to send in topics you’d like to hear about.

Every week, we get a ton of suggestions in our inbox — ranging from clothes hangars, to model airplanes, to mechanical bulls at dive bars. But there’s one request we get more than any other.

LISTENER GN: Hi, this is Greg from Long Grove, Illinois —

LISTENER J: Hi, my name is Justin, from Albuquerque, New Mexico —

LISTENER AH: Hello, this is Ann Hodges calling from Kingston, Jamaica —

LISTENER JC: Hi, this is James Cage from Duluth, Georgia —

LISTENER WL: Hi, my name is Winnie. I live in Dallas, Texas —

LISTENER TF: My name is Todd and I am from Phillipsburg, New Jersey, and I would love to hear an episode about podcasts! Thanks very much, and I can’t believe you’re on your 100th episode already.

This is the 100th episode of The Economics of Everyday Things. And for the occasion, we’re going to take the advice of Todd, Winnie, James, Ann, Justin, Greg, and many more of you, and turn the mic on ourselves. We’ll tell you about what goes into making an episode of this podcast: how we explore ideas, find experts to talk to, check our facts, make the result sound good — and, most importantly, how we keep the lights on.

TARTAGLIA: If we had our choice of no ads versus ads, we would take no ads — all of us would. But there’s this implicit trade-off. Content doesn’t just come to you for free, there’s a cost. 

For the Freakonomics Radio Network, this is The Economics of Everyday Things. I’m Zachary Crockett. Today: podcasts — particularly, the one you’re listening to right now.

*      *      *

These days, it seems like everyone has a podcast. There are more than 6.5 million of them, and they reach nearly 600 million listeners worldwide.

ROTH: There’s a gajillion podcasts out there. And they’re all made slightly differently. 

That’s Gabe Roth. He’s the executive director at Freakonomics Radio Network. And one of his many duties is to serve as the editor of The Economics of Everyday Things.

ROTH: There’s the stereotypical podcast of, like, two dudes in their garage recording and then uploading the unedited recording. But professional podcasting is not like that. And, there are different models for doing it. What we’re doing — making a scripted, properly edited, high quality podcast that comes out every week — that is a lot of work.

When we debuted this show in 2023, the idea was simple: everything you encounter in daily life, no matter how small or commonplace — be it a traffic light, a tube of toothpaste, or a hamburger at a fast-food restaurant — everything is interesting when you look at it closely. At the start, we decided on a few basic guidelines. The episodes would be published weekly, and around 20 minutes in length.

ROTH: The point of this show is not to be exhaustive. The point of this show is to give people something that fits into smaller increments in their day. “This is not a symphony — this is a pop song,” would be one way of thinking about it.

But even orchestrating a pop song is a lot of work.

LILLEY: We can’t produce an episode of this show in the course of literally one week — we take about six weeks to do it. So, in order to put out a new episode every week, we’re always working on six episodes at a time, each of them in a different stage of the process.

That’s Sarah Lilley.

LILLEY: I am the senior producer of the Economics of Everyday Things.

Sarah is my creative partner. We work together on each episode of The Economics of Everyday Things from start to finish. She lives in New York and I live in San Francisco, but we’re in constant communication through Slack channels, phone calls, and Google Doc comments. Now, 6 weeks might sound like a lot to produce a single 20-minute podcast episode. But each one has a lot of moving parts.

LILLEY: The production stages are essentially: 1- researching ideas and committing to a topic; 2- booking guests; 3- conducting interviews; 4- writing the script, which means choosing the right clips from the interviews and writing the narration; 5- editing that script and fact-checking; and 6- creating the audio mix.

Now, remember that we’re making six episodes at a time, each one in a different part of this process. Sarah’s the person who has to keep track of the whole thing. For that, she created an impressive — and formidable — spreadsheet we call “the grid.”

CROCKETT: It always stresses me out a little bit to open this document. 

LILLEY: You know, this document gives me calm. I feel like, if I remember nothing, if I woke up tomorrow with amnesia —

CROCKETT: Please don’t!

LILLEY: No, I’m going to try not to! — but, I could open this document and know where each episode is on that path to being completed.

One great thing about this show is that there is an infinite number of everyday things to explore. I have a list of probably 500 potential topics that caught my eye at some point. On the weekends, I’ll spend some time reading about red light cameras, divorce lawyers, or used tires. I set up calls with truck drivers, send Facebook messages to party clowns, and ask utility workers on the street about the cost of stop signs.

ROTH: The ideas that are the most exciting for the show are often the ones where you’ll pipe up with, like, “You know, I’ve been noodling around on this and it turns out there’s a really interesting thing about this or that.”

But a lot of our episode ideas also come from you — our listeners

ROTH: We get a ton of great emails from people.

CROCKETT: Yeah, it’s like — someone will see a cell phone tower on top of their aunt’s apartment building in Florida and think to themselves, “How does this thing make money?”

ROTH: Yeah. And then they’ll think, “I should email Zack Crockett and ask.”

Every Wednesday afternoon, our team meets to talk over new ideas — there might be about 30 or so, each week. And we make decisions pretty quickly.

LILLEY: We’ll each identify the ideas that intrigue us. And then we hash them out — which means we go down a bunch of rabbit holes, sift though what’s in there, argue about it. I mean, there are a lot of topics that start falling into familiar buckets — like, “this industry has been completely overhauled by the internet, or this one by private equity” — and we don’t want to tell that same story every week. 

A few things go into making this determination. For starters, it usually has to be something commonplace and recognizable: a cemetery, a houseplant, or junk mail. But, every now and then, we break that rule for something we just can’t resist.

ROTH: People got mad when we did an episode on Stradivarius violins because they were like, “This is not an everyday thing — this is a very rarefied, unusual thing,” which, yes, that is true.

We also usually veto ideas that are too big for our show, or involve more policy, than economics. It’s hard to do justice to something like rocket launches, or birth control in 20 minutes. Every now and then, we do something extra ambitious — like our episode on prison labor. It required flying across the country to a correctional institution, interviewing nearly a dozen people, and reading through a bunch of policy briefs. But we can’t do that every week.

ROTH: There’s times when we realize a bit too late that we’ve bitten off a little more than we expected, and then we have to do some triage. You and me and Sarah, we’ll talk through like, ” what are the angles that we absolutely need to cover? Who are the people who can cover that for us? What can we do in narration? What can we leave out entirely?”

When we’ve settled on a topic, we have to find the right people to interview. We try to focus on practitioners, rather than people looking at an industry from the outside.

LILLEY: We absolutely want to talk to whoever is actually driving the tow truck or writing the romance novel. It’s only afterwards that we might interview an expert who has a view on the big picture. But everything hinges on securing that first person, the one who’s, you know, elbow-deep in it every day.

That means that a lot of the people we interview have never been interviewed before.

LILLEY: If you’re a pistachio farmer, maybe you’re not used to all of a sudden being put in front of a microphone. And it can be intimidating. It’s a strange thing to ask normal people to do.

Getting people on board can be a long process. In some cases, it’s 15 emails and phone calls back and forth before someone agrees to be interviewed. A typical episode includes recorded interviews with 2 or 3 different people. And each one of those interviews might be 60 to 90 minutes long.

LILLEY: So, if you think about it, we’re collecting 4 hours of guest tape that we’re going to boil down into an episode that’s roughly 20 minutes.

Most of the time, I’m conducting these interviews from my apartment in California. I’m sitting in a tiny closet, surrounded by my wife’s clothes. I have a Shure SM7B microphone set up on top of a dresser. And I’m talking to someone in, say, Bowling Green, Kentucky. I’m able to do that thanks to our engineering team.

JOHNSTON: I’m Jeremy Johnston and I’m an audio engineer at the Freakonomics Radio Network.

Jeremy helps us find recording studios for guests wherever they live, and he’s able to patch me in remotely using a software program called Riverside. It’s like a fancy version of Zoom for high-quality audio.

JOHNSTON: We have a list of go-to studios that we’ve worked with before and there’s over a hundred studios on that list and they’re around the world. The average rate, I would say, is about $150 an hour. 

If a studio isn’t an option, we do what’s called a tape sync. That might cost $300 or $400 bucks.

JOHNSTON: That is an audio engineer that will travel to the guest and they’ll bring all the gear needed in order to connect to us for the interview. And there’s a whole network of people out there that are focused on doing that.

We’ve had to arrange some pretty complicated interviews. Early on, we did an episode on people who dig up and sell T. rex fossils.

LILLEY: We were talking to swashbucklers out in the hills of Montana who were nowhere near any city. A professional recordist had to drive for like four hours with their equipment to get to the guest.

There was also the time we interviewed someone who works with goat herders for an episode on cashmere.

LILLEY: We booked people in Mongolia, which is where the best cashmere originates. But it has a 12-hour time difference from New York — plus, we needed interpreters and a local studio in Ulaanbaatar. It was a lot.

When the interviews are done and transcribed, I take all of this material — along with my research notes — and spend around 4 or 5 days writing a script. I’m organizing selections of tape from these interviews into a structure that makes sense, and I’m writing chunks of narration between the cracks to help tell a story. A 20-minute episode usually works out to around 3,500 words, or 10 pages in a Google Doc. Then it goes off to Gabe for an edit.

ROTH: A big part of what I’m looking for when I’m doing the edit is, is the narration going to sound right when you’re saying it out loud? Like you started in this business as a writer, you wrote for the web and for newsletters and you wrote to be read. And as you know, writing to be spoken out loud is a completely different process. And so some of what I’m doing in that edit stage is is taking out some of your writerly touches and trying to make them more conversational and more written for the ear.

During this stage, the script also goes to another very scrupulous person: our fact-checker, Daniel Moritz-Rabson. His job?

MORITZ-RABSON: At the highest, most basic level, it is to make sure that everything that we put out is correct. 

Daniel is a trained journalist, and he goes through pretty much every script I write with a fine-tooth comb. He checks out any claims our sources make, and verifies that my citations, facts, and figures are accurate. There’s often not a definitive source on the revenue figures of, say, wine corks, tattoo parlors, or strip clubs. So, Daniel has to do some sleuthing.

MORITZ-RABSON: For any given story I will reach out to probably a minimum of four think tanks or subject area experts or government agencies. I want to get two people, at minimum, saying, “Here’s what the situation is.”

It might take Daniel anywhere from 3 to 6 hours to fact-check a script. Once the script is finalized, I record my narration, and Jeremy cuts the tape. That’s a term from the days when audio was recorded onto magnetic tape, and editors had to cut it with a razor blade. But today, it means he’s going through big digital files in an audio editing program called ProTools. And he’s carving them down to just the parts we want.

JOHNSTON: Sometimes the transition between those cuts doesn’t sound very natural, so we need to smoothen that out. 

For instance, Jeremy will start with something like this:

JOHNSTON: So, a lot of the times in an interview, it’s really easy to go off on tangents and, you know, we want to, kind of, get to the point and — and cut things down to a concise episode. And it’s so funny. I can hear myself doing it right now while I’m talking and I’m imagining all of these cuts that I’m going to have to do on my voice.

And he’ll turn it into this:

JOHNSTON: In an interview, it’s really easy to go off on tangents. We want to get to the point and cut things down to a concise episode.

Another part of the production process is selecting the music that gets interspersed throughout the episode.

LILLEY: All of our music, including our theme music even, is licensed from a massive digital library called APM Music. You can pay per use, or get a blanket arrangement if you use a lot. So I dig around their site and collect a whole bunch of music that I think might work for us at some point.

A lot of work goes into finding the right spots in the episode for music — and also making sure that music complements the mood of the content.

LILLEY: I read through the script and I’ll kind of hear where I think music wants to come in and what it should feel like. Maybe we’re moving into a new mini chapter of an episode where we’re going from meeting some protagonist in our story, to talking about the history of a topic. Or maybe we’re tabulating a bunch of costs, which requires a different type of listening.

When all of the ingredients come together, Jeremy sends out a rough mix for us to listen to. We make final edits, and the episode is scheduled to publish across all the podcast platforms at 8pm Eastern Time on Sunday night.

If you consider the cost of studio time, music licenses, and the many hours of labor that go into the creative process, producing an episode of The Economics of Everyday Things can be expensive. In order to make money back — and hopefully, turn a profit — we run ads, like the ones you’re about to hear.

So, how exactly does the ad side of the podcasting business work? That’s coming up.

*      *      *

Our show is owned, and paid for, by the Freakonomics Radio Network. We’re mostly a bunch of journalists and audio nerds. And when it comes to monetizing the show, we get some help from a much bigger company — SiriusXM. They run hundreds of satellite and online radio stations; they own the streaming music service Pandora; and they make podcasts.

TARTAGLIA: My name is Gabe Tartaglia, and I’m the Vice President of Podcast and Satellite Monetization, which is a fancy title for ad sales.

Tartaglia oversees a team of salespeople who sell ads on around 500 podcasts.

TARTAGLIA: The goal of our sales team is not to pitch and sell everything to every client. We need to understand what their goals are, who their consumer is. And then most importantly, for both us and for you, is the ad that runs on your show works. They get the results that they want, and they renew. And they can stay with a given show for years at a time if it’s successful. 

Tartaglia has been in the audio business for more than 30 years. He started in traditional radio and later shifted his focus to podcasts, when they first became popular in the early 2000s. In the old days, podcasters taped ads and content together and published them as a single continuous piece of audio. Technology now lets us tape the ads separately and serve different ads to different listeners.

TARTAGLIA: What happened about five years ago was “dynamic ad insertion” became more prominent, which allows us to be able to insert a specific ad based on the show that an advertiser wants to buy, or even an audience that they want to reach. It allowed for more brand advertisers to come into the space and made it a lot more lucrative for podcasters to bring in more ad dollars.

There are a few different types of ads on this show. For starters, there are host-read ads. An advertiser contacts Sirius and says, “I want to run an ad on The Economics of Everyday Things, and I’d like Zachary Crockett to read it.” There are some things I might refuse to do an ad for — like a cryptocurrency scam, or something political. But if I approve the request, I get a script of the ad, I record it, and it runs in a designated episode, or series of episodes.

TARTAGLIA: That is extremely valuable to clients. They want the host to be able to speak on behalf of the brand. Because you’ve got credibility with your audience and they want to tap into that organically.

There are network ads, which aren’t specifically designed for my show, and run across a variety of podcasts.

TARTAGLIA: Network-based buying allows for advertisers to buy a certain audience segment across any of the 500 shows in our network. So they might not have bought Economics of Everyday Things specifically, but that ad runs and finds a home there because your audience is the right audience to match that advertiser.

Tartaglia’s team has data that shows our audience skews a little more male, generally in the 35-64 age range, and higher than average income.

TARTAGLIA: For a lot of different brands out there, it becomes a really compelling sell.

And these ads can even be served at an individual level. If you and your spouse each download this episode on a different device, you might be served different network ads based on your demographics.

TARTAGLIA: We know by certain identifiers who the device owner is, and we say, “Okay, this is Zack’s phone” or “This is Gabe’s computer,” and we serve ads thinking that that’s the person who’s actually going to be consuming the podcast. You click “download,” those ads get prefixed and shipped to you, and you never knew that it was dynamically inserted.

So, how much do advertisers pay for these ads? Well, the industry measures the cost in something called “CPM.”

TARTAGLIA: CPM stands for “cost per thousand,” and Roman numeral M of thousand, so that’s where it came from. And that’s been a stalwart in the industry for decades and decades. That’s what TV is bought on, digital media is bought, et cetera.

Let’s say a podcast episode has 100,000 downloads. If the CPM is $25, the cost to run an ad would be $2,500. Now, if the show has a million downloads per episode, that same ad costs twenty $25,000. Tartaglia says he’s seen CPMs from $5 to $100 per 1,000 downloads.

TARTAGLIA: You could have one show that’s in a genre that’s hard to sell, that has the same size audience as a very well-regarded show that’s easier to sell. And, you know, the ad revenue will be different on the two shows.

But even within the same show, the rate varies based on the type of ad.

TARTAGLIA: Typically, network buys, where they don’t buy a specific show, those are lower CPMs. We could run that ad anywhere, so we’re just kind of filling the basement, if you will, of our supply. 

For advertisers, the cost of an ad also depends on where it is in the podcast. You may’ve noticed that there are three different spots in our podcast where ads run: the beginning, middle, and end. These slots are called pre-roll, mid-roll, and post-roll, and the first two categories are the most desirable and most expensive.

TARTAGLIA: Pre-roll is typically about 20 percent cheaper than mid-roll, and then post-roll is, you know, very nominal. Because there’s no guarantee that someone’s going to be listening and staying on at the very end of the podcast. 

Sirius often sells multiple ads in each of these slots. Let’s say a 20-minute show with 100,000 downloads has 5 60-second ads — two pre-rolls, two mid-rolls, and a post-roll. If you average a $25 CPM across those ads, that’s $12,500 in revenue. But because of the last-minute nature of how network ads work, it can be hard to reliably predict the income a podcast can generate.

TARTAGLIA: The ads that are running on that show specifically get placed first by the ad server. And then anything else, any open inventory that’s left over, the second tier of those “network ads” can run in those open slots. The network delivery happens daily. We can’t predict even tomorrow how much network revenue your show will get versus the next show because it’s literally dynamically inserted in real time. So every single day the system updates to say, “Okay, this show got, you know, $10,000 worth of network ads ran yesterday and that show had $20,000,” and it accrues and actualizes on a daily basis all the way through a quarter or a year.

In exchange for facilitating all of these ad deals, SiriusXM gets a cut of the revenue. They have an incentive to sell as many ads as they can on the podcasts they work with. But of course there’s a ceiling to how many ads people will listen to.

TARTAGLIA: Typically you might hear, you know, six to eight ads that might take four to six minutes. That’s probably getting towards max capacity for what’s comfortable in a listening experience.

Brands have different ways to measure the success of the podcast ads they run. Some include a website or promo code that listeners can use.

TARTAGLIA: And then in the next several days you look at, okay, how much traction to that website or visitors did we get? And the ad either worked or it did not. 

Podcast ads also often have less visible tracking tools at play.

TARTAGLIA: A lot of digital ads have a tracking pixel running in the background, where as long as you accept all cookies and, you know, agree to privacy compliance, that pixel will live behind the ad. You hear that ad for Wendy’s, and then your phone shows up in a Wendy’s — that content gets credit for helping drive your phone into a Wendy’s. So, it can get pretty sophisticated and to the point of almost creepy on the back end.

These improved tracking capabilities have boosted demand. Podcast ad revenue was up 26 percent last year — more than 3 times the growth for the digital audio industry at large.

But while Gabe Tartaglia and his team are selling those ads, we’re working on telling good stories. Our editor, Gabe Roth, hates it when I end an episode by saying “at the end of the day,” so this will probably get edited out. But at the end of the day, we’re in the business of helping people appreciate the ordinary things in life, whether it’s a plate of food on an airplane, a stock photo, or a porta-potty.

ROTH: The conceit of this show is that anything you see is interesting. And if that conceit gets into your head, then you start being interested in stuff that you might not have noticed before. You start paying attention in a different way to the world around you. And if we’re doing that, then that feels like a real success.

For The Economics of Everyday Things, I’m Zachary Crockett.

*      *      *

This episode was produced by me and Sarah Lilley, and mixed by Jeremy Johnston. We had help from Daniel Moritz-Rabson. There are other people here at the Freakonomics Radio Network who often contribute to this show. Morgan Levey has written and co-produced several episodes for us and is in our editorial meetings every week, pitching in ideas. Dalvin Aboagye helps us do research and publish all of our transcripts online. Greg Rippin and Jasmin Klinger often sit in on our recording sessions and always make sure my “Ps” don’t pop. And, of course, Stephen Dubner runs the whole shebang. Without him, this show wouldn’t exist. Thanks to everyone who’s been listening to The Economics of Everyday Things over our first hundred episodes, and especially the dozens of our listeners who suggested this topic. I’m sorry we can’t name you all here, but I hope this answered some of your questions. If you’ve got an idea of your own, feel free to send us a note at everydaythings@freakonomics.com. We’ll catch you next week.

ROTH: Let’s say we’re doing an episode on weasels. Here’s a weasel salesman and he knows all the ins and outs of weasel distribution and weasel promotion and weasel manufacturing and he’s willing to open up his books and say, here’s what I pay per weasel and here’s what I take in per weasel and here’s how much goes to overhead and here’s what my profit margins are on every weasel and if I sell weasels in bulk then it looks like this. 

CROCKETT: Are you lobbying for a weasel episode here?