Do Podcasts Publish Transcripts?

The same checks, wildly different odds

Every reportable platform in our corpus, ordered by the share of its shows that publish a machine-readable transcript. These are the same feeds behind every other figure on this site, split one more way.

Buzzsprout publishes 21 of its 35 shows with a transcript. Acast, ART19 and Simplecast publish 66 shows between them and not one of those feeds carries a transcript a machine can find. Nothing about those two groups of shows is different in kind: they are drawn from the same topics, scored on the same checks, and both ends contain shows that visibly book outside guests.

The mechanism is not mysterious and it is worth naming, because it tells you what a host could even do about it. Transcription is a feature some publishing platforms generate and declare for every episode automatically, sometimes only above a paid tier, and others do not offer at all. A host on a platform that does not is not withholding anything. They would have to transcribe the episode themselves and hand-edit their feed, and essentially nobody does.

Why a guest should care about this at all

An hour of your best material, published as audio, is invisible to text search. Nothing in the recording is indexable: not the point you made, not your name, not the company you run. The show's episode page carries a title and a paragraph of notes, and that is the whole of what a search engine has to work with.

A transcript changes that, and it is the only thing that does. It puts the full conversation into text on a page the show already owns and already gets crawled, with your name attached to whatever you actually said. That is a durable, searchable record of your expertise that outlives the week the episode aired.

Set it beside the other payoff guesting pages sell you. We measured the show-notes link separately and the honest answer there was that a guest episode carries barely more outbound links than the same show's other episodes. A transcript is a bigger prize than the link and is discussed roughly never, and unlike the link you cannot ask for it politely and change the outcome.

⚠️ Do not overread it in the other direction either. 18.5% is what is declared in the FEED. A show can publish a transcript on its website and tell no machine about it, and this study counts that show as a zero. What the number bounds precisely is reachability: how much of the podcast corpus hands a transcript to anything automated, which is the version that decides whether the text is findable rather than merely typed.

How to check one show in ten seconds

This is the rare check that does not need a tool, because the answer is published in the feed itself and visible in ordinary apps.

• Open the show in a podcast app that supports transcripts and look at a recent episode. If a transcript tab appears, the feed declares one and every future episode of that show almost certainly will too. This is the fastest check and it is the same signal we measured.
• If you want to be sure, open the show's RSS feed in a browser and search the page for `podcast:transcript`. A hit on a recent item is the exact fact this study counted.
• Check the show's own episode page as well. A transcript published there and not declared in the feed still helps you, and it is the case our number cannot see.
• Do the check before the pitch, not after, and do not weight it heavily. A show with a large, well-matched audience and no transcript beats a show with a transcript and the wrong listeners every time.

And if the show does not publish one, ask for the thing that is actually in the host's gift: an episode page that names you, a sentence of bio, and the one link you want. Those are written by hand either way, so a host can say yes to them. A transcript is a setting on software they may not be paying for.

How we measured it, and what would break it

Every figure above comes from a pass over the public RSS feeds behind our own directory on September 16, 2026. For each show we read the recent items and recorded whether ANY of them carried a <podcast:transcript> element, and separately which software the feed's <generator> names. A show counts as publishing transcripts when at least one recent episode declares one.

384 scored shows joined to a platform record and every rate above is over that denominator. 1 scored show had no platform record and is excluded from every rate rather than counted as a zero. 260 of them sit on a named platform; 103 feeds carried no <generator> at all and are reported as their own bucket rather than folded into a platform they might not be on.

Three things would change the answer. <generator> names the software that BUILT the feed, which for these platforms is the host that serves it, but it is never a claim about what a show pays or who owns it, and a show that migrated can carry a stale name for one build. A platform that starts declaring transcripts by default would move its row quickly, so this table is a reading of September 16, 2026 and not a permanent fact. And the tag itself is under-adopted relative to the practice, so every rate here is a floor on how many shows have a transcript somewhere and an accurate count of how many hand one to a machine.

The corpus is the same one behind every data page here: 408 shows, 385 with a feed we could read, gathered between July 16, 2026 and September 16, 2026. The rollup is a single pure module with its own test suite, and no figure on this page is typed by hand.

Sources and benchmark date

Every figure above is derived at build time from our own read of the public RSS feeds behind 408 podcasts. The transcript and platform figures were read on September 16, 2026; the corpus figures come from a separate read of the same feeds between July 16, 2026 and September 16, 2026. Method, limitations and the raw data are public:

• Which Podcasts Actually Book Guests? (our study)

  The full analysis behind the corpus: which shows publish visible evidence they book guests, how to reach them, and the full dataset under CC BY 4.0.

• Do Podcasts Link to Their Guests?

  The other durable artefact of an appearance, measured the same way: how many more outbound links a guest episode carries than the same show's own episodes.

• How to Promote Your Podcast Guest Appearance

  What to do with the episode once it airs, which is the half of the payoff you control rather than the half the show's software decides.

• How to Contact a Podcast Host

  The other finding from this same platform split: whether a show publishes a door a stranger can knock on is also substantially a property of its publishing software.

Frequently asked questions