Podcast Downloads vs Streams: Why the Numbers Differ
Downloads, streams and plays are counted differently by every platform. Learn what each number means and which metric to trust for your show.
By Don karlos Fru ยท Published 2026-10-25
Your host says an episode got 4,000 downloads. Spotify for Creators shows 2,600 streams. Apple Podcasts Connect reports a different figure again, and none of the three dashboards agree. Nothing is broken. Each platform counts a different action, with a different threshold, inside its own walls, so the numbers were never going to match.
QUICK ANSWERA download counts file delivery. Under the IAB Tech Lab guidelines (version 2.3, finalized September 29, 2026), at least one minute of content must be successfully delivered for a request to count as a download. A Spotify stream counts only after an episode has been played for longer than 60 seconds. An Apple play counts at 0 seconds or more, and Apple's "listeners" are unique devices, not people. Trust IAB-aligned host downloads for reach, platform numbers for engagement, and never add the three together.
What a download actually counts
Downloads are measured on the server side, from your host's logs, because most podcast apps do not report back exactly what a listener did after the file arrived. The shared rulebook is the IAB Tech Lab Podcast Measurement Technical Guidelines, finalized at version 2.3 on September 29, 2026. Under those guidelines, a download counts when at least one minute of the episode's content has been successfully delivered to the listener's device, and measurement providers filter out bots and duplicate requests. The rules apply to participating certified measurers, which is why two hosts can still produce slightly different totals if one follows the guidelines and the other does not.
Two things follow from that definition. First, a download is a delivery event, not proof of listening: an app can auto-download a new episode overnight for a subscriber who never presses play, and it still counts. Second, a download says nothing about how much was heard. A listener who receives the full file and quits after ten seconds and a listener who plays the episode twice both sit behind the same basic unit.
Editorial image from this article.What Spotify counts: streams, downloads and consumption hours
Spotify measures inside its own player, so its vocabulary is about playback rather than delivery. In Spotify for Creators, a stream is counted once an episode has been played for longer than 60 seconds. Episodes that a listener saves for offline or background listening appear as downloads in Spotify's reporting, which is a different thing from your host's download count. Spotify also reports consumption hours, the total time people spent playing your episodes.
Because a stream requires a minute of actual playback, Spotify's number will usually be lower than your host's download count for the same episode. Unplayed auto-downloads never become streams. The trade-off is scope: Spotify only sees Spotify. Listening that happens in Apple Podcasts or any other app is invisible there.
What Apple counts: plays and listeners
Apple Podcasts Connect uses its own definitions again. According to Apple's stats definitions, a play counts when an episode is played for 0 seconds or more. Pressing play, even briefly, registers. Apple's "listeners" figure is the number of unique devices that played your content, which is not the same as unique people: one person listening on a phone and a tablet is two devices.
That low play threshold explains why Apple's play counts often look generous next to Spotify's streams, while its listener counts can look smaller than your host's downloads. Apple is answering a narrower question: what happened on Apple devices, inside Apple's apps.
Why the same episode produces three different numbers
- Different actions. Delivery (your host), a minute of playback (Spotify) and any playback at all (Apple) are three different events.
- Different coverage. Your host sees requests from every app. Spotify and Apple each see only their own platform.
- Different units. Downloads count requests, streams count qualifying plays, and Apple's listeners count devices.
- Different filtering. Hosts measuring under the IAB guidelines remove bots and duplicates under shared rules. Platform dashboards apply their own.
How to compare your numbers fairly
You cannot merge the dashboards into one true number, but you can stop comparing unlike with unlike:
- Pick one source of truth for reach. Use your host's downloads, and check whether your host measures under the IAB guidelines.
- Compare each platform only with itself over time. Spotify streams this month against Spotify streams last month is a fair comparison. Spotify streams against host downloads is not.
- Use the same window everywhere. Set the same date range in each dashboard before comparing anything.
- Read platform dashboards for engagement, not reach. Consumption hours and how much of each episode gets played tell you whether people stayed.
- Keep a simple monthly log. Record each figure separately, labelled by source, so your trends stay honest.
Which metric should you trust?
Trust each number for the question it actually answers. For "how many times was this episode delivered, under a standard definition," host downloads measured to the IAB guidelines are the most comparable figure across shows and hosts. For "did people actually listen, and for how long," the platform dashboards are better, because they see playback. For "how big is my audience," treat every figure as an estimate with a known bias: downloads lean high because of auto-downloads, and platform numbers lean narrow because each one covers a single app.
The deeper problem is that all of these numbers live on rented ground. Platforms can redefine a metric, move it behind a different dashboard, or change what they share, and your history goes with it. The one audience asset no platform can redefine is your email list. PodSubs helps podcasters turn listeners into subscribers they can reach directly, so growth stops depending on whichever way a dashboard decides to count this quarter. Claim your show and start building the list alongside the downloads.
Frequently asked questions
Why are my downloads higher than my streams?
Because they count different things. A download counts once at least one minute of the episode file has been delivered to a device, and that includes auto-downloads that are never played. A Spotify stream requires the episode to actually be played for longer than 60 seconds, so unplayed downloads never show up as streams.
Does a download mean someone listened to my episode?
Not necessarily. Under the IAB Tech Lab guidelines, a download records successful delivery of at least one minute of content to a device. Delivery is measured from server logs, so it cannot confirm that anyone pressed play or how much they heard.
Which number should I show a sponsor?
Start with your host's downloads, provided your host measures under the IAB Tech Lab guidelines. That is the closest thing the industry has to a shared standard, so it is the figure that can be compared across shows. Use Spotify and Apple numbers as supporting evidence of real listening, not as the headline.
Why does Apple show more plays than listeners?
Because the two metrics count different units. A play registers every time an episode is played for 0 seconds or more, including repeats and resumes on the same device. Listeners count unique devices, so one device can account for many plays but only one listener.
Can I add downloads, streams and plays together for a total audience?
No. The figures overlap, since the same listener can appear in all three, and each one uses a different definition and threshold. Adding them double counts people and mixes delivery with playback. Track each number separately and compare each one only with its own past values.