A baked-in ad is part of the episode’s audio file, so every download of that episode carries the same ad, whether it is played the week the episode comes out or two years later. A dynamically inserted ad is placed into marked ad breaks by an ad server each time the episode is requested, which means it can be targeted by place and date, swapped partway through a campaign, capped, and taken out when the flight ends. Host-reads can be delivered either way. Baked-in suits evergreen offers on shows whose old episodes keep getting played, while dynamic insertion suits dated offers, regional tests and buying a set number of impressions.
How each method works
A baked-in, or integrated, ad is part of the audio file, so the only targeting available is the choice of show and episode, while a dynamically inserted one is picked by an ad server and spliced into designated breaks at the moment the file is requested. That is how the IAB Tech Lab’s Podcast Measurement Technical Guidelines (version 2.2, May 2024) describe the two, and they add that sponsorships which were traditionally host-read and baked in can now be inserted dynamically too. In that case the host records the read once as a separate file, and the server places it into the episodes the campaign has bought. (The current version of the guidelines is 2.3, released in September 2026.)
So choosing between the two is mostly separate from choosing between a host-read and a produced spot. It decides where the ad lives, inside the episode for good or in a slot that is filled again on each download.
How common each one is
Dynamic insertion is by far the more common of the two. IAB’s Podcasting Buyer-Seller Checklist (October 2023) describes it as the dominant form of podcast ad insertion, “at more than 80 percent.” Baked-in reads are still sold where a host prefers to record the ads with the episode, or where a brand wants the read to stay in the back catalog.
What each one costs
On a published rate card, a 30-second baked-in host-read costs about the same per thousand downloads (CPM) as a dynamic episodic placement, a 60-second baked-in read costs more, and a dynamic buy across a show’s whole catalog costs less. The figures below come from the Libsyn Ads rate card (2026 edition), which is a seller’s own published price list.
| Placement on the Libsyn Ads rate card, 2026 | CPM |
|---|---|
| Baked-in host-read, 30 seconds | $18 to $22 |
| Baked-in host-read, 60 seconds | $24 to $26 |
| Dynamic, episodic | $18 to $22 |
| Dynamic, full catalog | $14 to $16 |
| Programmatic | $12 to $15 |
Libsyn arrives at a campaign’s cost by multiplying reach by the CPM and dividing by a thousand, counting reach as the downloads an episode gets in its first 30 days. On those terms a baked-in 60-second read at $25 CPM, on an episode that reaches 20,000 downloads in its first month, costs $500. The listening that arrives after the first month is not counted in that price, which is the main argument for baking in.
Other sellers price differently. Acast’s pricing guide (2026 pricing, as of October 2026) quotes $25 to $40 CPM for host-read sponsorships and adds that each layer of targeting pushes the CPM up. A rate card tells you where a negotiation starts, and the price we pay depends on the show, the placement and the length of the commitment.
The back catalog
A baked-in read keeps playing for as long as the episode is downloaded. That can be good value on shows whose listeners work through old episodes, such as history, education and long-running interview shows, and it becomes a liability when the read carries anything that will date, such as a price, a seasonal offer, a product that may change, or a code that will expire. A baked-in read is a small permanent part of the show, and it should only carry things that will still be true next year.
With dynamic insertion, ads in old episodes can be sold again, replaced or removed, and a full-catalog buy spreads one campaign across a show’s whole archive at a lower CPM, as the Libsyn rates above show. The trade-off is that the read is no longer tied to the episode it sits in, so a host-read recorded once and inserted into many episodes cannot refer to what the host has just been talking about.
Targeting, timing and control
Dynamic insertion gives a buyer controls that a baked-in read does not. A campaign can start and stop on set dates, run in some regions and not others, cap how often one listener hears the ad, and change creative halfway through when a message is not working. A brand can also buy a set number of impressions and know when they have been delivered.
Baked-in reads give up those controls in exchange for permanence and context. The host can tie the read to the episode, everyone who hears that episode hears the read, and on a rate card like Libsyn’s the brand pays for the first month of downloads while the listening continues long after the flight.
How delivery is counted
Both are counted from the server’s record of what was downloaded or delivered, with no proof that anyone pressed play, because podcast apps report so little about playback. The IAB Tech Lab guidelines (May 2024) note that client-confirmed ad play would be the most accurate count, and that it is mostly unavailable. So when a seller reports impressions for a dynamic campaign, or downloads for a baked-in one, both are counts of files delivered, which is worth remembering when comparing podcast numbers with the viewability figures from other media.
Exclusivity
Exclusivity needs particular care with dynamic insertion, because a show can be selling its ad breaks to several buyers at once. The IAB checklist (October 2023) says exclusivity has to be defined campaign by campaign, and gives the example of a brand that has exclusivity for host-read ads while announcer-read ads for competitors still run in back-catalog episodes. We write exclusivity terms into each booking in plain words, covering which formats, which episodes and which dates.
How we choose
We weigh the same five things for each show in a plan.
- Whether anything in the read will date. Prices, seasonal offers and expiring codes point to dynamic insertion.
- How much of the show’s listening comes from its back catalog. If a large share does, a baked-in read may be the better buy, and we ask the show for its download curve before deciding.
- Whether the campaign needs dates, regions or frequency caps, which only dynamic insertion can provide.
- Whether the read gains from being tied to the episode. For hosts who weave the product into the conversation, baked-in keeps that context.
- What the show offers. Some shows only sell one way, and the price difference between the two can settle the question.
A plan can use both, with baked-in reads on the few shows where the back catalog is strong and the offer is evergreen, and dynamic host-reads elsewhere so the test has firm dates and the creative can change between flights.