Yes, Podcast Listenership Is Still on the Rise: Report

Photo: Eddy Tang/Getty Images/EyeEm

If you’re a longtime Hot Pod reader, you probably know that I hold Edison Research’s annual Infinite Dial study in high regard. The survey-based study of digital media usage has been the longest-running measure of podcast audiences going back to the medium’s earliest days, and as a result, the story they’re able to tell is the one I consider the most reliable.

That said, I didn’t spend as much time covering last year’s edition of the study for what should be obvious reasons by now: It was released at around the same moment that the United States began its descent into the COVID-19 pandemic. Leafing through the report at the time, it didn’t make much sense to me to allocate much attention to the year that had come before when what lay ahead felt so deeply unpredictable.

I don’t need to tell you that a lot has happened over the last twelve months. From a purely podcast standpoint, the wave of lockdowns that began last spring — then ebbed, then flowed, then splayed out into a messy patchwork system — resulted in some initial declines in listenership as the morning commute went away, along with a significant restructuring of work processes and mild consternation over whether there’ll still be a podcast business on the other side of the pandemic.

Eventually, though, podcast consumption rebounded as its structural advantages within the context of pandemic conditions came into sharper view. The medium lent well to remote-production workflows, which in turn attracted more participation from creators and celebrity talent and media companies, which in turn led to the creation of more podcasts and greater recruitment of their respective followings into the medium. Listening behaviors as a whole ended up adapting, moving away from the morning commute and towards more afternoon consumption. The case began to be made that podcasting, more so than many other new media infrastructures, was uniquely suited to meeting the moment. But the question was: To what extent?

The 2021 edition of the Infinite Dial study, published last week, gave an answer: to a considerable extent.

Let’s break the report’s podcast-specific findings out. To begin with, the study recorded gains in the major audience sizing metrics:

➽ 41{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} of the total U.S. population over the age of twelve, or an estimated 116 million Americans, can now be considered monthly podcast listeners, up from 37{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} the year before.

➽ 28{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} of the total U.S. population, or an estimated 80 million Americans, can now be considered habitual weekly podcast listeners, up from 24{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} the year before.

➽ Meanwhile, podcast familiarity — that is, the extent to which Americans are aware of the medium — continued to grow, present among 78{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} of the total U.S. population, or an estimated 222 million Americans, up from 75{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} the year before.

The American podcast audience was also found to have grown more diverse from a gender and ethnicity standpoint, with the study arguing that it has drifted towards a composition that more closely reflects the American population. (One specific finding that leapt out: There were exceptional gains among Hispanic listeners over the past year in particular.)

The report also found that the American podcast audience has deepened their engagement with the medium more generally. This is represented in the finding that weekly U.S. podcast listeners now average eight podcasts per week — typically interpreted as “podcast episodes” — up from six podcasts per week.

A quick note on some methodological progression here: This year’s report also includes a new “average podcast shows in the last week” measure, made distinct from a “podcasts per week” metric. The specific finding on that front: Weekly U.S. podcast listeners averaged 5.1 podcast shows in the last week.

Let’s pause on this beat for a second, because there’s a vast universe of analytical angles baked into this one data point. On a gut level, that feels like a small average number of shows per active podcast consumer especially when held up against an ever-expanding podcast ecosystem, with new shows launching just about every week (or day, or hour). At any rate, it’s worth introducing some level of complexity to that feeling: Not all shows possess a regular weekly publishing cadence, not all shows should be built to compete for everybody’s regular listening slots, and not all niches are adequately covered in the current spread of what’s available. In my mind, there’s room to grow in all directions, and besides, I’d be curious for this ratio of average consumption per user to average production of whole medium to be weighed against other media, whether it’s books, video games, or even the ever-increasing preponderance of products distributed over streaming video services.

Anyway, as always, I highly recommend you go through the report in full, if only to get a better sense of the change over time. But before we move on, I wanted to flag a few other things from other parts of the report.

It should be clear by now that the podcast ecosystem is being fundamentally stitched into other media systems, whether we’re talking about the medium’s competition for listening time against other audio formats (like audiobooks) or how it’s being increasingly absorbed by competition between the large audio streaming platforms.

To that end, here are some of the relevant findings that I’m tucking away in the back of my head:

➽ The report argues that “Spotify has solidified its spot as the largest single-source for online audio, and has played a role in the growth of podcasting (especially with younger listeners).” The platform leads in all the important measures, with Pandora consistently coming in second place.

➽ Audiobook listening seems to be flattening back out. After a spike in the 2019 study (50{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} of the total U.S. population, up from 44{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} the year before), that measure now hovers at 45{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} and 46{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} of the total U.S. population over the past two studies.

➽ Some interesting findings within the context of in-car media consumption. Of course, the broader point to consider is the fact that folks are driving less during the pandemic, but it’s still interesting to see that AM/FM radio has dropped to 75{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} of population from 81{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} of population in the “audio sources currently ever used in the car” measure and that half of the total U.S. population engages in online audio listening in the car through a cell phone, up from 45{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} of the population the year before.

Finally, shout-out to the new survey questions on Twitch live streaming in there. I know I’ve been watching a hell of a lot more random streams since the pandemic began.

➽ In case you missed it, I spent two weeks reporting out a feature for Vulture that looked into what happened with Reply All, Gimlet Media, and the union push that took place in the lead up to the podcast company’s acquisition by Spotify in early 2019. That piece dropped last Wednesday.

➽ And as it turns out, I wasn’t the only one poking around. The New York Times and The American Prospect also published pieces on the same day, with the former containing specifics on just how much money various key figures in the Gimlet story made off the sale to Spotify.

➽ Coincidentally, later in the week, the Gimlet Union would secure its first contract with Spotify after two years of bargaining, announcing the deal over its Twitter account early Friday morning. The Ringer Union also secured its own first contract a day later. Parcast, the third content division within Spotify that’s formed a union, is still in the bargaining process.

➽ Entercom, the broadcast radio company that acquired Cadence13 and Pineapple Street, is moving to acquire podcast advertising marketplace startup Podcorn, the Wall Street Journal reports. The deal will value Podcorn at $22.5 million.

➽ CAA is promoting Josh Lindgren as the head of its podcast department. My impression is that this is merely formalizing what Lindgren has already been doing since he joined the talent agency in the summer of 2018. Lindgren’s clients include NPR’s Ari Shapiro, iHeartMedia’s Stuff You Should Know, Jane Marie and Dann Galluci’s Little Everywhere, Maximum Fun, and the Futuro Media Group, among others. Here’s the Variety write-up.

➽ Apple Podcasts is apparently shifting away from “Subscribe” to “Follow,” Podnews found last week. The thinking being — or at least it’s thought to be — that the word “Subscribe” is generally associated with media products that aren’t free.

➽ Keep your eyes peeled on this. From Insider: “Amazon’s ad boss, Alan Moss, told advertisers that the e-commerce giant plans to roll out ads in podcasts.” How and through what means, exactly, remains unclear, but I imagine there’s quite a bit of runway between the Amazon Music platform and the possibility of podcast ad tech-related acquisitions in the months ahead.

➽ Shout-out to the crew over Nieman Lab, who’ve been syndicating RQ1, a new monthly newsletter going over the latest academic research around journalism. A recent issue covered Gabriela Perdomo and Philippe Rodrigues-Rouleau’s paper, “Transparency as Metajournalistic Performance: The New York TimesCaliphate Podcast and New Ways to Claim Journalistic Authority,” which takes a scalpel to the performance of transparency in Caliphate specifically, but in narrative audio more generally. Really worth the time.

➽ I think this may very well qualify as a first. From the New York Times: “‘Nobody Wants to Be There, Dude’: How a Juror’s Podcast Led to an Appeal.” To fill in the blanks a little bit, the juror is a standup comedian and the podcast has a following of like… a hundred people.

From TechCrunch: “Apple discontinues original HomePod, will focus on mini.”

It’s been some two years and change since Leon Neyfakh left Slate to sign a Luminary deal, pop open his own production studio, and set forth to find out what greener pastures may grant him. As you would expect, he’s spent the bulk of that time being plenty busy, and when I checked in on him last week, I’d learn that his studio, Prologue Projects, seems to be evolving rather unexpectedly.

The studio continues to traffic in the sort of narrative nonfiction series you’d expect from the originating host of Slow Burn, dulcet tones and historical perspective and all. At the heart remains Fiasco, its flagship series that’s gearing to launch its latest season later this spring, on the Benghazi attacks. (As far as I can tell, it’s still a Luminary exclusive.) Then there’s The Edge, a sports series hosted by the journalist Ben Reiter that spent its debut season on the Houston Astros cheating scandal debacle. The studio was also attached to a project last fall with Cadence13 and Alex Gibney’s Jigsaw Productions, called Whirlwind, which provided a sprawling look into U.S.-Russian political intrigue that stretches to the present day.

But beyond the limited-run documentary series stuff, Prologue has also been rounding out its portfolio with shows that adhere to more regular publishing schedules. This includes the intensely strange 365 Stories I Want To Tell You Before We Both Die from the author Caveh Zahedi, which drops tiny dispatches every day and, weirdly enough, has a lot in common with the structurally similar 365 days with mxmtoon, though, of course, the differences between the two hosts there couldn’t be more dramatic. It also includes a decidedly more straightforward comedy podcast called Celebrity Book Club with Lily Marotta and Steven Phillips-Horst, which comes out of a co-production with Headgum and has the additional challenge of going up against an assortment of other podcasts of the same name and concept, including one hosted by Chelsea Devantez and produced by Stitcher.

Anyway, all of this is to say that Prologue appears to have spent the past two and half years figuring out its angles and mixing things up when necessary, and there’s no more representative example of these dynamics than what the studio recently had to do with another one of this regularly publishing shows, 5-4, which I think also serves as an interesting window into one instance of how things are going on the mid-range ad-supported side of the podcast business these days.

Broadly speaking, 5-4 is a conversational podcast that pitches itself as a show “about how much the Supreme Court sucks,” with each episode dedicated to diving deep into a different Supreme Court case that stands out for what the show would consider its terrible legal reasoning or effects on society. It’s hosted by three progressive lawyers who maintain some level of anonymity beyond their first names — Peter, Michael, and Rhiannon — in large part due to the fact that, well, they’re practicing lawyers. The podcast launched about a year ago, making its debut with an archetypal episode about Bush v. Gore, before proceeding to bounce between a mix of foundational “terrible” cases and cases that bring the discussion close to contemporary news developments.

In the beginning, Prologue positioned 5-4 as a fairly straightforward weekly podcast that would be supported by advertising. They pitched the show around to various distributors and ultimately ended up working with Westwood One, the Cumulus Media-owned radio group that’s been working to build out a podcast business over the past few years. They forged a standard distribution deal, with Westwood One handling ad sales as well as providing promo spots over their radio stations and some of their own podcasts. As is usually the case with this type of arrangement, the deal paid out what Neyfakh described as a “modest” minimum guarantee to cover the cost of production.

5-4 went on to attract a solid following. Its core constituents, according to Neyfakh, seemed to be some mix of law school students, law professors, and civil rights lawyers. In other words, a kind of listenership that some people would consider pretty monetizable, given the presumed long-term economic status of lawyer and law-school types. By the end of last year, the podcast grew to around 40,000 to 50,000 downloads per week, with monthly downloads across the show’s entire catalogue averaging around the 200,000 mark.

For all intents and purposes, that’s a pretty respectable number for a podcast that doesn’t feature widely known names attached to the production. (In fact, as previously mentioned, there were no full names at all, save for Neyfakh, who is listed on the show as a “presenter.”) And as far as I can remember, that download volume was enough to get you a fairly decent stream of advertising interest back around say… three or four years ago.

Of course, much has changed in the past three or four years, and despite 5-4’s solid first-year showing, Neyfakh told me that the podcast wasn’t able to make reliable advertising money through their Westwood One deal. (Westwood One, by the way, declined to comment when I followed up for more details.)

This struck Neyfakh and the Prologue team as incongruous with the opportunity that 5-4 was exhibiting. Advertising interest or no, they felt a strong sense that the show was cultivating a particularly engaged community around it. There were a few metrics that suggested as much, but the one that really grabbed their attention was merchandise sales: In the first week after opening an online store, they sold a thousand orders. Surely that’s an indicator of something.

After the deal with Westwood One ended, the team briefly considered going back out to the marketplace and finding a new distribution partner. But as they surveyed the scene — ever filling up with bigger names, more shows, and greater consolidation — they felt like a show carrying 200,000 downloads per month isn’t really in a position to make serious ad dollars. And so, taking a cue from an increasing number of podcasts that find themselves in this peculiar in-between space, they shifted their efforts to Patreon.

For the Prologue team, the choice to roll with the membership platform in particular was partly informed by who was already on there. “Patreon is already home to many other lefty podcasts,” said Neyfakh. “We had a sense that a lot of the 5-4 fanbase already knew how to use it, in some cases maybe they already had their cards saved in.” They moved the show from Megaphone (where it was previously hosted under the Westwood One deal) to Acast (which recently launched a direct integration with Patreon), and constructed an offering around two tiers: one at $5 per month, which provides exclusive episodes and discounts on merchandise, and one at $10 per month, which adds on a Slack channel with the hosts, exclusive events, and even deeper discounts on merchandise. They also went in with fairly conversative expectations. “In terms of a rule of thumb, people talk about converting anywhere between 0.5{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} to 5{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} over the course of the first year,” said Neyfakh.

The payoff turned out to be near instantaneous. Within a week of launching the Patreon, the podcast converted almost 3,000 supporters at almost $200,000 in annual revenue. At this writing, almost a month after launch, the 5-4 Patreon is a hair over 3,100 subscribers, generating around $16,900 per month in revenue. That’s far from, say, Chapo Trap House’s 38,400 Patreon supporter base and $170,000 monthly revenue, but hey, it’s just the first month, and it’s a start.

5-4’s experience is an interesting one, though I’m personally wary of the extent to which we can extrapolate conclusions. For one thing, this could very well be a story about Westwood One’s effectiveness in third-party ad sales, or more likely, a story about the impact of the pandemic on advertising around a new show launch. But 5-4’s difficulties garnering advertising interest does seem to square more generally with the stories of many other similarly sized shows that I’ve been hearing a lot about lately. Shows of a certain considerable but not exorbitantly large size that could’ve gotten a response from advertisers just a few years ago now seem to be increasingly shut out of that monetization opportunity, at least in its current form — things may well change as programmatic options open up over the next few years.

After its pivot to Patreon, 5-4 is now significantly better positioned for the next stage of its life, and its experience stands as a good example of the fact that a race for scale isn’t the only way to run a shop in this business. In any case, it’s not like advertising is off the table completely. “Maybe if the show grows to a certain size, advertising could be a good option for us once again,” said Neyfakh.

In tomorrow’s Servant of Pod Sarah Marshall of You’re Wrong About and Why Are Dads? is on the show this week.

It turns out I’ve been a fan of Marshall’s work well before I ever picked up You’re Wrong About, which I should say makes what Edison Research would consider one of the five shows in my regular rotation. (J/k, I ping-pong between maybe fifteen regular shows, though I do go through phases, and obviously I’m an irregular customer.) Back in 2017, she published a fascinating piece that dug deep into the backstory of Titanic — the James Cameron epic, not the ship itself, though I’m sure that’s interesting — which I kept regularly returning to in my Pocket app because it’s exactly the “behind the scenes of an outrageous Hollywood production shit” that tends to hook me in the gills.

Anyway, point is: It makes perfect sense to me that she’s the very same Sarah Marshall who makes one of the two hosts of You’re Wrong About, which isn’t just a great example of a strong independent podcast, but also a show I would personally consider as being the purest form of what an indie conversational podcast can be: a few people talking, good vibes, smart as all hell.

You can find Servant of Pod on Apple Podcasts, Spotify, or the great assortment of third-party podcast apps that are hooked up to the open publishing ecosystem. Desktop listening is also recommended. Share, leave a review, so on.

By Aria Bracci

When you think of the last podcast episode you listened to, do you remember where you were? Doing laundry? Taking a walk around the block? Perhaps you would’ve had a different answer a little over a year ago, before the pandemic vastly limited the amount of places you could freely be.

As the anniversary of the first U.S. lockdowns rolled around, I began to wonder how audio creators — in particular the ones who hope their work will be consumed intentionally and with few distractions — felt about how commutes, long said to be a primary context of podcast consumption, had been scaled way back. Did this radical shift end up resulting in more of what these folks might consider “intentional” listening? So I started asking around, and what I ended up finding was some reinforcement of a whole other idea: that ideal, “intentional” listening environments don’t actually exist.

Let’s start here: Listening on the subway isn’t how Brendan Stermer, host of Interesting People Reading Poetry, would prefer his show be heard. Instead, this is what he recommends: “Good headphones, quiet space, normal speed, and minimal distractions.”

Interesting People Reading Poetry features well-known figures reciting poems (two times through, as a professor might advise) and revealing a bit about what the writing means to them. For it to mean something to listeners, Stermer says, they need to put in a little work — which, he acknowledges, is a lot to ask.

“I think appreciating poetry requires really careful attention to detail,” he says. And if you’re cringing at the prospect of undistracted listening, he has a qualifier: “We don’t put out very many episodes at all,” he says, only publishing installments that he and brother Andy, who also composes original music for the show, think are the best of the best.

With so much going on behind the scenes, audio creators expect — or at least hope — that their work will shine through when someone listens. Take, for example, the podcast Radio Drama Revival. Similar in structure to Interesting People Reading Poetry, it’s an anthology of pre-written works, and the creators of those works get to say a bit about their process. Host Elena Fernández Collins hopes that listeners will both respect the discussions and feel immersed in the audio drama — and, to that end, she does a lot to set the scene.

“It’s like being taken to a movie theater,” says Fernández Collins, who assumes the responsibility of transitioning the listener from the real world to the show’s “mysterious, noir-ish” space. “When I record, I record fairly close-mic’d,” she says, in addition to lowering the pitch of her voice, which is then backdropped by jazz music. “The vibe that I imagine in my brain is Alan Cumming when he’s hosting MASTERPIECE Mystery!”

To achieve the optimal effect, Fernández Collins says, listeners would ideally consume Radio Drama Revival like a glass of wine: “at night, doing something calm.” That’s at least what she hopes. She, for one, listens to podcasts “on one-time speed without silences trimmed,” but even she can’t guarantee that she’ll always be able to listen in an ideal setting, let alone the serene circumstances she imagines for her audience.

Neither, it turns out, can Stermer.

When I asked him about his listening habits, he paused, then replied — “That’s really embarrassing.” He listens to primarily interview-based podcasts, he says, and the way he listens is in stark opposition to what he’d suggested: “The way that I listen to most shows, aside from my own, is through my internal iPhone speaker, while I’m doing dishes or something.”

Jennifer Mills, a producer for Everything is Alive and Wait Wait… Don’t Tell Me!, described a similar contradiction. As a creator, she says, “I love to take in the final product on a walk,” which is also what she found herself doing as a listener. She made a habit of listening to a daily news podcast during a morning stroll, but sometimes, she says, it just didn’t work.

“I find that, once in a while, the weather or the vibe of the walk is [in] too great of contrast with what the episode is about, that I actually have to turn it off,” she says.

On the website for Everything is Alive, the first thing that appears is a message: “If you’ve just found us, we recommend starting at the beginning,” after which a caveat — “But it’s up to you” — appears. “Personally, I don’t think there’s an ideal way to listen to it,” Mills notes; but, she says, wading through the personalities and narratives that the show builds out for its subjects “is like having a relationship with different people,” in that the order that they enter your life matters. But that’s not all.

“When I listen to the show, I think about a void, and I think about the conversation existing in an imaginary space,” Mills says, then laughs. “So maybe if everyone can get into a sensory deprivation chamber… ”

Being a listener herself has illuminated how hard it can be to live up to these kinds of expectations, particularly those that creators have for their audience, and even those less extreme than the (facetious) suggestion to enter an expensive, high-tech vessel.

“I think about my own listening habits and how experimental and erratic they are right now, and when I do sit down with podcasts, I feel like that’s the only way that it could possibly be,” Mills says. “However people are listening to podcasts right now is exactly how they should be listening to them.”

For some audio creators, acknowledging the limitations of their surroundings goes further than causing them to loosen their expectations for how their product will be consumed; it helps them create a better product in the first place.

Andrés López Azpiri, a composer and sound designer for NPR’s Spanish-language podcast Radio Ambulante, recalls a message that CEO Carolina Guerrero forwarded to him from a natively English-speaking listener, opining that the background music was too loud. This had the effect of obscuring the dialogue, which the listener had to consciously listen for and digest. Notably, non-native Spanish speakers make up 20 to 25{540ccc4681f92a8237c705b0cdebbb9da373ec200da159e6cc1fd9f393be00be} of Radio Ambulante’s U.S.-based audience. But even if that weren’t true, a lack of clarity could become a challenge for any type of listener.

López Azpiri got to work tweaking the sound quality of the show, experimenting with software plugins and how much time he spent on certain parts of each episode. Then, listening in his downtime gave him added insight for how to improve.

When he consumes podcasts for leisure, he says, he chooses shows that aren’t in his native Spanish, but rather in English or German. And when doing so, he says, “I find that if I leave my cell phone a couple feet away from me, and the interview itself has a lot of dynamic” — meaning change in volume — “I cannot understand.” He might not have been able to fix this in the shows he listened to, but he could prevent what appeared to be similar obstacles within the show he helped produce.

Through “compression,” a common treatment that engineers give to audio after it’s recorded but before it’s published, he drags the loudest and quietest sounds closer to a happy medium (since, if the loud moments are painfully loud, you might turn down the volume and find yourself unable to hear the quiet ones). “If you compress into oblivion, it sounds artificial,” he admits, but if you opt out entirely and leave a huge contrast between loud and quiet, the latter gets lost.

These practices have gone a long way, both in the three years that López Azpiri was working at Radio Ambulante before COVID-19 took hold, when people spent hours of each day in moving vehicles, and now, when people don’t — yet still get distracted.

“I haven’t been the host in a not-pandemic time,” says Radio Drama Revival’s Fernández Collins, yet the calm, focused listening environment that she dreams of for her listeners still doesn’t appear to be embraced. In fact, she says, if people mention their listening environments at all, they often say something like, “I was listening to this episode while I was chopping vegetables.”

As López Azpiri also points out, listeners who are parents “can’t just throw in their earphones and ignore the kids.” He imagines that shows are probably played aloud, perhaps from a phone balanced near the kitchen sink. “Pandemic or not,” he says, “they’re listening in non-ideal environments.”

In his four years with Radio Ambulante, López Azpiri has fine-tuned his curation of the show’s sound. Part of it ensures the content is clear. The other part, building on many years as a musician, ensures the content is compelling. He and another composer now create original music for the show’s episodes, which he is as careful to preserve as he is the speech. In fact, addressing the original listener’s concern about volume didn’t mean just turning down the music; it meant, over time, creating better-fitting, more complementary scores.

“I want to make sure that that seriousness is there,” he says, since that mood tends to be elicited by low bass tones, like in an action movie. And once you’ve added an element like low tones, he advises fellow engineers, make sure your technical choices, like compression, don’t compromise it. “Listen to the content really, really quietly,” he says. “If it doesn’t engage with you, and you don’t find it emotional and exciting, there’s a problem.”

Radio Ambulante has to be comprehensible, but that isn’t the only reason it exists. While López Azpiri’s role is to crisp up the audio itself — in his words, “to make it as dry as possible and as clear as possible” — it’s also, he says, to make the show “as compelling as possible — and as beautiful as possible.”

Listening while making dinner, it turns out, can be just as incompatible with this effect as listening on the train. But many audio creators moonlight as listeners; they may have anticipated, realistically, how their work would ultimately reach you, and they may have planned accordingly.

López Azpiri says that messages from Radio Ambulante listeners still come in, but they sound different now. “If they don’t say anything about the audio, I take it that they’re understanding everything perfectly,” he says. “As long as people write to say they were engaged or they cried or they learned something new,” he says, he’s done his job.