PodSchool Podcast | What’s the difference between podcast downloads and plays?

Hello. Welcome to the show.

A couple of questions on the same thing I have received, so I thought I'd mush them all into one episode. Today Rosanna asked, what is the difference between downloads and plays? And Donna asked, how can you know downloads versus whether people listen to the whole episode?

The magic of podcast analytics. It is unfortunately not as transparent as we would hope.

However, it is pretty good, and a lot of people whinge about podcast analytics, but there's a lot of people who spend a lot of money on outdoor advertising. And do you know if anybody looked at that or do you just know how many people might have driven by?

I'm sure there's probably some way to digitally measure how many cars went by, but do you know who looked at it? I don't know. You know, podcasting has pretty good analytics compared to some of the ways that advertisers buy things.

And it's usually the advertisers, you know, that complain about these kinds of things. Understandably. I get it, you need to show ROI. It's very important.

But at the same time, I think that while there is always room for improvement, we do have some better metrics than some of the other mediums out there. However, having said that, there are some limitations to how we look at things.

Now, the first thing I would say is that a stream doesn't really exist in podcasting.

It feels like you are streaming a podcast the way that you listen to it, because, you know, if you go into your podcast app, you press play, you don't go in and physically download the file and listen to it. But what is happening behind the scenes is that that file is being downloaded.

So a download in and of itself is just when a podcast episode is downloaded. And that can happen without anybody playing. Some podcast apps, if you are subscribed to a show, will automatically download episodes of that show.

This is what got the industry into such a pickle recently with the Apple iOS 17 update. You might have heard that spoken about if you've listened to any podcasting industry things or read any podcast industry news. That was basically Apple updating their software because traditionally they used to automatically download, if you were subscribed to a show, quite a significant amount of that show's back catalogue. Now, those weren't real listens. Nobody was listening to an episode of your show from 2019, even though in the back end of your podcast host it said that they were. That was likely a podcast app like Apple automatically downloading that episode. But it counts as a download. And that is what the industry goes by in terms of measurement.

Now in the iOS 17 update, they decided to turn that off, but they also decided not to tell anybody about it. So we woke up one sunny, I don't know, September morning, I think it was September or something, and all of a sudden everybody's downloads had fallen off a cliff. And by cliff, I'm talking like 30%. You know, it wasn't like two downloads, it was a significant amount.

And unfortunately, a lot of forecasting, at least on networks, is based on, "Hey, here's where we were at that year and that's what we plan." And so a lot of networks found that they were under-delivering in terms of what they'd promised advertisers and they were significantly off their targets. If you are an at-home podcaster, that's devastating in a way because, you know, you're looking at every download and every download counts. If you're a network, you've got a lot of obligations to advertisers, download goals, all of those kinds of things. So those kinds of things happening overnight can really stuff with you. And I can tell you it did stuff with us and the industry as a whole.

So that is what a download is in terms of it happening automatically. That still happens on some apps. It doesn't happen to the extent that it did in terms of that update.

It was a good thing because there's no point in having a bunch of downloads that aren't real. It was good. The main issue that the industry had with it was that it just happened like a little surprise. We got a little gift one morning and it was a poo in a bag.

So, you know, the upshot of that is that we now have a much cleaner system in the back end. You're not downloading multiple episodes. It's still not perfect. There are still auto-downloads that happen. And so yes, there are some shows that get downloads that might not be listened to.

And you still can't tell if somebody has actually listened to an episode if it's downloaded. But it's better than it was. The play thing, when you are pressing play, that is like a user-generated download.

So it is still downloading the file, but it has been downloaded because somebody pressed play. So it feels like a stream to you, but it's actually still downloading in the backend.

So we term, like, a listen or a play as when somebody has actually physically listened. In apps like Spotify and Apple, where they have first-party data, where people are logged in, they can tell what those people are doing. There is much more information about, has this person actually listened? How much have they actually listened to?

That is much harder to get from your podcast host, where these platforms don't pass that data through.

So you can often go into your podcast platforms like Spotify and Apple and go into their podcaster dashboards and actually see some better analytics than you can get in your podcast host. And if you have, like, 80% of your audience on Spotify, that can be a really decent measure of your audience.

When you just go into Spotify and look at those analytics and look at your consumption data and how many people are listening to your episodes, because that is actually the majority of your listeners are on that platform. Where it becomes difficult is when you're kind of going in and getting that consumption data but only 7% of your audience listen on Spotify.

So it's quite hard to extrapolate that out to your entire audience, you know. But yes, that is a download. It can happen automatically or it can happen via somebody pressing play.

And a stream doesn't really exist because even if you are pressing play and having a streaming experience, you are still actually downloading a file.

Listeners is another sort of triangulated measure that isn't 100% accurate because platforms, at least podcast hosts, are trying to measure that on the basis of IP address.

But if you are listening to a show, when you leave the house and you catch a bus and you ping 50 different cell towers, and then you get to work and you get on the work Wi-Fi…well, you might have pinged as five different people, when in reality you were just one person listening on multiple different IP addresses.

So that's not a really accurate kind of measurement either. There are a lot of platforms that are coming out now that are trying to fix this sort of problem of attribution and, did somebody actually listen?

And there's so much work going into the space. This is often what happens, this wasn't a problem at all, right, before advertisers came into the space.

And then all of a sudden advertisers want answers, understandably, because they're dealing with big budgets. And they're like, "I'm not going to spend my money again with you if I don't know that it was worth spending my money with you." This is where, kind of, effectiveness studies and doing some work at the end of a campaign can be really helpful.

You can survey your audience and see if that advertising made an impact. But still, the majority of people just want to know: did a lot of people listen to this?

And so there is a lot of work being done to try and get a more accurate measure of that. But at the moment it's, you know, sort of accurate-ish.

You know, you can tell that a file was downloaded, but can I tell you 100% that somebody listened to it? No, I can't.

So that's a bit of a bummer, because if you have a hundred downloads on your episode, that doesn't mean that 100 people listened to it or that, you know, even 50 people listened to it twice. That might mean that, you know, there were 20 automatic downloads and 80 actual listens of your show but you just can't tell that data.

But like I said, if you go into some of those apps that have first-party data, that have logged-in users that can track what their listeners are doing, you can get some really useful information and then just use the information that you have to build a picture of what's actually happening on your podcast.

Ultimately, I think the way to go for any independent podcaster is to just look at, is your audience growing? Is it going in the right direction?

That is the thing to focus on as opposed to round numbers. And ideally you just want to make sure that your show is constantly growing if you can make that happen.

Hopefully that has helped to answer some questions or potentially made you even more confused than you were to begin with. But that is the punish of podcast analytics. Sorry about that.

If you would like a little help with your podcast, then make sure you head to my online podcasting course, podschool.com.au. You can join the waitlist there if it's waitlisted. If it’s open, you can jump on in. Depends what time of year it is, but check it out.

I've got all of the details for what's included in the course there at the website. And if you have a question, don't forget to head to the description of the episode, click on the link, and submit it.

All right, see you next week. That's all for today.