Substack podcasts: what the platform does, and where it stops

Substack hosts your audio file, generates an RSS feed, and distributes that feed to Apple Podcasts, Spotify and anywhere else that accepts one. Episodes sit inside your existing publication, so a subscriber gets the audio in the same email they already open.

Three things follow from that, and they are genuinely good. You keep the subscriber list rather than renting an audience from a podcast host. You can put an episode behind the same paywall as your writing, with no separate product to set up. And there is one publish button for text and audio instead of two systems that have to be kept in step.

For a writer who already has a paying list, that bundle is hard to beat. Nothing below argues you should leave.

Analytics. Substack reports plays and little else. There is no IAB certification, which matters the moment you want to quote a download number to a sponsor, and there is no breakdown of where in an episode people stop listening. If you are trying to work out whether your twelve-minute intro is costing you the audience, the data is not there.

Editing. You can record into Substack or upload a file. What you cannot do is meaningful repair: removing a long pause, cutting a sentence you fluffed, evening out a recording where you drifted off the microphone. In practice most writers record once and publish whatever happened.

Sound quality. There is no noise reduction, no levelling, no loudness normalisation. A recording made at a kitchen table sounds like a recording made at a kitchen table, and next to a professionally produced show in the same podcast app, the difference is obvious in about four seconds.

Transcripts. Substack will generate one, but it is not an editing surface. You cannot change the words and have the audio follow.

It depends entirely on what the audio is for, and the honest answer for a lot of writers is that it does not matter much.

If the audio is a convenience for existing readers, something they put on while walking the dog because they already trust you, record it in Substack and stop reading here. Polish is not the point and nobody is comparing you to anything.

It starts mattering when the audio has to win people who do not already know you. An episode that shows up in a podcast app sits next to shows with producers. It matters when you are charging for the audio specifically, because paid subscribers apply a different standard to a thing they bought. And it matters the moment a sponsor asks for numbers.

The workaround is not to move your podcast off Substack. It is to stop treating Substack as the place where the episode gets made.

Record and produce somewhere built for it, export a finished file, and upload that file to Substack. Substack keeps doing the part it is good at, hosting, distribution, paywall and list. The audio arrives already fixed.

This costs you one extra step per episode and nothing else. You do not lose your subscribers, your paywall, your archive or your feed, because none of those move.

Substack lets you mark any episode free or paid. A paid episode is delivered to paying subscribers through a private RSS feed they can add to their normal podcast app, which is the single most useful thing the platform does for audio and the least well known.

Two practical notes. Put at least some audio outside the paywall, because a podcast app listing is one of the few ways strangers find you, and a feed with nothing playable converts nobody. And be consistent about which episodes are paid, since subscribers who cannot predict what they are buying churn faster than subscribers who get less.