My friend, mentor (the goddamn whippersnapper), colleague, and former boss Marco Arment wrote three posts about producing podcasts and the value of podcast networks over the last month, two in the last day, that I think warrant a response that is milder than you might imagine. He and I have gone hammer and tongs on Twitter about this at times, but his latest posts come closer to what I believe the statistics show it's all about. (His posts are about Mule Radio shrinking [May 29], "Podcast Networks Are The Wrong Model" [June 22], and "The Elephant In the Podcast Studio" [June 23].)
Podcast networks rose, as Marco notes, when things were hard. He made two main points yesterday related to blogging that he says are now true for podcasting:
Writers don’t need blog networks to be successful today because of two major shifts since blogging began:
The tools required to host and publish are now widely available and free (or very cheap).
Services now exist for publishers to make money without requiring them to be a member of a network.
(What he doesn't precisely mention, and I'll talk about it a bit much later, is that very few podcasts reach a level at which being in a network is an option. Chris Gale, who produces a podcast, explains that quite well in a blog post today.)
I've been part of two podcast networks in different, probably unique relationships because of the point in their development at which I joined, as well as because I've always owned my show.
Let me talk about those points in turn, and then move on to subject matter and advertising.
Means of Production
Every single part of podcasting except the actual speaking part was a pain in the ass or expensive for a long time.
Skype remains the worst best way to record podcasts with one or more remote guests, which is the case in, I wager, most podcasts outside of radio networks or publication-produced ones. Dan Benjamin has written over the years about how in order to achieve quality on Skype, he built a system that committed one Mac to each remote guest running its own copy of Skype! That was mixed in the studio to the host and back to guests.
This sounds like crazy anal behavior one level, unless you record from Skype regularly. Then it starts to sound sadly reasonable. Dan was building a network, and this reduced so many variables that it was a perfectly sensible investment. He made what was an unreliable crappy-sounding experience for most of us into something good and repeatable. (I believe that Leo Laporte's separation of remote participants is similar for his audio/video network. Mule Radio had a studio for its in-house recordings, but wasn't involved in recording the initial audio for many of its programs, such as The Talk Show.)
(Skype hasn't gotten any better, but the "double-ender" recording method has become more sensible as audio tools have gotten cheaper and better. With double-ending, each party (two or more) on a Skype or other kind of call record their own isolated mic stream into a local file. We do this for New Disruptors, and my audio engineer mixes those into a track. It can sound almost radio-studio quality when all goes well. And we record Skype as local/remote isolated tracks as a backup, too, using Audio Hijack Pro. Jason Snell just had 11 people on a recent The Incomparable call! More about The Incomparable later.)
Bandwidth was certainly a cost issue a few years ago and remains so unless you hook into the right place. Libsyn is dedicated to podcast hosting and SoundCloud has a long-running beta alongside its bigger interest in supporting musicians. Squarespace has an excellent combination of web hosting and podcast features. I use Squarespace to host my personal blog, The New Disruptors podcast, and The Periodicalist podcast, and host all my audio files on SoundCloud. Libsyn has fees in some cases above monthly rates (for some kinds of organizations, for podcasts that have advertisers, etc.), but all three offer essentially unmetered, unlimited bandwidth for a low monthly subscription price.
These firms' business models let me pay them a flat rate and then consume $200 of Amazon S3 bandwidth — where most of these companies turn to cloud storage — because most customers only use a couple bucks' worth, if not. (Marco, Casey Liss, and John Siracusa's ATP and John Gruber's The Talk Show each eat about $2,000 of monthly S3 fees.) There are other cheap options, too. I have virtual private servers (VPSes), virtualized instances of servers running on high-end hardware, at Linode. The cheapest option it offers at $10 a month includes two terabytes of data transfer.
So sophisticated web hosting is easy and cheap: Squarespace's unlimited bandwidth plan is $20 per month, and it's simple to find other similar offers elsewhere. Audio file hosting is essentially free or very cheap, or you could DIY and run a VPS.
Maybe equally importantly and less obviously, audio tools started to dramatically improve a few years ago, and there are many good options, along with Adobe's rental option with Creative Cloud. GarageBand (Mac) and Audacity (many platforms) are free; other tools are modestly priced, like Amadeus Pro. They're all much easier to learn to use for basic editing and many have good options to clean up spoken audio, too. (Check out Andy Affleck's newly revised Take Control of Podcasting on the Mac ebook.)
I've seen what Marco is talking about here happen directly. (I first told this story in a Squarespace sponsorship spot last spring when they advertised on New Disruptors.) My friend Swoozy had an idea for a podcast and picked my brain by Twitter and then on the phone for maybe 30 minutes. She had no previous podcast experience and no real audio-editing background.
After we talked, she set up a Squarespace account, and filled in the details to create an iTunes-listed podcast. Later the same day the podcast showed up in the iTunes directory. Swoozy recorded an episode, used Garageband to edit it, posted the installment via Squarespace and, voila, she was a podcaster. It cost her nearly nothing except time, and it sounded pretty good. She rapidly improved with the editing over a few episodes — and grew to hate Garageband like the rest of us. (Like many people, real life's demands intruded on side projects, and she put the podcast on hiatus, but plans to resume it eventually with a co-host.)
The Network Effect or Lack Thereof
I've been part of two networks with the same show: Mule Radio and Boing Boing's podcast family have both hosted The New Disruptors, although I have always owned the show outright and maintained my own website for it to ensure I could post other items, full show notes, maintain archives, and be portable if I chose to be. I've also talked to many, many other podcasters privately about traffic, networks, ad dollars, and the rest, but I can't reveal any of that, and I would hate to tell you to "trust me"; this is my experience calibrated against what others have found, too.
Despite any public pissy-ness between Mike Monteiro and me about other matters (mostly our personality mismatch), he, Jim Ray, Caleb Sexton, Angela Kilduff, and David McCreath (to mention just a few) were instrumental in encouraging me to create the show and fostering its ongoing production. A combination of moral support, managing the hosting and bandwidth and content-management system, and, initially, editing the show gave me the room to turn it into something I'm really proud of. Episode 81 goes up this week.
And, of course, they got their piece of the action, as they should. I had no dispute and still don't about what their percentage was. I don't want to disclose details, because I came on early, and I don't know if my deal (which changed over time) was better or worse than anyone else's. I did have substantial ad sales thanks to my friend Lex Friedman (more about him in a moment, too) during the spring and summer of 2013.
Did Mule build the audience for the show? It's hard to tell. Certainly, I received attention from being part of the network and when there were somewhat fewer tech podcasts in late 2012 than there are now, this led to more people trying me out. But I believe it was the guests I had on the show and the personal networking I do that led to most of the solid base of listeners. Each guest would bring via his or her or their network more first-time listeners, and that has helped sustain the program, to be sure. (I picked a hard show format, too: every week is different! Which means listeners pick and choose, as opposed to shows that focus on the same subject or have the same people each week, even if there is an occasional guest.)
I felt like I'd hit a threshold in listeners at about 5,000 to 8,000 regular downloads. And, months before, I'd taken over audio production with my brother in law, who has 20 years of audio engineering under his belt and a day job in the field. We were delivering a distribution-ready item, and the network wasn't delivering value in continuing to help increase the audience size.
Mike wrote in announcing retracting Mule Radio to ad-free, in-house shows, they had hired a full-time, dedicated employee, but "it really wasn’t a real business. It was a side project. A side project that was now playing with real-ish money." Mule's main business is as a design studio. I felt this. They worked hard for me, but it wasn't core to what they do, they weren't building a publicity and marketing arm, and their own in-house ad sales (which were modest when I arrived) came on fully months after I left.
I was the first show for which Lex Friedman sold ads outside of his own program, Unprofessional, and he went on to sell sponsorship across all of Mule Radio. I had purchased The Magazine in June 2013, and wanted to bring everything under that umbrella, hoping for my own personal network effect, and it didn't make sense to have the podcast in someone else's space. I needed Lex for ad sales and also had to conserve as much expense as possible to run The Magazine.
(However, because I am not an asshole — or at least don't think I am — I didn't leave quickly. I knew many thousands of ad dollars were booked in June and July, and I set a date two months later (August 1), to make sure Mule got its piece of the healthiest period in sponsorship for the show so that they reaped more of a reward of their initial investment of support and time. I used that period to research and set everything up for the transition. And, as I noted then, Mule was a class act: they supported me in the move and helped ensure my RSS feed was redirected to my website so I wouldn't lose subscribers.)
For me, that's the sole remaining measure of a network: can it deliver more listeners to some effect, which is typically the reach required to increase sponsorship value and thus revenue.
A few months after I left Mule, Mark Frauenfelder of Boing Boing, who had appeared on an early episode of the show, asked if I wanted to be part of their expanding network. I agreed, and again kept my independence and my own feed; they were more like a distributor than a network, and I appreciated having more reach. I saw an immediate increase in downloads: 15,000 for the first episode that was hosted there (about a movie about Dr. Demento); that episode now has a cumulative 21,000 downloads.
From about October to May, I was also distributed on Boing Boing, and again don't want to disclose my deal with them as it's private, and the dollars aren't precisely relevant. But what was interesting to me on the ad side was the response from sponsors: despite downloads having leapt to 12,000 to 15,000 downloads over eight weeks at Boing Boing for most episodes compared to about 8,000 to 10,000 on my own after leaving Mule, sponsors were clearly not seeing an impact proportionate to the increase. People got my podcast through the main Boing Boing feed, but it wasn't a good fit and they were likely hardly listening to it. (Boing Boing has a pretty amazing set of podcasts across a lot of creative, geeky, brainiac, and sci-fi/fantasy topics that you should absolutely check out.)
I went fully independent again a few weeks ago, again with no hard feelings: a fit is a fit or not. The number of downloads of episodes had already declined substantially at Boing Boing (to 7,000 to 8,000 per episode after a couple of months), and I thought, once again, better to be fully my own animal than to hook up with a network. Current listenership is about 4,000 in the first few days after an episode posts, and tracking towards 8,000 over eight weeks, the time period generally used to calculate ad rates.
The Quantified Joys of Lex (Friedman)
Can everyone go it alone as Marco suggests? No, but the bar is so much lower than it used to be, as he describes. The early rise of podcast networks can easily be attributed to the complexity and expense. Marco, John Siracusa, and many others joined up with networks where a host handled the interview because they weren't experts or specialists yet.
As time passes, none of us needs to be as much of an expert — note my friend Swoozy, certainly a technically adept person, but not a programmer, and the ease with which she launched her show. But we still have to be willing to take on those tasks. (I custom tweak my RSS feed, for instance, because the automatically generated ones at Squarespace and SoundCloud don't include everything I need or in the right format.)
What we get into isn't really about independence at this stage in podcast evolution. It's about money and percentages. Podcast networks as presently constructed and — to my knowledge about the standard arrangements of either ownership or revenue sharing they engage in — simply want far too much relative to the value that they add for shows that a creator could produce without network help.
Podcast networks, including the tech ones I'm familiar with and those like Earwolf, an exceedingly popular comedy network that pupped off The Midroll, which sells my ads, are constructed too much like old terrestrial radio production companies, and not enough like the loosely federated nature of Internet association in which the means and methods of production and distribution have become simple and cheap.
My area of interest at The New Disruptors centers on new ways to fund, manufacture (goods) or produce (digital), and distribute. If your podcast doesn't require a lot of equipment and expense, and you can handle the production, distribution is a snap. If you goal is to make money, you need bring in much less, perhaps by a factor of two or three, than in any podcast network to achieve the same net results. And you own the upside, if and when that occurs, and anything else you want to do with the program: T-shirts, conferences, partnerships, and the like.
This idea dies hard. Alex Blumberg, one of the folks behind Planet Money and executive producer of This American Life just announced the formation of a for-profit podcast network. He will raise a bunch of money, but the creative folks will be partners or contractors, not independent owners:
Blumberg aims to give podcast creators partial ownership of their shows, possibly via individual contracts, while serving in an “editor-developer” role himself.
That seems ridiculous to me in 2014, but Blumberg is trying to create radio-style shows in podcast format in which the overhead for writing, reporting (and thus salaries or fees plus travel expenses), and production will be quite high, and require a significant investment and significant risk. If they're in the mold of the two other shows he's involved with, the iceberg's top is the podcast; the part underneath that one never hears "on the air" consumes a lot of cash.
For me, the value lies in advertising networks, which is apparent in that most podcast networks have split off their advertising portions and new ad-only networks have formed. Earwolf pupped The Midroll, which purchased Lex's Podlexing (named by yours truly; you're welcome), and brought Lex in as an executive. Dan Benjamin runs Archer Avenue to sell ads against shows in and out of the 5by5 network. Dave Wiskus's Standard Broadcast launched with The Talk Show, Unprofessional (his co-hosted show that itself launched Lex's ad career), and, surprise, Accidental Tech Podcast!
These ad networks take significant but highly varying percentages to handle the part that is still tedious and hard: finding advertisers, dealing with legal stuff, signing contracts, shmoozing, making phone calls (horrible), billing, and the nasty occasional part of collection, which reportedly does not involve the removal or rearrangement of kneecaps.
They take a much smaller portion than a podcast network, because they don't deal with production or typical handle marketing to or growing an audience. Their part of the value chain is well defined, hard, and useful.
Now a small sidebar. When Marco says anyone can do this, it's true. Can anyone sell ads and make good money from a podcast? No. It takes an audience and the right skills and persistence. Building an audience is tough. Marco notes that the Neutral car podcast that he, John, and Casey Liss launched as an experiment only hit about 1,000 regular listeners after 10,000 downloads of the first episode. ATP has climbed way up to 75,000, and it sells sponsorships at $3,500 each, three per show — $10,500 per episode, or $40,000 to $50,000 gross per month. They have a dedicated, highly professional audience that buys stuff. Very, very few podcasts gross anything like that amount. (Some public-radio shows have hundreds of thousands of downloads per episode.)
New Disruptors, over its history, has grossed from about $0 to $5,000 a month, with about $1,500 to $2,000 seemingly stabilizing as about the most sustainable amount, including Patreon patronage, The Midroll sales, and the Cards Against Humanity underwritten indie ads. (My brother in law and I put in about six to eight hours an episode, or roughly 25 to 40 hours a month collectively.)
You need several thousand listeners and to be in the right market for most ad networks to be interested in you, and you have to be able to deliver results for your advertisers. There's no way to do that except only pick advertisers that you think align with your audience, do good ad reads with good copy, and hope listeners find the products and services appealing. Those are all uncontrollable factors. Advertisers typically want results; a few act as true sponsors, trying to make great things happen by providing underwriting without any particular return in mind. MailChimp underwrote a pilot season of six episodes of The Periodicalist, for instance.
At times, Marco has implied or stated outright that people could sell ads themselves, as he has done at times, but the time and effort involved in learning to sell ads and doing so effectively is going to be outweighed by the potential return for most podcasts. Neutral would have been a poor investment for Marco of time to sell sponsorships against; I imagine (he hasn't said explicitly) that the combination of a lack of potential revenue and the notion of having a relatively small amount of interest led to that.
Not all shows need to sell ads, though, which is the broader point, and relatively few people make a large part or their entire living from podcasting outside of folks who come from celebrity or mass-media culture, such as Adam Carolla and the like. Podcasting can be an important and interesting component in a career, a great side project that's fulfilling, or something important for a business getting the word out.
An Incomparable Value
Is there any sense to being part of a podcast network, then? Is it just taking money you don't need to give it (in the most horrible construction I could put on it) or providing expertise that people mostly no longer need?
There's definitely a value to be had, but as with so much of the Internet, it comes from loose affiliation and general independence (a federation of anarchies) rather than from strong affiliation and lack of ownership.
I'd point to Jason Snell's The Incomparable Radio Network that he's been building slowly for fun over years. The flagship show, on which I'm a frequently appearing panelist, will shortly distribute its 200th episode. Our "bonus tracks" of outtakes developed a following and have their own separate feed now. More recently, after some Incomparable D&D episodes, Total Party Kill appeared, in which participants play through a game.
Teevee is the new flashcast format for talking about TV programs right after they air, that any of us can host at will. And more recently, Not Playing with Lex and Dan (two other regular Incomparable panelists) migrated from Boing Boing, and Scott McNulty launched Random Trek, in which he talks about a randomly selected episode across all Star Trek series with a guest.
Jason started The Incomparable as a fun thing to do with friends in his spare time, and it's become a well listened-to program with a really wonderful fan base, with whom we all regularly interact. (The difference between us and fans is minimal, and sometimes fans become panelists!)
Jason is launching shows of his own interest and helping others promote their programs through both the association and a master feed for all shows. In this case, it is abundantly clear that The Incomparable as a podcast has contributed to these shows gaining listeners quickly; the complement of The Incomparable as a network allows fans of one show or panelist to find these other projects easily.
I believe we'll see more of this in which a percentage of revenue isn't what's important; Jason here is sharing back-end infrastructure he built with people he already knows and enjoys. Rather, building an affiliation of interest brings listeners the value of recommendation and one-stop shopping in subject areas or genres. (One could argue that the Podcast Thing recommendation site is a kind of loosely affiliated network of the sort I mean, even though it doesn't involve the participation as such of any listed podcasts.)
I would gladly bring New Disruptors back again into a network in which I felt that I and the other programs could cross pollinate and cross contribute. And I believe such networks are being built.
Podcast networks have a future, but it's in casting off more of the shackles and models that continue to constrain them. And this future vision may be so different from what we currently understand a podcast network to be that Marco may be right: the new model may need a new name that better defines its purposes. A pod of podcasts, not a network.