A Little Thinking On Big Podcast Stats
As a general rule; don’t stress out over your podcast’s stats. Your podcast will probably get the audience it deserves. And when you start looking closely, your numbers start making a lot less sense.
Which podcast listening app is responsible for most of your downloads? No, you don't need to check. Assuming your show is trending like just about every other podcast, the answer is Apple Podcasts, right?
One of the things I lament most about podcasting is our lack of a common language around stats and performance. Sure, we use the same words to talk about stats and performance, but those words we use often radically different things. So I return to the topic of stats from time to time to offer some clarity. Or, as with today’s episode, to add more mud to the already murky waters.
Apple Podcasts Has The Biggest Presence In My Stats (And Probably Yours)
For the better part of 15 years now, Apple Podcasts (was: iTunes) has been the biggest app used to download podcast episodes across the board. In aggregate, they remain the leader. Thanks to aggressive moves by other big-money and big-user-base apps, the gap between Apple and it’s rivals is shrinking. But they still lead.
To quantify that for my show, I looked at all aggregate downloads for all episodes going back to July 6th, the start of Season Three. True to form, Apple Podcasts is the app most often used to download my episodes. Specifically, 33.3% of all downloads since that day were attributed to Apple Podcasts.
If that number looks a little small - Apple is routinely touted as having over 50% market share - I remind you that my podcast is different than the run-of-the-mill podcast. More to the point, my audience is different, being made up of almost exclusively working podcasters. Many working podcasters are also power podcast consumers, making them less likely to use consumer-grade listening apps like Apple Podcasts. (Though Apple Podcasts remains my personal listening app of choice.)
Podcast Nerds In Strong Second Place
Examining the same time-frame, I see that 21% of downloads were attributed to Overcast, an iOS-only app that is held up as the gold standard for power listeners of podcasts. It’s my wife’s app of choice and the one we use together when we’re driving in the car listening to podcasts we only listen to when we’re together. True love right there, baby.
If that number looks large to you - Overcast is often reported as the top “indie” podcast listening app, but usually with single-digit market share - I remind you again that my audience over-indexes on podcast-savviness. So I’m not at all surprised that I have 3x the percentage of Overcast users than an “average” podcast.
Bringing Up The Rear
At 10.3%, downloads attributed to Spotify are - pardon the pun - spot-on with the aggregate marketshare usually attributed to Spotify. The remaining 35% of downloads reflect the intentionally-diverse distribution methodology I’ve established, covering smaller podcast listening apps and browser-based requests. All normal stuff.
Until I start looking at individual episode downloads, that is.
Not-Quite Quantum-Level Weirdness Exposed
That the percentages shift drastically when drilling-down to individual episodes didn’t surprise me. Individuals buck averages all the time. What did surprise me, however, was the consistency at which the anomaly occurred.
The hosting company for Podcast Pontifications is Captivate.fm, of which I am an advisory board member. As enabled in most quality podcast hosting companies, I have the option to drill-down and just look at the downloads of any single episode, effectively ignoring the myriad other episodes downloaded in a given time frame.
I started with yesterday’s episode. Normally, a single day (more like 20 hours) isn’t enough to capture a good picture of the apps used to download an episode. But daily shows have different usage trends than shows of less-frequent release cycles, so it’s valid in my case.
I was surprised to not see Apple Podcasts at the top of the list. I was more surprised that only 23% of downloads for yesterday’s episode were attributed to Apple. Which app was in the #1 spot for yesterday’s episode? Overcast, collecting almost 40% of all downloads of yesterday’s episode.
Weird, right? No, as it turns out, not weird at all?
I examined the prior day’s episode and saw almost the same result. Overcast on top (though down to 33%) and Apple trailing at still less than 25%. The day before that continues shows Overcast in the lead, followed by Apple, albeit with a slightly narrowing gap between.
And the day before that. And the day before that. Rinse and repeat, and it’s the same story with each episode. I have to go back a full 10 episodes - nearly three weeks - before Apple Podcasts finally pulled alongside Overcast for downloads of a single episode.
Phantom Spotify Downloads?
Remember earlier when I said that Spotify has been responsible for over 10% of my downloads since Season Three started? Well, I can’t see any appreciable Spotify download activity on the 10 most recent episodes of my show. It’s just not on the list. So whatever Spotify users are hitting, it’s not my new stuff.
Different Strokes for Different Apps
At the risk of offending my friends, Steve, Ted, Zach, and others who work on Apple Podcasts, I worry the inherent behavior of apps is so vastly different that it’s impossible to evaluate them on an even playing field.
Yes, we have IAB 2.0 guidelines that hosting companies follow and are supposed to eliminate confusions like this as they impart a standard for counting downloads. But people smarter than me have looked at the guidelines and say there is too much room left to interpretation.
My friend and fellow working podcaster JD Sutter offered this explanation:
It sounds like most of your daily "grab the show as soon as it releases" listeners are on Overcast. And then the more casual listener uses Apple Podcasts and then they grab several episodes at a time.
He may be right. I know I’ve seen spikes of downloads where one or two “people” using Apple Podcasts will download all 300 episodes in my show’s RSS feed. Is that enough to throw off my aggregate stats? Maybe. Or maybe it’s just weird.
To me, it’s more evidence that the things we think of as universal truths are anything but. It shows the weakness of making blanket assumptions about overall behavior and assuming those same behaviors take place on a very small scale.
Stats are weird, right?
Perhaps you know someone in your podcasting clique who is either stats-philic or stats-phobic who might either get a terrified by or a kick from this episode? Send it to them, along with a personal note of why they should listen. Your personal recommendation goes a long way and really helps this show grow.
And if you like the crazy, sometimes rambling ideas I come up with four days a week, please go to BuyMeACoffee.com/EvoTerra buy me a virtual coffee or twelve to show your support.
That's it for today, and I don’t publish episodes on Friday, because I agree with Andrew Yang on many things.
But I shall be back on Monday with yet another Podcast Pontifications.