I always had a hunch that people didn’t really listen to podcasts on YouTube (not in full, anyway). I’ve crunched the numbers based on the last 9 months of data from our customers, and apart from one significant outlier, most auto-posted YouTube videos receive no views.
I want to be clear right from the start, that this is a study on listener behaviour, not a comment on the value of the content being uploaded.
A little background
I wanted to gather data to determine whether or not to continue supporting our YouTube integration, as it’s slow and, due to a lack of clarity and support from Google, unreliable.
I know some of our customers like the integration because it’s a one-click method for automatically creating videos from their content, and gaining potential listeners from a popular platform, but as someone who deals with customers everyday, it’s more often a source of frustration for them than it is a way of finding new listeners, and I’d rather focus on what works well — and what I can control — than what doesn’t.
Over the past 10 months, over 700 videos have been auto-posted to YouTube, from Podiant’s integration. A handful are duplicates, some have been removed by the channel admin or taken down by YouTube, and some I haven’t included in my analysis because they’re not podcast content*.
What’s a view?
Before I get into the details, let’s just clarify what we’re talking about. I’ve gathered “view counts” for each video in the study, which YouTube defines as a play of the video for at least 30 seconds. Given that most podcast episodes are between 30 and 90 minutes, we don’t really have a sense of how much of an episode people have actually “watched”.
To the numbers!
Across the 480 videos I’ve analysed, the average view count for the lifetime of each video is 15.3. Now that’s not nothing, but let’s look at the outlier I mentioned.
It was a true-crime podcast, and one of Podiant’s most listened-to shows. Listeners were, if I recall correctly, directed to the YouTube videos rather than the podcast (which is unusual). It was a real outlier too, with view counts in the thousands.
If we exclude that podcast as an aberration, our average view count drops to 10 , across the life of the video (roughly one view a month). Again, nothing to be sniffed at if you’re a new or niche podcaster and you’re playing a long game. But while there are videos getting double- or even tripe-digit view counts, this is not the norm.
So let’s look at the mode: the number which appears most often in our list of view counts. That number is 0, which means the majority of videos auto-posted to YouTube from Podiant are not played. Again, I’m not judging the content: I’m simply stating that podcast listening doesn’t happen all that much on YouTube, based on our data. Plus, as podcasters, we tend to focus our marketing push on platforms like Apple Podcasts, or increasingly Spotify.
Why does this matter?
I welcome data from other hosting services as it may well contradict mine, but this isn’t an academic paper. I have to make a business decision based on where this small company should spend its limited time. I’m not saying that “no-one listens to podcasts on YouTube”, as that’s clearly not true, however, given that the vast majority of videos have received less than a handful of views, and that the integration fails more often that it works, I can’t justify continuing to support it.
What’s the problem?
Part of the issue is that the YouTube integration is very hard to debug, as it’s something of a black box, and a slow one at that (our kit is optimised for audio, not video). Audio goes in one end, along with an image, and most of the time, video comes out the other. Sometimes video doesn’t come out, and I can’t always know why without spending a day writing more code to investigate. Usually however, a video does come out the other end, but then it gets rejected by the YouTube robots. Sometimes there’s a reason for this, much of the time there isn’t, and when that happens, customers rightly get frustrated, and I’m the one who has to deal with that frustration, not Google.
Options for existing users
Audioship provides what they call “the easiest way to upload your audio to YouTube”, and offers email support for their service. You can also batch-upload your episodes to them, and you get an hour’s upload time for free, then after that it’s only $1 per hour.
Back to the original question
So, is it worth auto-posting your podcast to YouTube? It will depend almost entirely on your audience, and where you expect to find them. But is it worth Podiant investing computational and human resources into providing this service at no extra cost? I’m afraid not. So I’ll be turning off the YouTube integration at the end of October, and removing all reference to it from our marketing literature.
I'm looking forward to focusing my efforts in new areas to help podcasters attract new audiences, and serve listeners better.
- Excluded videos are from one channel, and are simply full-length pieces of music that have evaded YouTube’s Content ID algorithm, but for which we haven’t received copyright notices. As such, they’re not valid podcast content, so have been excluded from the study.