What Share of Podcast Listeners Finish an Episode?

Can this question be answered at all?

Partly, and only by the platforms. As examined elsewhere in this series, a podcast delivered as a file reaches a device that, in the IAB's words, cannot send data about consumption back. Server logs record what was requested, not what was heard, so a publisher working from hosting statistics cannot answer this question about their own show.

The exception is the platform that controls the player. Apple, Spotify and other apps observe playback directly, and can report how much of an episode was actually played because the playing happens inside their software.

This produces an awkward split in the market. The numbers used to price and sell podcast advertising come from server logs, and the numbers that describe whether anyone listened come from platform dashboards that advertisers generally do not see.

So the honest answer to the headline question is that a general industry figure for episode completion does not exist in any verifiable form. What exists are per-show, per-platform measurements held by creators, and a set of circulating benchmarks whose provenance is mostly untraceable.

The split also explains a recurring frustration for creators. A show can watch its measured consumption improve while its download figure stays flat, or the reverse, because the two numbers describe different populations doing different things — one counts devices requesting a file, the other counts people playing it.

What does Apple actually measure?

Three distinct things, each defined precisely, and the precision is what makes them useful.

Plays: “the total number of times people pressed play on your episode.” This is an action taken by a person inside the app, not a file request made by software.

Engaged Listeners: the “total number of people who listened to or watched at least 20 minutes or 40% of an episode.” Two thresholds joined by or, so a short episode qualifies on the percentage and a long one on the elapsed time.

Average Consumption: “a percent average indicating how much of the episode users played per device, based on the listeners' consumption and the duration of the episode.” This is the closest thing to a completion rate that any platform publishes a definition for, and Apple describes the underlying data as aggregated listening and viewing completion rates from unique devices. Apple is explicit about the contrast with the rest of the industry, noting that downloads have traditionally been the key reach metric while its analytics focus on completion instead — and confirming it does not supply download figures at all, directing creators to their hosting provider for those.

How far apart are the two thresholds?

Far enough that they are not comparable, and the gap is the central fact of podcast measurement.

The IAB guideline counts a valid download once the ID3 tag plus roughly one minute of audio has been transferred, justifying that floor on the grounds that anything smaller could not have been listened to. Apple's engagement bar is twenty minutes, or forty per cent of the episode.

On a sixty-minute interview, one threshold is satisfied by transferring about one-sixtieth of the file to a device, and the other requires a third of an hour of actual playback. Both can be described in a deck as a measure of audience.

This is why a show's download figure and its engagement figure can tell entirely different stories, and why comparing two shows on downloads alone compares almost nothing. A programme whose listeners routinely reach twenty minutes and one whose listeners drop at ninety seconds can report the same downloads, and only one of them has an audience in the sense a sponsor is paying for.

Why do the circulating benchmarks disagree?

Because they measure different things and are rarely labelled with which. In preparing this report we found completion figures in wide circulation ranging from around half to above ninety per cent, with several presented as industry averages.

We are not reproducing them as findings, for a specific reason. Almost none arrives with a definition attached. A figure derived from a survey asking listeners whether they finish episodes is a self-report, and self-reports about attention are known to flatter. A figure from a hosting platform's aggregate dashboard measures something real but is specific to that platform's user base and its own definition of completion. A figure from a single network describes that network's shows.

Compounding this, completion depends heavily on episode length, format and genre in ways that make a single industry average close to meaningless. A twelve-minute daily news briefing and a three-hour conversation cannot share a benchmark.

The reliable statement is therefore narrower than a number. Platform-measured consumption exists, is defined, and belongs to the creator. Any figure quoted without naming its source, its definition and its sample should be treated as decoration.

There is also a selection effect worth naming. Completion figures are published when they flatter, by parties with something to sell, and a benchmark assembled from voluntarily disclosed numbers describes the shows willing to disclose rather than the medium.

What should you ask a show before booking?

For the platform numbers, because they exist and the creator has them.

Ask for Average Consumption on recent episodes, and for the Engaged Listeners count rather than downloads. Both are defined by Apple, both are measured from playback rather than inferred from logs, and a show serious about its audience will have them to hand.

Ask where in the episode listeners drop off. Apple describes its analytics as showing engagement patterns including whether listeners skip intros or mid-roll segments, which bears directly on whether a guest segment twenty minutes in will be heard at all.

Ask what share of the audience uses the platform being reported. Apple's figures describe Apple listeners, and a show whose audience is mostly elsewhere is showing you a partial picture — an honest one, but partial.

A show that answers these is measuring its audience. A show that responds only with a download total is quoting the metric that requires the least of its listeners, and that is worth knowing before rather than after.

None of these questions is adversarial, and a working creator will not read them that way. They are the questions someone who intends to appear on the show would naturally ask, and the answers shape what a sensible appearance looks like.

What this means for anyone buying a podcast placement

Stop treating downloads as audience, and stop accepting completion figures without a definition. Those two habits account for most of the disappointment in this market.

Note what Apple's own design tells you. The platform that can see actual playback built its analytics around consumption and does not report downloads at all, directing creators elsewhere for them. The company with the best view of listening declined to build its reporting on the industry's headline number, which is the same signal its chart rankings send by excluding automatic downloads.

Where a placement is a guest appearance, position matters more than any aggregate. A segment placed before the average drop-off point reaches a materially different share of the audience than one placed after it, and that is a question with an answer the creator can look up.

Media Morpheus books podcast appearances. What a booking reliably delivers is an appearance on a real programme and a durable, citable artefact. How much of it anyone hears is a property of the show rather than of the booking, it is measurable by the show, and asking for that measurement before you buy is the single most useful thing a buyer can do.

Methodology and limitations

This report draws on Apple's published listener analytics definitions in its Apple Podcasts for Creators documentation, and on the IAB Tech Lab Podcast Measurement Technical Guidelines version 2.1 for the download threshold used in comparison. Five limits apply. First, no industry-wide completion statistic is reported here, because no figure meeting this series' sourcing standard was located; circulating benchmarks are discussed but deliberately not reproduced as findings. Second, Apple's metrics describe Apple Podcasts listeners only, and other platforms define and report their metrics differently. Third, the 60-minute episode used in the threshold comparison is illustrative; Apple's 'or' condition means the binding threshold changes with episode length. Fourth, completion varies substantially by genre, format and duration, so no single benchmark would be meaningful even if one were published. Fifth, platform documentation changes without notice and this describes the position at the date of publication.

Conclusion

There is no trustworthy industry figure for what share of listeners finish an episode, and the reason is structural rather than accidental: the metric the market trades on is produced by server logs that cannot observe listening, while the measurements that can observe it sit inside platform dashboards belonging to individual creators. The gap between the two standards is stark. Apple counts an engaged listener at twenty minutes or forty per cent of an episode; the IAB counts a download once about sixty seconds of audio has been requested. Both get described as audience. For a buyer the practical move is to ignore aggregates entirely and ask one show for its own measured consumption, which exists, is defined, and is the only version of this number that means anything.

Frequently Asked Questions

Sources

1. official Apple. Listener analytics, Apple Podcasts for Creators, for the definitions of Engaged Listeners, Plays, Followers and Average Consumption.
2. trade body IAB Tech Lab, March 2021. Podcast Measurement Technical Guidelines v2.1, for the valid download threshold used in comparison.