What a Podcast Download Actually Counts
A podcast download is a request for at least 60 seconds of an episode file, not a completed listen and not a unique human. Here's what that means when you're checking a show's real audience before you agree to appear.
A podcast download is a request for at least 60 seconds of an episode's audio file, made by something that passed a bot filter, counted once per device in a 24-hour window. That is the IAB standard definition. It is not a completed listen, and it is not confirmation that a real, unique human heard the show. It is a filtered server log entry, and the gap between that and "audience size" matters when you are deciding whether to spend an hour on someone else's show.
What counts as a download under the IAB standard
The Interactive Advertising Bureau's Podcast Measurement Technical Guidelines (IAB Tech Lab, version 2.2, 2024) set the industry definition most reputable hosting platforms build their analytics on. A request counts as a download when three conditions are met: the file request pulls at least 60 seconds of audio (roughly 960,000 bytes at a common 128kbps encoding, per the byte-range math IAB uses), the request doesn't come from a known bot, crawler, or flagged server network like a cloud data center, and repeat requests from the same IP address and device get deduplicated within a rolling 24-hour window so one listener doesn't count twice for the same episode ("What Is an IAB Download?", IAB Certified Download, 2024).
Nothing in that definition confirms a person finished the episode, or even that a person was on the other end of the request at all. It confirms a file was requested past a minimum threshold, from something that looked human enough to clear a filter. That's a useful, standardized measurement. It is not the same claim as "this many people heard this episode."
Why raw numbers run higher than the filtered count
Two things routinely inflate a raw server number before any filtering happens.
Podcast apps pre-cache episodes. A listener's phone can ping an episode file to buffer it for offline playback before the person ever presses play, and depending on the app, that ping can generate a server request that looks like a download if nothing downstream filters it out. Server logs that don't apply IAB-style filtering count every one of those requests as if a person listened ("IAB Podcast Measurement Explained", Castos, 2026).
Bots and crawlers do the same thing from the other direction. Search engine bots, monitoring services, and traffic from known server networks like AWS or DigitalOcean can generate file requests that never involved a listener. IAB-compliant measurement filters these out by checking user agents and known server IP ranges before a request counts (IAB Certified Download, 2024). A hosting dashboard or a self-reported number that skips this filtering is reporting raw server traffic, not an audience figure, and the two can differ by a meaningful margin.
This is why "IAB-certified" is a stronger claim than "IAB-compliant." Compliant means the platform's methodology follows the guidelines. Certified means an independent auditor checked that the implementation actually does what it claims, on an ongoing basis. A host who can produce a certified report is showing you a number someone else verified.
Why a number from day one looks nothing like the 30-day figure
Downloads accumulate. An episode doesn't get all its listens on release day, it keeps picking up requests for weeks. That means the same episode produces very different numbers depending on when you ask, and those numbers are not interchangeable.
| Window | What it captures | Where you'll see it |
|---|---|---|
| 24 to 72 hours | The fastest-moving portion of the audience, people who follow the show closely enough to catch a new episode within a couple of days | Rarely quoted on its own, but often the number a host has handy right after an episode airs |
| 7 day | An early, still-incomplete read on an episode's pace | Occasional pitch or media kit figure |
| 30 day | The industry standard window advertisers and networks use to evaluate a show's reach ([Podcast Growth Benchmarks](https://podcastingauthority.com/podcast-growth-benchmarks), Podcasting Authority, citing Edison Research's Infinite Dial) | Media kits, sponsorship decks, ad rate cards |
| Lifetime | Every download an episode has ever generated, growing indefinitely as older episodes keep getting found | Total show downloads, sometimes used to make a young show look established |
A show that quotes a strong 24-hour number for a brand-new episode isn't lying, but it's showing you the smallest slice of the eventual total. A show that quotes lifetime downloads for the whole feed is showing you a number that includes episodes from years ago and tells you little about what a guest would actually reach this month. Neither is directly comparable to a competitor's 30-day figure, and treating them as equivalent is how a founder ends up overestimating reach.
How to sanity-check a number before you say yes
A download figure in a pitch email or a media kit is a claim, not a source. Ask for one of two things before you take it at face value: a recent IAB-certified download report, or a dated screenshot from the show's hosting platform dashboard showing episode-level numbers, not just a total the host typed out.
Both are reasonable requests and most hosts running a real show can produce one without friction. A host who can't, or who pushes back on a simple ask like that, is telling you something too. Pair that check with a look at the show's actual download threshold for the tier of guest they're pitching to, and compare it against the podcast directory entry for the show if one exists, rather than relying on a single number handed to you in isolation.
None of this means small numbers disqualify a show. A tight, engaged niche audience can matter more than a large, unverifiable one, and this is one input among several when you're weighing audience fit against your goals for the appearance. It just means the number itself deserves the same scrutiny you'd apply to any other unverified claim before you commit your time.
Why this matters before you commit an hour
An interview takes preparation, travel or setup time, and a real conversation. Founders who guest regularly learn to spend that time on shows where the audience is real and the fit is right, not on whichever pitch had the biggest number attached. How Convokast evaluates shows before a client ever approves a target follows the same logic laid out here: verify the number, check the fit, then decide. If you'd rather not run that check yourself on every inbound pitch, get in touch and we'll vet audience fit and show quality before any target reaches your approval list.
Common questions
What actually counts as a podcast download?
Under the IAB Podcast Measurement Technical Guidelines, a download is a request for an episode file where at least 60 seconds of audio is retrieved, from a source that isn't a known bot or crawler, deduplicated so the same IP address and device only count once per episode in a rolling 24-hour window. It measures a filtered server request, not a completed listen and not a confirmed unique person.
Why do podcast download numbers seem inflated?
Raw server logs count every file request, including repeat pings from podcast apps pre-caching an episode before anyone presses play, and traffic from bots and crawlers that were never filtered out. A host quoting an unfiltered server total, rather than an IAB-compliant or IAB-certified figure, is often quoting a number well above what a real audience actually generated.
Why is a show's download number different in the first 24 hours than after 30 days?
Downloads accumulate over time. A number pulled a day or two after an episode drops only reflects the fastest-moving portion of the audience, while the 30-day figure, the industry standard window advertisers use, captures the slower listeners too. Comparing an early number from one show to a 30-day number from another compares two different measurements and produces a misleading gap.
How do I verify a podcast's download numbers before agreeing to appear?
Ask the host for a recent IAB-certified download report or a screenshot from their hosting platform's dashboard, dated and showing the actual episode-level numbers. A number typed into a pitch email with no source behind it is not verifiable and should be treated as a claim, not a fact, until you see where it came from.
Work with us
Want to be the guest, not the reader?
We pitch, book, and prep you for the shows your buyers already listen to.
Free 20-minute call. If we are not a fit, we will say so.