Help · Guides
Why podcast download numbers differ between hosts
Last updated 14 October 2026
Move a podcast between two hosts and the download counts can change by a third without a single listener coming or going. Nothing happened to your audience; the two hosts count differently. Knowing the rules lets you read your own numbers, compare quotes and talk to sponsors with confidence.
A download is a server event
Apps fetch audio files from the host's servers. A download is counted when a file is served, which is the only point at which the host can see anything. There is no reliable way to know whether the file was played, or for how long; apps keep that to themselves. So every download number is a count of deliveries, filtered by rules that remove the deliveries that clearly were not a person listening.
The IAB rules
The IAB Tech Lab's Podcast Measurement Technical Guidelines, version 2.3, are the rules most of the industry has settled on. Their core points:
- One download per listener per episode per day. The same device and app fetching the same episode several times in a day counts once. Seeking, pausing and resuming produce several requests for one listen.
- A minimum amount of audio. A request that fetches only the first few seconds, as apps do when checking a file or showing a preview, does not count. The guideline uses one minute of audio, in practice about 1 MB.
- Bots, crawlers and known non listeners removed. Directory crawlers, search engines, monitoring tools and the directories' own catalogue checks are excluded by their identifying strings and behaviour.
- Counted at the server, not by scripts. Numbers from the host's own delivery logs, not from a player's reports or estimates.
A host can follow these rules without being certified by the IAB. Certification is a separate paid audit of the implementation; "follows the guidelines" and "IAB certified" are different claims, and an honest host says which one it is making.
Why hosts still differ
Even among hosts that follow the guidelines there is judgement: how a listener is identified (by address and app string, with or without further detail), which crawlers are on the exclusion list, and how partial requests are treated. Hosts that do not follow the guidelines at all count every request, which inflates figures by a large margin: a single app checking an episode's file size becomes a download. If a host's numbers look generous, that is usually why.
How to read your own numbers
- Use the 7 and 30 day windows for trends, and the first 7 days of an episode's life to compare episodes with each other.
- Keep imported history separate from measured history. A figure loaded from a previous host was counted under its rules.
- Quote sponsors a per episode figure over a defined window, measured under IAB rules, and say so. It is the number they are used to.
- Expect a drop when moving to a stricter host, and expect it once. The audience did not change.
What Podmaniac does
Podmaniac counts at delivery on its own servers, one per listener per episode per day, with a 1 MB minimum and crawlers removed, following the IAB v2.3 guidelines; it is not certified. Downloads are shown per day, per episode, by country and by app, with imported history labelled. Details in How we count downloads.
Common questions
What counts as a podcast download?
A delivery of an episode's audio file from the host's server to an app, after filtering: one per listener per episode per day, at least about a minute of audio served, with bots and crawlers removed.
Is a download the same as a listen?
No. A download shows that an app fetched the file, usually because a subscriber's app downloaded a new episode. Whether and how long it was played is known only to the app.
What is a good number of downloads for a podcast?
Numbers vary enormously by topic and country. The useful comparison is your own episodes against each other over the same window, for example the first 7 days after release.
Why did my downloads drop after changing host?
Almost always a difference in counting rules. A host that followed the IAB guidelines less strictly, or not at all, reported higher figures for the same audience.