Dynamic ad insertion has genuine costs, and most of them land on the listener rather than the publisher. That asymmetry is the problem: the person who benefits from inserting more ads is not the person who has to sit through them.

We sell dynamic ad insertion. We also think shows that adopt it without a policy make their podcasts measurably worse, so this page is the case against — followed by the mitigations that actually work.

1. It removes the friction that used to limit ad load

With baked-in ads, every additional break costs you an edit, a re-export and a re-upload. That friction was doing quiet work: it made you ask whether the fourth ad break was worth the effort.

Dynamic insertion removes the cost entirely. Adding another marker takes seconds and applies everywhere at once. Nothing in the system pushes back, and the incentive runs one way — more slots, more inventory, more revenue — right up until listeners start leaving, which shows up in your numbers months after the decision that caused it.

The mitigation: decide your ad load once, write it down, and treat it as a constraint rather than a default. One mid-roll under 45 minutes, two up to 75, three at the absolute outside. Revisit it deliberately, not opportunistically. See pre-roll vs. mid-roll vs. post-roll for where the slots should sit.

2. You can hear a bad splice

An inserted ad is a different audio file joined to yours at playback. When the two do not match, listeners notice — not consciously as “that was dynamically inserted”, but as a jolt.

Three things cause it:

  • Loudness mismatch. The ad is noticeably louder or quieter than your show. This is the most common and most irritating version, especially for anyone listening in headphones or a car.
  • Room and codec mismatch. A spot recorded in a different space, at a different bitrate, drops in with an audibly different texture.
  • Hard cuts. No breathing space on either side, so the ad begins mid-sentence-feeling and ends abruptly back into your content.

The mitigation: normalise ad audio to the same loudness target as your episodes, ask sponsors for a file that meets your spec rather than accepting whatever they send, and top and tail every ad with a short silence so the join has somewhere to land. If you read the ad yourself in your own studio, most of this disappears.

3. The ad no longer fits the show

Podcast advertising works because of intimacy. The host is a familiar voice, and the recommendation sounds like a recommendation rather than a commercial break.

Dynamic insertion makes it trivially easy to drop in a spot that has none of that — a pre-produced ad, recorded by a stranger, for a product that has nothing to do with your show or audience. It plays perfectly and it is worth much less, because it is a radio ad in a medium that does not reward radio ads.

The mitigation: prefer host-read creative even when it is inserted dynamically. The two are not opposites — you can record the ad yourself and still insert it dynamically, which gets you the endorsement and the ability to change it later. That combination is the best of both and is what most shows should be doing. Programmatic vs. host-read covers the trade-off in full.

4. Repetition inside a single episode

The complaint listeners voice most often is not about ads existing. It is about hearing the same ad three times in forty minutes.

This happens when one campaign has enough budget to fill every slot and nothing prevents it. Each individual placement is correct; the aggregate experience is grating.

The mitigation: do not let one campaign occupy every slot in an episode. If you only have one advertiser, run them in one slot and fill the others with your own promos, or leave them empty. An empty slot costs you a little money. Three repetitions of the same spot costs you listeners.

5. Measurement gets more complicated

Because the file is assembled per request, the relationship between “an episode” and “a file” stops being one-to-one. Episode duration can vary between listeners. Byte-range requests, CDN caching and partial downloads all interact with ad delivery in ways that make reconciliation harder than it looks.

This rarely matters to a small show and matters a great deal once a sponsor starts auditing delivery against their own numbers.

The mitigation: be conservative and consistent in what you report, use IAB-compliant measurement, and never present a number you cannot explain the derivation of. A sponsor who catches a discrepancy will not ask for a correction — they will not renew.

6. Re-listening is a different experience

Someone returning to a favourite episode hears different ads than the first time. Most listeners never think about this. A minority find it faintly unsettling — the episode is supposed to be a fixed object, and it turns out to be slightly liquid.

There is no real mitigation. It is inherent to the model. It is a small cost and worth naming honestly rather than pretending it does not exist.

So should you use it?

For most shows, yes — but as a tool with a policy, not as an upgrade.

Dynamic insertion is clearly right when: you have an evergreen back catalogue that still gets downloads, you want campaigns that can end without re-editing, or you want to promote your own things across your archive.

Baked-in may genuinely be better when: you are making a short-run narrative series with the sponsor woven into the script, your show is small enough that one flat-fee sponsor covers it, or the ad is a genuine creative collaboration that belongs in the recording.

The full comparison is in baked-in vs. dynamic ads.