By many measures, Castro might seem modest. The company that makes the freemium podcast-listening app doesn’t disclose usage numbers, stating only that it’s among the top 10 podcast apps for iOS—a market segment dominated by Apple’s bundled Podcasts, which overshadows everything else. Castro might have as many as the low hundreds of thousands of users, the vast majority paying nothing.
But Castro has ambitions. While the podcast industry raked in over $300 million in ads in 2017, that number is estimated at $400 million for 2018, and is expected to cross $600 million by 2020, according to the Interactive Advertising Bureau. That’s a tiny sum relative to newspaper and radio ad sales, which are around $17 billion each. Still, it’s up from nearly nothing in just a few years.
Counting advertising alone misses an important and growing part of podcast revenue, however. Other forms of income come from paid content that’s either charged as a premium or included as part of a subscription. This money is collected by companies as varied as Audible, Spotify, and Stitcher. They don’t disclose details about the income they receive that is solely attributable to selling podcast content, but some observers believe it could reach billions in a few years, based on rapidly growing demand.
Castro would like a piece of that future, and the two-man band that developed the app— Pádraig Ó Cinnéide and Oisín Prendiville—recently sold their creation to Tiny, a Victoria, B.C., venture-capital firm that specializes in boosting the fortunes of generally profitable middle-stage startup firms and app makers. Castro’s developers will remain at work on their creation. Tiny’s cofounder Andrew Wilkinson also founded MetaLab, a user-interface shop that’s designed products and handled launches for firms like Slack, Coinbase, Amazon, and Google.
Castro’s revenue currently comes entirely from users who opt to pay for Castro Plus at $3 a quarter or $9 a year. That in-app upgrade brings a number of minor improvements, including an option to recognize and trim silences in podcasts—offering a minor speedup of listening time—and per-podcast customization of settings, such as how many episodes to retain.
Tiny believes in Castro as a product and thinks it’s positioned to take advantage of burgeoning revenue in a rapidly growing space. “We think Castro is the best [app] we’ve used,” says Wilkinson via email. “So it was a no-brainer to team up with Pádraig and Oisín to help them keep going to be able to capture that opportunity versus continuing the brutal indie app-store grind with limited resources.”
The potential listening market is huge. In 2006, in podcasting’s relative infancy, just 11% of Americans 12 or older had ever listened to even one episode. But by 2018, according to Edison Research, that number had climbed to 44%, or about 124 million people. The firm said 17% of Americans who are 12 and older currently listen to podcasts at least once a week. There’s more growth potential yet to come in other countries. The U.K.’s telecom and broadcast regulator, Ofcom, noted in September 2018 that weekly podcast listeners there had doubled from 7% of people aged 15 and over in 2013 to 11% in 2018–nearly 6 million people–but they still have a long way to go to match the U.S.
Wilkinson notes that Tiny is happy to invest in listeners without obsessing about short-term profits for now, but will seek “revenue over the long term.” The goal is to become one of the “top 10 players.” The landscape of podcast apps, advertising, and consumer-paid content doesn’t make that a boast or implausible. Castro is just the latest in a series of moves on the app side of things, as more companies enter the fray.
Here’s a map of the current rocky terrain.
You can’t talk about the podcast ecosystem without starting with Apple. It’s the 100,000-watt gorilla radio station, if that radio gorilla were benign and shared its bananas by broadcasting any tape sent in.
The Apple Podcasts app ships as part of iOS, and thus is in the hands of a billion or more iPhone and iPad users. By many reports, listeners who use this app form 70% or more of podcast consumers. All other apps, no matter how popular, have just a few percentage points of listeners, at least for podcasts that aren’t behind a walled garden.
Over the years, Apple’s interest in podcasts has ranged from indifferent to supportive. Currently, it’s in a supportive arc. The company doesn’t offer any way for podcasters to charge for content, but it puts them in its directory without charge, and offers a streamlined way for listeners to subscribe to new episodes and select from back catalogs.
Apple shows podcasts in its audio search results with the same level of visibility as music, spoken word, and audiobooks. The Podcasts app is not great, but it’s good enough, and it fits neatly into this ecosystem. The Podcasts app’s existence is certainly a reason podcast listening has grown steadily over several years from a niche audience. (Google released its own Google Podcasts app for Android in June 2018, but it has to be downloaded. Google has no plans for an iOS version, nor is Apple likely to offer anything for Android–although given that it just announced plans to bring formerly proprietary services to LG, Samsung, Sony, and Vizio TV sets, who knows what the future might bring?)
For all the benefits of Apple’s support of podcasts, its dominance has led to an advertising monoculture. Podcasters couldn’t charge directly for episodes or shows, and couldn’t assemble rigorous information about listenership or, critically, how long into a podcast people listened before dropping off or whether they skipped over ads. Advertisers on terrestrial and satellite radio rely in part in listening data collected by Nielsen’s tracking systems.
In this monoculture, “host-read ads” leapt to the fore. A throwback to old-time radio, they involve the podcast’s host reading ad copy–often improvising a bit or speaking from personal experience–in the same natural cadence of the rest of the show.
That sort of read “will not necessarily scale to billions of dollars, we don’t think,” says Erik Diehn, the CEO of Stitcher, an E.W. Scripps company that offers its own subscription-based app, sells ads for podcast creators, and produces its own shows for free and premium distribution. “If you’re Starbucks and want to reach people in five cities for a seven-day holiday promotion,” he said, a host-read ad doesn’t meet the bill. That’s despite the success of Midroll, a pioneering podcast ad-sales firm acquired by Scripps before it bought Stitcher.
These ads also typically benefit advertisers who can measure direct response from codes or URLs read by the hosts. There’s a reason why it seems like direct-to-consumer businesses such as Audible, Casper, Harry’s, MeUndies, and Squarespace are the only companies paying for podcast ads. For programs that don’t have many regular listeners or aren’t focused around a host, these sorts of ads are a poor fit. (Some podcasts rely on voluntary support or a patronage model, but this largely works best for those with low production costs or vast listenership.)
In the last few years, there’s been some shift. A number of companies offer software that can insert audio ads dynamically for each download of an episode. Some podcasts use host-read ads for an initial release, and then mark those sections to deliver dynamic ads when an episode becomes part of the back catalog. Some popular podcasts may have as many cumulative downloads from a back catalog in a month or so as any new episode.
And Apple opened its kimono at least partly. In mid-2017, it began to offer podcast creators access to analytics for their shows, aggregating information from iOS users who opted in at a system-wide level to sending diagnostic and usage information to Apple. Charts and lists reveal overall and per-episode data, such as unique devices and time listened.
But most importantly, Apple displays listenership across the duration of a podcast episode, providing a graphical and quantitative insight into when people stop listening. That lets podcasters know (and prove to advertisers) that listeners keep listening rather than abandoning an episode after a few minutes.
With Apple’s new features, podcasts fed out through a feed, rather than through custom apps, could get unfiltered and direct insight for the first time. There were fears that this knowledge would poison the podcast pool. What if it turned out listeners tune out before the first ad? Those concerns were misplaced. However, people generally listen through episodes at a declining rate that largely matched expectations.
But dynamic advertising and partial metrics don’t offer enough options for every podcast that’s after revenue to thrive and expand. That’s led to a proliferation of podcast apps with specialized features—some aimed at listeners and some for show producer and podcast networks—designed to create listener loyalty, accept payments, and gather metrics outside of Apple’s constraints.
Hungry for data
A few dozen apps, like Apple Podcasts, have no monetization or network affiliation built in. In iOS, that includes Castro, Overcast, Downcast, Pocket Casts, and others. These can use public or private podcast feeds. (A consortium of public-radio groups, including NPR and This American Life, bought Pocket Casts in May 2018.) These apps typically have an in-app purchase, likely a recurring but modest subscription, to unlock certain features, or request a voluntary contribution. But none offer premium content or pass-along payments to podcast creators.
Dozens of other podcast apps and streaming-audio apps that include podcast support have a huge diversity of models. Some, like Spotify, embed ads in third-party podcasts unless someone is a premium Spotify subscriber. Others, like Stitcher, allow subscriptions to a significant percentage of free podcasts, but also offer access to premium programming through a monthly subscription. And NPR One, created by that public-radio network, is a combination of public-radio-only podcasts and streaming audio. iHeartRadio has a similar app that centers around streaming radio from its stations, but also allows podcast subscriptions of programs it produces and others it lists in a directory.
Still others, like the podcast-hosting platform Libsyn, create custom or template-based apps for programs or networks that can optionally collect subscription fees or other revenue.
Many of these apps gather the kind of listening data that advertisers and media-company executives love seeing: granular and comprehensive. It’s used for selling ads, but also analyzing the popularity of programs, especially for subscription-based services.
This tendency to track listeners and their behavior will take a strong tick upwards with NPR’s release of a a podcast measurement standard called Remote Audio Data (RAD). RAD will allow podcast producers to tag their content, dropping markers in audio files at specific time stamps using an existing metadata format. The standard comes with wide support among other app developers, podcast platforms, and both public and commercial radio and podcast networks.
When a listener uses any of these apps with RAD support and the app encounters a marker, it shoots back an anonymized bit of data to an analytics URL that’s part of the marker. RAD will let podcasters, advertisers, and other members of the ecosystem aggregate listening data across a variety of apps into a single dashboard. (Apps will have to work through how they disclose or allow opt-in or opt-out of RAD data collection in a way that conforms with Apple’s and Google’s privacy policies, too.)
That will overcome some of the fragmented podcast app landscape. But Apple hasn’t signed on, and some podcast app developers are opposed to the idea. Marco Arment, whose Overcast app typically has more listeners than any other independent podcast app, has long taken a strong privacy stance. On Deccember 11, he tweeted about his objections to RAD.
Yes. I understand why huge podcast companies want more listener data, but there are zero advantages for listeners or app-makers.
I won’t be supporting any listener-behavior tracking specs in Overcast. Podcasters get enough data from your IP address when you download episodes. https://t.co/mplhnrmCsc
— Marco Arment (@marcoarment) December 11, 2018
The Stitcher example
The focus on listener metrics and ad delivery means that most podcasting companies haven’t delved deeply into other possible forms of revenue. This is partly because the firms generating the most revenue with subscription-only podcast series and premium content for shows that are otherwise freely available share almost no information about revenue attributable to podcasts.
Stitcher is a rare exception, as its parent company, E.W. Scripps, breaks out the business’s revenue, which was $13 million in the most recent quarter, or a 90% year-over-year increase. That combines ad commissions and subscription/premium content. (E.W. Scripps put its podcast ad sales, original series development, apps, and premium services under the Stitcher brand earlier this year, folding in Midroll.)
Stitcher combines a bit of nearly every existing revenue model in one place. It pioneered the “Netflix for podcasts” model, which has gradually become a value-added extra in a number of audio-subscription services, including Audible, Spotify, and Pandora. Listeners to those networks can subscribe to podcasts available for free on the larger internet, but just as Netflix supercharged its subscription growth by producing programming in-house in ever-larger quantities, it seems like that’s the direction for audio-subscription services with podcasts as well.
According to Stitcher CEO Diehn, “There will an increasing volume of content that will not be available via RSS,” the venerable technology used to push out free podcast feeds. Partnering with Marvel, Stitcher just coproduced the first season of a Wolverine series, available initially only to paid subscribers. The cast had fairly well-known TV and movie actors, including Richard Armitage in the lead role. (Stitcher also launched a Conan O’Brien podcast with his production company, but it’s broadly available and free.)
Despite the scope of listenership, none of these models have fully shaken out yet. At some point, people may own enough mattresses and razors, and stop responding to ads. Locked in by premium content, listeners subscribing to one network may not want to subscribe or buy audio from others.
Even with Apple’s domination of mass listenership, the fragmentation of the rest of the market makes it unlikely that podcasts turn into what’s happened in streaming media, with original, subscription-only programs available uniquely at Hulu, Netflix, Amazon, and TV and movie company apps such as CBS All Access and Disney’s Disney+. Netflix and Hulu have scored some remarkable subscription increases, but at some point–as with cable TV bills–people look at what they’re paying each month and make hard decisions.
This gives Castro and a host of other apps with existing user bases a toehold to help explore the future. Tiny’s principals didn’t want to discuss the details of Castro’s future on the record, but they’re eager to be part of riding the podcast wave and shaping its direction. And as Stitcher’s Diehn says, “I don’t think it’s all shaken out yet.”
Listeners currently reside in the catbird seat. Everybody wants to cater to their listening interests, nobody knows what to charge or fully how to collect money from them, and their listening habits remain only partly tracked. Whatever model or models emerge will have to contend with a lifetime’s worth of already-released podcast episodes that hundreds of millions of people have yet to listen to.