We need to talk about podcast "listeners"
To the Podcast Technical Working Group at the IAB Tech Lab,
I follow a lot of podcasts. Please don’t tell my friends who make these shows, but… I don’t listen to every episode they publish.
I just checked my phone’s storage to learn I have a whopping 29.34 GB of stored podcast episodes. That’s easily 500+ hours of audio, most of it downloaded automatically.
I don’t think I should count as a “listener” of those episodes unless I actually listen. But according to the IAB’s current podcast measurement guidelines, I’m counted as a daily “listener” even when I don’t listen.
I’m writing today in response to your request for public comment on version 2.2 of the Podcast Technical Measurement Guidelines. I applaud your aim to “provide clarity in the marketplace by developing a common language around measurement.” This is important, necessary work.
The word “listener” is sorely in need of clarity.
As far as I can tell, the podcast industry regularly uses at least three different definitions of “listener.”
Definition 1: Plain English
First, there’s the simple dictionary definition of listener: “a person who listens.” This is simple, non-technical, and straightforward. I can easily picture a concert hall filled with 1,000 listeners. Each of them a person. Each listening. In plain English, by definition, listeners listen.
Definition 2: Platform-specific
Second, we have platform-specific definitions of listener. For example, the Apple Podcasts Connect dashboard defines listeners as:
The number of unique devices that have played more than 0 seconds of an episode.
The Spotify for Podcasters dashboard defines listeners similarly:
The number of unique people on Spotify who played this episode 0 seconds or more.
For both Apple Podcasts and Spotify, listeners are qualified by consumption, and verified by device playback telemetry. When Apple and Spotify talk about listeners, they’re talking about a reasonable proxy for “people who hit play on an episode.”
Crucially, Apple and Spotify’s definitions of listener don’t conflate delivery with consumption. The Apple Podcasts app on my phone might automatically download an episode of a show I follow. But unless I actually hit play, I don’t count as an Apple Podcasts listener for that episode. Again, by definition, listeners listen.
Definition 3: IAB Tech Lab
Finally, we have the Podcast Technical Measurement Guidelines definition of a listener. From the Version 2.2 Public Comment Draft:
Listener: data that represents a single user who downloads content (for immediate or delayed consumption). Listeners are represented by the unique combination of IP address and User Agent as described in section 5, step 3. Listeners must be specified within a stated time frame (day, week, month, etc.).
Using the IAB’s definition, podcast “listeners” don’t need to have actually listened to anything to count as “listeners.” I think that’s confusing.
I run a podcast growth agency called Bumper. Our team works with many different types of podcasters, including large networks and smaller independent creators. All of our clients want to understand the true reach of their podcasts, as measured in people.
In our experience, the podcast industry’s various definitions of “listener” – one that measures delivery, and others that measure consumption – is deeply confusing. Especially when different listener numbers disagree with each other.
For example, in September 2023, when Apple changed automatic download behaviour in Apple Podcasts, many of our clients saw a significant decline in downloads attributed to Apple Podcasts. Here’s what that looked like for one of our client’s shows, starting the week of iOS17’s launch on September 18, 2023:
Because IAB “listeners” are derived from downloads, these same clients saw a significant decline in the “listeners” reported in their hosting provider dashboard:
But when those same clients checked their daily, weekly, and monthly listener counts in the Apple Podcasts Connect dashboard, they saw no such decline. In fact, many of our clients saw an increase in Apple Podcasts listeners over the same period:
One set of numbers decreased. The other set of numbers increased. And both sets of numbers used the same word: “listeners.”
We need better language. I believe the IAB Tech Lab can and should play a role in that.
The IAB’s current definition of “listener” is most closely tied to file transfers, and doesn’t directly measure listening behaviour. I believe this is at odds with a plain English definition of the word listener. In many cases, it can be misleading. When creators see the word “listeners” or “unique listeners” displayed in an IAB-certified hosting provider’s dashboard, it suggests actual listening has occurred, even when it hasn’t.
In everyday conversations with our clients, the team at Bumper regularly has to disambiguate between these multiple definitions of listener. To make sense of it all:
We refer to IAB-defined listeners as “downloaders”
We refer to Apple Podcasts and Spotify listeners as “verified listeners”
I believe “downloader” is an accurate term to describe “a single user who downloads content.” It makes sense when constrained by time (e.g. daily downloaders, weekly downloaders, monthly downloaders, all-time downloaders).
In summary, I believe downloaders download, and listeners listen. Our industry shouldn’t conflate delivery with consumption, and the IAB should consider renaming podcast “listeners.”
Thanks you for your ongoing work, and the opportunity to share this feedback.
Dan Misener
Co-founder
Bumper
Author’s Note: I wrote this after the IAB Tech Lab invited the public to share feedback on version 2.2 of their Podcast Technical Measurement Guidelines. For years, the IAB’s guidelines have been widely used within the podcast industry to define important terms like “download” and “listener.”
Disclosure: In 2017, I was employed by an IAB Tech Lab member company, and participated in the development of the IAB’s Podcast Measurement Technical Guidelines v2.0.