I've started development on the next major revolution of My Podcast Reviews (the leading service to see and share all your global podcast reviews), and in the process of collecting data, I've noticed some interesting correlations between the seriousness of podcasts and the hosting companies they are on. Here's that data.
(I'm writing this while my toddler is sleeping, so attractive charts and more detailed explanations will be missing for now.)
All of the following data is from Apple Podcasts and is updated as of December 1, 2018. I estimate only a 1% or smaller margin of error.
First, how many podcasts are in Apple Podcasts (AP)?
- About 609,000 podcasts listed
- A weekly fluctuation of +/- 5,000–7,000 based on new or removed podcasts (not conclusive on this number yet)
For data on podcast-hosting companies, I looked at only the podcasts' RSS feed URLs. This is not 100% reliable and will, for most podcast-hosting companies, result in under-estimated numbers. This is because not every podcaster chooses to use the feed-publishing tools offered by their hosting company. For example, while Libsyn offers a powerful suite of offsite podcast publishing tools, many podcasters may use Libsyn only for hosting and stats, but generate their RSS feed elsewhere (such as with PowerPress on WordPress). This also doesn't account for feed-mirroring or feed-enhancement services, such as Podcast Mirror, FeedBurner, Podtrac, and such, because such services sometimes obscure the source feed URL. Thus, Blubrry will be noticeably absent from this list because Blubrry's own domains show up rarely in feed URLs, but usually only in the individual episode enclosures.
Now, on to the podcast hosting companies! (Due to the obscurity of some data, this is not a complete list of the top hosting companies, since Blubrry is obviously missing.)
- SoundCloud: 83.3K podcasts (13.7%)
- Anchor: 72.6K podcasts (11.9%)
- FeedBurner (not a podcast host): 49.8K podcasts (8.2%)
- Libsyn/Libsyn Pro: 41.0K podcasts (6.7%)
- podOmatic: 37.6K podcasts (6.2%)
- Podbean: 28.5K podcasts (4.7%)
- Spreaker: 15.2K podcasts (2.5%)
- Buzzsprout: 8.4K podcasts (1.4%)
- AudioBoom: 4.4K podcasts (0.7%)
- SimpleCast: 3.3K podcasts (0.5%)
- Omny Studio: 2.9K podcasts (0.5%)
- Acast: 2.4K podcasts (0.4%)
- Art19: 1.1K podcasts (0.016%)
- Fireside, Podtrac, FeedPress, and Megaphone are all under 1K
- Probably PowerPress (not a podcast host): 28.9K podcasts (4.7%)
Each podcaster can customize the number of episodes they keep available in their podcast's RSS feed. Publishing tools have different defaults, usually 10, 50, 100, or 300. Thus, episode counts are not a foolproof way to determine a show's commitment or longevity. But anything under 10 would be notable as it's below the default limit of probably all publishing tools.
1–10 episodes individually:
- 1 episode: 76.3K podcasts (12.5%)
- 2 episodes: 44.1K podcasts (7.2%)
- 3 episodes: 35.0K podcasts (5.7%)
- 4 episodes: 29.0K podcasts (4.8%)
- 5 episodes: 25.3K podcasts (4.1%)
- 6 episodes: 21.1K podcasts (3.5%)
- 7 episodes: 17.4K podcasts (2.9%)
- 8 episodes: 16.4K podcasts (2.7%)
- 9 episodes: 14.9K podcasts (2.4%)
- 10 episodes: 29.7K podcasts (4.9%)
Note the decline in counts up to 9. But because 10 is a popular limit, there are many more podcasts with that many episodes in their feed. (Remember, there may be more episodes beyond that limit. This is a measurement of only what's in the feeds.)
Now, some groupings:
- 1–9 episodes: 279.4K (45.9%)
- 10–50 episodes: 218.1K podcasts (35.8%)
- 51–100 episodes: 58.9K podcasts (9.7%)
- 101–200 episodes: 25.8K podcasts (4.2%)
- 201–300 episodes: 22.5K podcasts (3.7%)
- 301 or more episodes: 42 podcasts (.0068%)
There are about 4.3 million episodes published in Apple Podcasts. Thus, the average number of episodes per feed is about 7. But again, this is highly affected by self-imposed episode limits in feeds and thus not reflective of the true number of podcast episodes. For example, many highly popular and long-running podcasts like This Week in Tech, TouchArcade, and No Agenda limit their feed to display only the latest 10 episodes (16 currently in No Agenda‘s feed).
“New,” “active,” and “inactive”
As a kid in the '80s and '90s, I remember going to the video rental stores (imagine if Netflix/Hulu/etc were physical stores you had to drive to and pay for each movie you spent hours deciding to watch). There would always be a “new releases” section, and it amazed me how long movies would stay in there. So I've always wondered how long something is truly “new.”
For this post, I'll define such labels as follows:
- New: launched within the past 3 months (not enough data, yet)
- Active: published an episode within the past 3 months (229.8K podcasts)
- Inactive: no published episodes within the past 3 months (372K podcasts, and yes, this may be skewed against some seasonal shows)
As I'm publishing this on December 3, 2018, I used September 1, 2018 as the split date. Any shows with episodes on or after that date I label “active,” and any shows without episodes on or after that date are labeled “inactive.”
Segmenting by podcast hosting/publishing tool
Now this is where it gets really interesting! Let's look at the same kinds of numbers as above, but broken down by podcast hosting or publishing tool.
Percentages in the following will be of the number of podcasts generated by each host/tool.
- 83.3K podcasts
- 33.7K active podcasts (40.5%)
- 49.6K inactive podcasts (59.5%)
- 27.0K with 3 or fewer episodes (32.4%)
- 21.6K with 4–9 episodes (25.9%)
- 23.7K with 10–50 episodes (28.5%)
- 11.1K with 51 or more episodes (13.3%)
- 48.8K point to podcast page on media host (58.6%)
- 72.2K podcasts
- 30.1K active podcasts (41.7%)
- 42.1K inactive podcasts (58.3%)
- 41.6K with 3 or fewer episodes (57.7%)
- 15.3K with 4–9 episodes (21.2%)
- 13.0K with 10–50 episodes (18.0%)
- 2.2K with 51 or more episodes (3.0%)
- 72.1K point to podcast page on media host (99.9%)
- 49.7K podcasts
- 12.9K active podcasts (26.0%)
- 36.8K inactive podcasts (74.0%)
- 8.4K with 3 or fewer episodes (16.9%)
- 9.8K with 4–9 episodes (19.6%)
- 22.8K with 10–50 episodes (45.9%)
- 8.8K with 51 or more episodes (17.6%)
- 126 point to podcast page on media host (0.3%)
- 40.9K podcasts
- 25.3K active podcasts (62.0%)
- 15.5K inactive podcasts (38.0%)
- 3.3K with 3 or fewer episodes (8.1%)
- 6.5K with 4–9 episodes (15.9%)
- 18.2K with 10–50 episodes (44.5%)
- 12.9K with 51 or more episodes (31.6%)
- 18.3K point to podcast page on media host (44.9%)
- 28.3K podcasts
- 15.7K active podcasts (55.3%)
- 12.7K inactive podcasts (44.7%)
- 7.0K with 3 or fewer episodes (24.7%)
- 5.3K with 4–9 episodes (18.6%)
- 11.4K with 10–50 episodes (40.2%)
- 4.7K with 51 or more episodes (16.6%)
- 26.6K point to podcast page on media host (94.0%)
- 15.1K podcasts
- 7.9K active podcasts (52.5%)
- 7.2K inactive podcasts (47.5%)
- 3.1K with 3 or fewer episodes (20.7%)
- 3.6K with 4–9 episodes (24.0%)
- 5.2K with 10–50 episodes (34.4%)
- 3.2K with 51 or more episodes (21.0%)
- 13.6K point to podcast page on media host (89.8%)
- 8.3K podcasts
- 6.4K active podcasts (76.8%)
- 1.9K inactive podcasts (23.2%)
- 1.9K with 3 or fewer episodes (22.7%)
- 1.9K with 4–9 episodes (23.3%)
- 2.9K with 10–50 episodes (35.3%)
- 1.6K with 51 or more episodes (18.8%)
- 565 point to podcast page on media host (6.8%)
- 3.3K podcasts
- 2.4K active podcasts (72.7%)
- 894 inactive podcasts (27.3%)
- 411 with 3 or fewer episodes (12.6%)
- 693 with 4–9 episodes (21.2%)
- 1.5K with 10–50 episodes (44.8%)
- 700 with 51 or more episodes (21.4%)
- 232 point to podcast page on media host (7.1%)
- 2.4K podcasts
- 1.3K active podcasts (52.9%)
- 1.1K inactive podcasts (47.1%)
- 186 with 3 or fewer episodes (7.8%)
- 444 with 4–9 episodes (18.7%)
- 1.0K with 10–50 episodes (43.4%)
- 713 with 51 or more episodes (30.1%)
- 1.1K point to podcast page on media host (48.2%)
- 1.1K podcasts
- 844 active podcasts (80.2%)
- 209 inactive podcasts (19.8%)
- 35 with 3 or fewer episodes (3.3%)
- 114 with 4–9 episodes (10.8%)
- 416 with 10–50 episodes (39.5%)
- 488 with 51 or more episodes (46.3%)
- 831 point to podcast page on media host (78.9%)
- 935 podcasts
- 736 active podcasts (78.7%)
- 199 inactive podcasts (21.3%)
- 41 with 3 or fewer episodes (4.4%)
- 110 with 4–9 episodes (11.8%)
- 315 with 10–50 episodes (33.7%)
- 469 with 51 or more episodes (50.2%)
- 10 point to podcast page on media host (1.1%)
- 618 podcasts
- 465 active podcasts (75.2%)
- 153 inactive podcasts (24.8%)
- 63 with 3 or fewer episodes (10.2%)
- 119 with 4–9 episodes (19.3%)
- 265 with 10–50 episodes (42.9%)
- 171 with 51 or more episodes (27.7%)
- 513 point to podcast page on media host (83.0%)
- 28.9K podcasts
- 14.4K active podcasts (49.8%)
- 14.5K inactive podcasts (50.2%)
- 3.4K with 3 or fewer episodes (11.8%)
- 5.4K with 4–9 episodes (18.8%)
- 15.2K with 10–50 episodes (52.7%)
- 4.8K with 51 or more episodes (16.7%)
- 0 point to podcast page on media host (0.0%)
- 850 podcasts
- 482 active podcasts (56.7%)
- 368 inactive podcasts (43.3%)
- 73 with 3 or fewer episodes (8.6%)
- 231 with 4–9 episodes (27.2%)
- 331 with 10–50 episodes (38.9%)
- 215 with 51 or more episodes (25.3%)
- 8 point to podcast page on media host (0.9%)
- 954 podcasts
- 331 active podcasts (34.7%)
- 623 inactive podcasts (65.3%)
- 97 with 3 or fewer episodes (10.2%)
- 156 with 4–9 episodes (16.4%)
- 345 with 10–50 episodes (36.2%)
- 356 with 51 or more episodes (37.3%)
- 2 point to podcast page on media host (0.2%)
- 2.9K podcasts
- 2.1K active podcasts (72.7%)
- 790 inactive podcasts (27.3%)
- 155 with 3 or fewer episodes (5.4%)
- 324 with 4–9 episodes (11.2%)
- 924 with 10–50 episodes (32.0%)
- 1.5K with 51 or more episodes (51.5%)
- 53 point to podcast page on media host (1.8%)
- 4.4K podcasts
- 2.0K active podcasts (45.9%)
- 2.4K inactive podcasts (54.1%)
- 552 with 3 or fewer episodes (12.5%)
- 790 with 4–9 episodes (17.8%)
- 1.8K with 10–50 episodes (39.6%)
- 1.3K with 51 or more episodes (30.2%)
- 4.0K point to podcast page on media host (91.0%)
- 37.6K podcasts
- 3.8K active podcasts (10.2%)
- 33.7K inactive podcasts (89.8%)
- 15.4K with 3 or fewer episodes (40.9%)
- 11.5K with 4–9 episodes (30.7%)
- 8.9K with 10–50 episodes (23.8%)
- 1.7K with 51 or more episodes (4.6%)
- 37.5K point to podcast page on media host (100.0%)
Data is subject to interpretation, so here's how I interpret this information I've shared.
- Free podcast-hosting companies may boast high numbers of podcasts, but they usually have very low active podcasts. This is probably because “free” asks for very little strategy or commitment.
- Premium podcasting tools seem to show more committed, active users.
- FeedBurner has more inactive users than active, probably because a lot of podcasters who ended their shows have not bothered to migrate to better publishing tools.
- Anchor and SoundCloud are the most popular media hosts, but they also have the highest percentages of inactive podcasts. And a shockingly high number of Anchor users have 3 or fewer episodes, indicating very new shows or very uncommitted podcasters.
- A disappointing majority of podcasts on a majority of hosting providers have the feed pointing back to the podcast webpage on the host. This could be laziness on the podcaster's part to not change it to their own domain, it could be ignorance of the ability (and necessity) to change such setting, it could be that the tool doesn't even allow that setting to be changed, or it could be that the podcaster is intentionally using the media host's website as their podcast's primary website. (But you should still get your own domain and update the website URL in your feed!) The most shocking of these is Anchor-hosted shows where all but 3 were pointing at the podcast's page on Anchor. The only 3 that weren't doing so appear to be migrations from other hosting services and are thus still pointing back to those previous services.
- As could be expected, this data suggests that the kinds of people willing to invest in their podcasting tools are the kinds of people likely to keep the podcast active.
- Inversely, it also appears that providers with extremely low-barrier tools and no costs to the podcaster result in a lot of “flash in the pan” podcasts that publish only a couple episodes and then podfade. Anchor's high number in Apple Podcasts is also likely due to their automated submission of their users' podcasts to Apple, but through Anchor's own account (and not the podcasters').
I have no doubt this data will be interpreted in other ways, and some companies might be unhappy with what the data reveals.
What's your interpretation?
[…] And how about this one from Daniel J. Lewis of The Audacity to Podcast: […]
Great article, Daniel! Many podcasts start, and then podfade a few episodes in. While that may be bad for the reputation of podcasting, it is good to see more people getting their feet wet in this medium.
[…] posted the research to his website, Tuesday, and what really caught our eye was how many podcasters still use Soundcloud as their […]
Thanks for doing the research, Daniel. Insightful.
Thanks for pulling all these stats together. I’m guessing that many people will be surprised by your findings. I’d love to know the date that each of these hosts started offering podcast hosting. I know that both Soundcloud and Anchor shifted at some point and it would be helpful to know the time frames for these numbers.
Budding Podcasters and Podcaster “hopefuls” will find that podcasting is a lot of work (despite the promise of easy podcasting) so they end up quitting! If might be simple to create a few episodes with a free host, but growing an audience and improving the actual substance of the content is easier said than done! With that said, I think that exposing people to podcasting this way is a really good thing for the overall growth the medium because the people that end up podfading also end up with a newfound respect for the what it means to be a podcaster. In the long run these people will become the BEST future podcast listeners ever.
And we can all use more loyal podcast listeners.
Yes. It’s a tough balance between making the tools and process easier (which I’m a huge fan of), and yet also ensuring we’re not pushing people into something they’re not ready for, yet. I don’t think licenses or certificates should be required, but maybe these super-easy tools could have some kind of “staging” or “playground” mode, allowing someone to essentially practice podcasting, but without worrying about putting it out there and giving up.
I have chatted with a lot of podcasters on Anchor and many of them find that it is so easy to start a show, they create 3 or 4 shows and just stick with the one (s) they enjoy the most. When that happens, I think it’s a good thing.
Good for them, yes. But do you think it’s good for those abandoned shows to be publicized or left public if the podcaster was only experimenting?
I don’t see the harm. It’s not like anyone will be able to find them on Apple Podcasts anyway. There are thousands of abandoned podcasts — even those that people continue to pay hosting fees for. Look at all the other mediums that have abandoned shows, like YouTube for example, I think the bigger problem is the way Apple Podcasts charts are being gamed.
[…] Daniel’s awesome post on podcast hosting data: https://theaudacitytopodcast.com/what-new-data-suggests-about-podcast-hosting-customers/ […]
[…] this point, Pandora will not be including everyone who submits — there are over 600,000 podcasts out there! — but the presumption is they will eventually follow Spotify's lead and include everyone who […]
[…] recently written by podcast industry expert, Daniel J. Lewis helps tell the story. In “What New Data Suggests About Podcast-Hosting Customers,” Daniel makes some interesting discoveries and […]
[…] interesting thing to note is from data collected over at TheAudacityPodcast.com. It seems that the people most likely to keep their podcast active are those who were willing to […]
I keep statistics on the ones I listen to, as well. Of the 404 podcasts I subscribe to, 54% are active, 37% are definitely inactive (> 90 days since an update), and another 8% are in that in-between ( 30 < x <= 90 days) where they might be dead but maybe not. (My "x" was 3/3/19 because that's when I did the data collection most recently)
[…] 45.9% of the podcasts in Apple Podcasts had produced just one episode […]
one year later I’m curious if you would do this research once again to have all the new Media Hosts like https://www.podigee.com/en/ or https://theaudacitytopodcast.com/captivate etc. included? (meaning to have an updated base of the Data?)
And maybe to compare it to this research from 2018?
Thanks and have a great time!
Yes! I plan to post future updates. They’ll most likely be on https://mypodcastreviews.com/ going forward, but I may bring the data over to https://theaudacitytopodcast.com/ for more opinion and commentary.
Wow, very very fast reply! Thanks, I’m looking forward to it at https://mypodcastreviews.com/ *thumbs-up*
[…] found that 38% of Libsyn’s podcasts were inactive and 8.1% had three or fewer episodes. This is […]