The Discovery Problem

There aren’t too many podcasts. There are too many podcast FEEDS.

Jun 04, 2024

Hi there. It’s been a minute. While I don’t feel overly obliged to explain the long gap between posts, I did want to share this short email exchange with the ever-brilliant Helen Zaltzman, creator and host of The Allusionist.

Mia: I'm finding it hard to follow up my substack series over-enthusiastically called "My Big Idea." What comes after that? My much smaller and less significant idea? More of the same but not as interesting? A treatise on how impossible it is to execute said "Big Idea"? Phew.
Helen: Big ideas are usually a bunch of small ideas mixed with mud and leaves, aren't they? When you look closely?

As someone who loves to play in the dirt, that was all the encouragement I needed to get writing again. The execution is a slow process. I might not ever get there. But every conversation I have, every book/article I read, every podcast I listen to, and every doc I watch brings me one step closer. For now, here’s another provocation.

A brief history

At the top of many of my classes and workshops, I offer a whiz-bang tour of podcast history. I start way back, with Guglielmo Marconi’s very first radio transmission in 1897 and Reginald Fessenden’s first broadcast of voice and music on Christmas Eve, 1906.

But I very quickly fast forward to 1999 and the development of the RSS Feed.

RSS stands for Really Simple Syndication (or Rich Site Summary) and it’s essentially the technology that makes podcasting, as we know it, possible. It allows creators to upload new content whenever they want and have those updates reach anyone, anywhere, who’s paying attention to that feed.

The best part about RSS is that it’s open source. So anyone can use it for free. Back in the day, you could write your own XML code and get your audio up on a simple website or blog where users could find your most up-to-date material. Now of course, there are dozens of ways to get your feed out to listeners. Typically you’ll use a feed-hosting service like Libsyn, Buzzsprout, Acast, or Megaphone. And these companies will distribute your feed to platforms like Apple, Spotify, Overcast, and Pocketcast. Descript has a nice explainer of the whole thing here.

When Serial launched in October 2014 there were somewhere around 150-200K podcast RSS feeds. Many of them were created by hobbyists with little-to-no training in the art of audio storytelling. So when Serial arrived with its public radio-trained host/reporters and superior production value, it wasn’t hard for them to get everyone’s attention. In a bit of synergistic timing, October 2014 was the same month Apple started preloading the podcast app onto every iPhone. So it was easier than ever to subscribe to Serial’s feed.

By 2016, when we launched Revisionist History, the number of podcast feeds had more than doubled, to somewhere around 500K. But the number of podcast listeners had also gone up significantly. With the combined engine of a stellar marketing team and Malcolm’s existing fans, we were able to build a substantial audience for that show. “The Golden Age of Podcasts” had begun. Between 2016 to 2019, I helped launch half a dozen new shows that you could reasonably call “hits.” And this trend was repeating in production houses everywhere.

Chart of the number of podcasts which have grown 781 since 2018 — Chart of podcast growth over time from resound.fm

Fast forward to today, and there are more than 4 MILLION podcast feeds. Granted, only a fraction of those are actively publishing new episodes, but still. It’s a lot to weed through, much less keep up with. Of course there are podcast networks that bundle those feeds in various ways. (How many times have we heard that a new network wants to become the Netflix/HBO/etc of podcasting?) But listeners still tend to find shows based on subject matter or what their friends recommend. They are network agnostic.

I just checked my own podcast player (Overcast) and I’m currently following 205 shows (professional hazard). These shows are all over the map in terms of networks — from indie to iHeart. Clearly I’m not listening to all 205, and many of them are inactive. Inertia keeps me from going through and cleaning up that list. Plus I like to have a record of what I’ve listened to in the past. But when I have such a deep bench of material already in my feeds, why on earth would I go looking for something new? There are only so many shows I can listen to with any regularity. And now with so many of them being “always on,” I’m even less inclined to seek out new feeds.

And what of all those inactive feeds? Not just the ones in my podcast player, but the millions of feeds just sitting there on Apple and Spotify? Sure — some of them are probably practice feeds or one-offs. But a good portion of them are shows that had, at one point, engaged listeners.

When you let a feed sit idle, you are essentially leaving money on the table. There must be a better way to organize and curate podcast RSS feeds so listeners can more easily find them and creators can more easily monetize them over time.

Over the last few months I’ve been laying out some ideas for a cooperatively-run podcast network where creators can make a sustainable living and shows are monetized through a grassroots salesforce.

Part One reimagines how podcast sales might work differently – based on my experience selling Cutco knives as a teenager.

Part Two looks at the cost of podcast production in the wake of the downfall of The Messenger.

Part Three questions the value of the CPM as a metric for determining the true value of a podcast / podcast audience.

If you’ve read those you’ll know that the type of podcast we’re most at risk of losing in the current market (and the one I love the most) is the long-form narrative.

A look at how long-form narrative feeds work

Let’s say a long-form narrative audio-storytelling team does everything right. They spend a year or so making 10 excellent episodes. They market the show well and they build an audience of people who share the show with their friends and talk about it on social media. They see week-on-week growth. Maybe they even land some solid advertising deals. Everyone is happy. Then what happens? The run ends. It takes another year at least to make a second season. Those advertisers turn their attention to the next big thing. The audience moves on. And meanwhile, that feed just sits there.

Executives know that this is a terrible plan. It’s why they’re now pushing so hard for “always on” weekly shows. In the case of long-form narratives, they attempt to keep the listeners and advertisers engaged by packing the feed with cross promotions and bonus content, which is usually terrible (or at least disappointing compared to the core content). Or worse, a show ends and the company that paid for that content decides they just don’t want to support it anymore. But they don’t want to give up the feed either. So they hold that feed hostage and the content creator can either try to negotiate some ridiculous sum to buy it from the company or they can try to get another company to buy the feed for them. Meanwhile it sits idle. Everyone loses.

But what if there was something equally compelling ready for the feed right away? What if there were a better way to organize all the feeds, so you could listen to multiple shows in one place without it turning into a jumbled mess? And without the feeds being “owned” by a network or other entity who just want to make a profit off the shows under their banner? What if the producers and podcast fans were in charge?

It’s time to reimagine how RSS feeds are used for podcasting

Instead of an individual podcast feed for every show on the planet, what if feeds operated more like radio stations – one trusted stream of content targeted to one type of listener? When I turn on my favorite radio station, I can be pretty sure they’re going to play something I like before too long. So I keep coming back to that station whenever I’m in my car. I might not be fully tuned in every moment and I might scan around a bit. But soon enough they’ll play something that has me singing along, and I’m reminded why it’s my favorite station. I’ll come back to it again and again. Podcast feeds could work like this too.

There are a number of places who are working on some version of this.

Radiotopia Presents (formerly Showcase) and Pineapple Street’s The 11th (now defunct) have both tried a curated feed of sorts. I’m a fan of the material that appeared in both of those feeds. But there are a few problems. First, their content is/was too few and far between. On Radiotopia Presents, months will pass between shows, so every time they have a new series in the feed, they have to work to get my attention all over again. (The 11th was once a month, but still.) It’s not enough to build a habit. Also, the content in both of those feeds is pretty esoteric, geared toward a more artistic sensibility. That’s fine if it’s mixed with more popular fare, but as an independent feed, it’s not sustainable for creators or funders.
The Slate Daily Feed puts all their content for each day in a single feed, in addition to independent feeds for each show. This is great if you want to listen to everything Slate does in one convenient location. But the content varies wildly from news and politics to parenting and pop culture. I love Slate but I’m not interested in all those subjects. Similarly, WNYC has the NYC Now feed – updated three times a day with breaking news and other headlines. This is a good idea in theory, but when one of my former colleagues made a brilliant, award-winning, five-part investigative series in partnership with WNYC, they put the show in this feed where it got completely buried. The reporter created her own Spotify playlist of the five episodes where listeners could more easily access the series. WNYC did make the series available as a stand-alone tab on their website, but who listens to podcasts on a website??
Ben Riskin and Bianca Grimshaw of Room Tone have created an interesting workaround with their curated “Selects” channel on Apple Podcasts. If you subscribe to their channel ($2.99/month or $29.99/year) you get access to a series of shows, each populated with content on a theme. For example, “Memoirs,” curated by the legendary producer Dmae Roberts. Or “Baby It’s Cole Outside,” – all pieces by This American Life stalwart Sean Cole. The curators offer commentary in the show notes and Room Tone pays a licensing fee for use of the audio, so something gets kicked back to the creators. I like that this puts the curation in the hands of audio tastemakers; if I like them and their work I’m likely to take their recommendations seriously. I also like that once I’m subscribed to the channel I get access to whatever shows are in there. I can tune in to the channel and see “what’s on” similarly to how I use Netflix. But I’m paying for something I could otherwise find for free if I tried hard, the channel is exclusive to Apple (ie: no access for Androids), and the shows only update about once a month. Again, not terribly habit forming. Plus Apple gets a cut of any revenue. I’ll stay off my soapbox here.

A number of organizations are addressing the feed/discovery problem by creating listening apps.

I love what Hark Audio is doing with their Hark Daily Feed: “Hark editors handpick amazing podcast moments and curate them into ‘Harklists’ - podcast mixtapes that bring you serendipity, wonder and new perspectives each day.” In this version of curated feeds you actually hear the voices of the curators telling you why they chose their selections. And Hark repackages the lists in all kinds of creative ways. Plus the app is free. But they focus on 3-6 minute clips, and I often want to hear full episodes. (You can get to full episodes and even import your favorite podcasts into the Hark app, but that’s not really what they’re going for.)

The New York Times recently launched an audio app where you can hear all NYT audio content (and some of what used to be on Audm, RIP) including the Daily, This American Life, Serial, Hard Fork, Ezra Klein, and their new interview show. They have a ton of great content, but I don’t want to listen to the NYT or even NYT-adjacent shows all day every day. I like to change it up.

Similarly, the flagship NPR app has been around for a long time. In 2019, NPR launched the NPR One app with this description: “Catch up on news, or discover a new podcast picked just for you. Share what you like and skip what you don't.” Notably, they would drop shows from other networks into their feed – something that was negotiated with the NPR marketing team. Non-NPR shows that were featured on the NPR One app would see a notable bump in their download numbers. But as good as this was for “discoverability” it wasn’t so good for NPR. They folded the NPR One app at the end of 2023, focusing all their efforts on their flagship app. I love NPR, but I don’t want to listen to NPR shows 24/7. Plus, using the NPR or NYT app (or even Hark Audio) requires a brand loyalty that feels limiting, and not in the spirit of open RSS feeds that kicked off this industry to begin with.

How might podcast RSS feeds be networked to make a better experience for both listeners and creators?

As I mentioned in my last post about the death of the CPM, I’ve been doing a lot of brainstorming recently with Kristen Hayford. She’s a veteran marketing professional who’s been working with The Kids Should See This, a membership-supported website offering curated educational content for kids, helping parents “Avoid 'the Wild West' of YouTube algorithms.”

Kristen and I meet up on zoom every other week or so and share things we’ve read or heard related to this shift in the media industry. We write things up and then we poke holes in each other‘s ideas. I have a similar relationships with Sara McCrea, an early-career producer who checks all of my assumptions and makes me think differently about my work. I’m so grateful to both of them for all their time. If you’re a freelancer, I highly recommend finding a buddy like this to engage with regularly.

Kristen‘s basic idea is a public network of curated feeds.

Here’s how it works

Very similar to Room Tone’s Selects, each feed is curated by an influencer — essentially a podcast super fan. The influencer acts like a DJ — recommending their favorite stories or shows.

So let’s say I’m the influencer and the curator of the busy working mom feed. I’ll use my own habits to determine what goes in the feed. First thing in the morning on weekdays, I’ll run something news-related. Then I won’t have anything else in that feed for a while. (I’m working!) Much later in the day, I’ll run something a little lighter — an interview show maybe — to listen to while I’m driving the kids around, shopping, or cooking. And maybe in the evening I’ll run my favorite kids show to play during wind-down time.

In the middle of the week, say every Wednesday, I’ll run a wellness show — like a little pick-me-up for midweek. And on Fridays, it’ll be all comedy. (Think about TAL’s yearly Poultry Slam. Sure, you could listen to it on demand any time of year, but you really want to listen to it around Thanksgiving. It’s curated for you that way.)

On the weekend I’ll run my narrative shows when I have a little more time to really listen. If something notable is happening one week, or if there’s breaking news, I might pull something special into the feed and explain why it’s there.

The feed is personalized, specific, and flexible.

There are other feeds, too. Maybe there’s a feed specifically for millennials, curated by a millennial. A feed specifically for a certain racial demographic, curated by someone IN that demographic. A feed for students or musicians or trans people or athletes. Each feed is curated by that demographic, for that demographic.

Could this curation be done by an AI algorithm the way Spotify’s music DJ feature works? Probably. But for this:

The influencer also acts as a salesperson

(This is the grassroots salesforce.)

As the host of the busy working mom feed, I’ll read all the ads and cross promos. I’ll talk directly to my listeners. I’ll explain to them why I’m playing what I’m playing and how I feel about it. And yes, I’ll try to sell my listeners some stuff. But it’s stuff that will be totally relevant to working moms. I’ll have a consistent and ongoing relationship with that audience: I’ll engage with them through my social media channels, I’ll invite them to relevant live events, I’ll know them and they’ll know me.

I’ll also have direct relationships with the advertisers. I could invite them onto my feed and talk to them about their products and why the advertiser thinks they’re a good fit for the audience I know so well. I can maintain complete separation between the content creators and the ad sales. It’s similar to how public radio underwriting works, but without the dispassionate voice. I can even run pledge drives – encouraging individuals/members to support the feed in various ways. Listeners and advertisers get the personalized recommendations that have been proven to work so well for selling stuff. And creators can get the editorial distance they require along with the ad sales support they need to monetize their work. I’ll say more about this in the next post.

OK so what now?

Most importantly, we need technology to allow us to create a feed of feeds (a super feed!) that’s easy to search and navigate. I don’t think creators should have to give up their existing RSS/audience for this vision. As I’ve written about here, advertising alone will never support a sustainable living for content creators. So they should be able to keep what they’ve built to date and still participate in this experiment. There must be a way to run a single RSS feed in multiple channels… to “borrow” an existing RSS feed and port it into this curated feed… allowing the listener to keep their existing fans and adding them to the larger pool of listeners. Any coders out there want to show me how this could work?
We need clearer metrics about listener behavior. The folks at Bumper are doing some wonderful work here and are sharing their data publicly with their Podcast X-Ray tool.

We need podcast super fans who can serve as influencers. Folks from various demographics who want to search the podcast landscape for shows that will delight and sustain their listening audience.

We need collective-minded, independent content creators and storytellers who are open to sharing their work in this way without the certainty of minimum guarantees, and the willingness to collectively share whatever revenue can be generated from these super-feeds.

And finally, we need advertisers to pay higher CPMs, knowing they’re going to get better engagement with their audience.

For next time

How do we do that? Kristen Hayford says we should be charging thousand dollar CPMs. Or better yet, $1 per listener. (For context, right now it works out to about $.23 per listener.) Or even better than THAT? Let’s value listeners’ attention, not just their download.

Thank you for reading Freelance Cafe. This post is public so feel free to share it.

This episode of Search Engine: How do we survive the Media Apocalypse with Ezra Klein
This episode of Bubble Trouble: The Post-Bubble World of Podcasting with James Cridland and Jonas Woost
This Digiday article: Inside The New York Times’ plans to correlate attention levels to other metrics
Edison Research’s 2024 Podcast Consumer Report with an emphasis on audience engagement
Spotify’s 2024 Trends Report with a focus on Podcast Influencers

Freelance Cafe