Are Anchor podcasts different?
This article is at least a year old
While Anchor isn’t the only free podcast host, it’s certainly the largest.
In July 2022, the Podcast Index, an open directory of podcasts, said that out of the 4,063,248 open podcasts in their database, 1,811,666 - 44% - were hosted on Anchor.
But, are they different? Some podcast hosts claim that the service is full of dead podcasts, and that 'serious’ podcasters wouldn’t use them. But maybe that’s what podcast hosting companies would like you to think. Anchor has clearly sucked money away from the podcast hosting companies. Even if 20% of Anchor’s shows are committed enough to pay for a $20/month hosting package - a number we’ll come back to - that’s a missing $87 million a year from the paid podcast hosts. So maybe it’s worth their while to talk Anchor down.
I thought it worth taking an impartial look at the data.
As impartial as I can. My editorial disclosures are here; Podnews is part-funded by a variety of supporters and sponsors, many of whom are podcast hosts. Anchor’s parent, Spotify, has been a supporter and sponsor in the past; so have their competitors.
Anchor podcasters make fewer episodes
44% of all podcasts are hosted by Anchor. But, Anchor shows post fewer episodes.
The average number of episodes per podcast2 is 21.7; and the average number of episodes for shows hosted by Anchor is 13.6 (a third fewer).
Podcast Index also automatically removes many inactive shows from Anchor; so the difference may be even higher.
Anchor podcasters are more likely to podfade
Many rival podcast hosts talk about how many shows have 'podfaded’ on Anchor - that is, shows that are no longer being updated.
It’s not a great measure - it’s absolutely valid for a limited-series podcast to be produced, and podcasting is full of those. Are Serial, S-Town or West Cork all worthless now that they are no longer posting new episodes? Clearly not - and, in the case of West Cork at least, it is still earning revenue from dynamically-inserted ads.
It’s also relatively easy for Anchor to appear worse if you measure podfading. An abandoned show on a paid podcast host will cost a monthly fee; so, most podcast hosts will eventually stop paying, and the show will fall off the internet. However, an abandoned show will live on Anchor forever, because it’s free to be there.
But - let’s see what the data says.
Out of the 325,141 podcasts that have posted new episodes in the last 30 days, 94,757 (29%) of these podcasts are hosted by Anchor.
If you remember above - 44% of podcasts in this dataset are published by Anchor; so yes, Anchor podcasters are less active.
Even committed regular podcasts podfade more on Anchor
Given the limitations of the above data, perhaps it’s better to look at committed podcasts: that 20% figure that we guessed at earlier. Perhaps, a committed podcast is one that has posted more than ten episodes; and one that’s regularly produced means it’ll have been updated in the past 30 days.
There are 282,444 committed regular shows: 6.9% of all shows have more than ten episodes, and have been updated in the last 30 days.
There are 83,911 committed regular hosted on Anchor, just 4.6% of all Anchor-hosted shows, which have more than ten episodes and have been updated in the last 30 days.
In other words, yes, Anchor podcasters are more likely to have podfaded: even when you look at committed, regular podcasters.
That said: if only 83,911 shows on Anchor are being run by “committed regular podcasters”, that means the paid podcast-hosting business might only be losing $20m a year because of Anchor. Rather smaller than the $87m we guessed at earlier.
Anchor shows are shorter, too
While we don’t have updated data from the Podcast Index, which doesn’t keep episode durations, in 2020, Podcast Ranking tells us that the average length is 31 minutes. 1
But, the average length of an Anchor-hosted episode was just 21 minutes - a third shorter.
Some of this may be due to Anchor’s mobile tools - unusually for a podcast host, Anchor’s app contains all the tools you need to record a simple podcast just using a phone.
And, perhaps, this is a good thing: in 2018, we learnt that 50% of Americans say that podcasts are too long; and our own advice is to make your podcast as long as it needs to be, but not a second longer.
So is Anchor different?
Yes, Anchor is different: because Anchor is free.
When viewed as a whole, Anchor users post fewer episodes, and episodes are shorter. Anchor users podfade more than other podcasters: even, committed regular podcasts podfade faster on Anchor.
Anchor doesn’t offer many of the services you can get with a paid podcast host. While it doesn’t own your podcast, it does have some clauses in its terms of service that could cause seasoned podcasters to think again. As we note in our article giving advice on choosing a host, Anchor is free, but doesn’t offer everything other podcast hosts do.
Additionally, Anchor seems to host a lot of pirated podcasts which means all Anchor shows shoulder that reputation.
But perhaps there’s a plan for Spotify
It could be the case that Anchor users who become committed, regular podcasters end up switching to Spotify’s enterprise host, Megaphone. Our regularly updated host switching data shows that most shows that get transferred out of Anchor get transferred to Megaphone.
So, perhaps this shows that Anchor is working for Spotify: pulling in new, inexperienced podcasters; helping them learn their trade and become more committed and regular; before on-boarding them to Megaphone.
Where else might this be the plan? Why, Acast, which has a free plan.
So perhaps the reason Anchor is different is because it’s working as designed.
Boring notes:
- Simple averages are different to median episode lengths, which give a better view into how long podcasts are. Daniel Misener looked at this in depth in late 2019.
- Where not otherwise mentioned, our data for this article came from The Podcast Index. The snapshot of the data was updated on Monday July 18 2022, at 07:00:19 GMT (a Unix timestamp of 1658127619). “Last 30 days”, etc, are produced from this timestamp.
