Podcast Sponsorship Rates: What Brands Actually Pay

Podcast sponsorships are priced on CPM — cost per thousand downloads of the episode — with mid-roll host-read spots the most expensive slot and pre-roll the cheapest. As a working range, expect roughly $18–$30 CPM for a 60-second host-read mid-roll, $15–$25 for a pre-roll, and $10–$15 for a post-roll. Those are the ranges direct deals and mid-size networks tend to land in; they are conventions, not measurements, and niche B2B shows break them badly on purpose.

The useful thing about CPM pricing is that it makes a small show affordable and a big show expensive in proportion. The unhelpful thing is that downloads are a weak proxy for listens, and the gap between the two is where most of the disappointment lives.

What the slots cost, and what they're worth

Slot	Position	Typical CPM	Why
Pre-roll	First 1–2 min	$15–$25	Everyone hears it; some are still settling in
Mid-roll	Middle of episode	$18–$30	Listener is committed; highest completion
Post-roll	After the content	$10–$15	Cheapest, and the drop-off is real

Mid-roll costs more because a listener who is fifteen minutes into an hour-long episode is unlikely to leave. That premium is genuine, not a network invention — but it is a 30–50% premium, not a 3x one, and on a small show the difference between slots may be $40.

Ad length matters as much as position. A 60-second read is roughly double a 30-second read on most rate cards, and a 60 is what you want for anything that needs explaining. If your product takes a sentence — a snack, a sock — buy 30s and buy more episodes.

Do the arithmetic before you talk to anyone. A show averaging 20,000 downloads per episode at $25 CPM is $500 for one mid-roll. A show doing 2,000 downloads is $50, which is below most hosts' floor: expect small shows to quote a flat $150–$500 per episode regardless of what CPM implies, because the admin of running an ad doesn't get cheaper as the audience shrinks.

Host-read vs produced, baked-in vs inserted

Two axes, and brands routinely buy the wrong quadrant.

Host-read means the host says it in their own words from your talking points. Produced (or announcer-read) means you supply finished audio. Host-read costs more — sometimes 1.5–2x — and it is worth it, because the whole reason podcast advertising converts is that the host's endorsement transfers. A polished 30-second radio spot dropped into a two-person conversation gets skipped. Buy host-read or don't buy podcasts.

Baked-in means the ad is permanently in the audio file; dynamically inserted (DAI) means it is stitched in at download time and can be swapped or removed later. DAI is how networks sell campaigns by flight dates, and it means your ad also runs on the back catalogue — often a real bonus, since evergreen shows keep getting downloaded for years. Baked-in is common on small independent shows and gives you a permanent spot for a one-time fee. For a limited-time offer, insist on DAI. For an evergreen product, baked-in on a strong back catalogue is quietly one of the better deals in creator marketing.

Buying direct vs going through a network

Networks and marketplaces (the ad sales arms attached to most large shows) will handle trafficking, provide download verification, and enforce delivery. They also carry minimums — commonly $5,000–$25,000 for a campaign — and their rate cards are the top of the range.

Buying direct from a host means emailing them. For shows under about 25,000 downloads an episode, that is usually the only option, and it is where the value is: independent hosts negotiate, throw in a newsletter mention or a YouTube version, and care about a brand that treats them well. Expect to pay 10–30% under network rates for the equivalent audience.

What to ask for before agreeing a price:

1. Downloads per episode, 30-day window, IAB v2 certified — not total downloads, not "monthly listeners", not subscribers. If they can't produce that number, price the deal as if the audience is half what they claim.
2. Audience geography. A show with 60% of its audience outside the countries you ship to is priced on downloads you can't use.
3. Category exclusivity for the flight. Standard on host-read deals and cheap to ask for.
4. A promo code or vanity URL, unique to the show. Not shared across your podcast buys.
5. Ad copy approval, and the recording before it airs if the host is willing — most are.

Measuring it, given that you mostly can't

Podcast attribution is the worst of any creator channel: no click, no pixel on the audio, and a listener who hears you in the car and searches your brand two days later. Three practical approaches, in ascending order of usefulness:

• Vanity URLs and unique promo codes. Undercounts badly — most people don't type a URL from audio — but the direction is real and the code is cheap.
• Post-purchase survey. One question at checkout: "where did you hear about us?" Podcast is the channel this rescues, because it catches the people who searched your brand name instead of typing the code.
• Pixel-based podcast attribution vendors, which match the IP that downloaded the episode to the IP that later hit your site. Directionally useful at scale, noisy on a single show, and worth it once you're spending five figures a month.

Buy a minimum of three to four episodes on the same show before judging it. One-episode tests are how brands conclude podcasts don't work: the first read is the host's first time saying your name, and the audience has heard of you for eleven seconds. Repetition is the entire mechanism.

The operational side is where podcast buying quietly eats time — a spreadsheet of shows, download claims, quoted rates, which host has your copy, which episodes have aired, which code belongs to whom and who has been invoiced. CreatorCast tracks each show as a creator relationship with its deal terms, approvals and payouts on one record, which is the difference between running eight podcast deals and running eight separate email threads from memory. For comparing this against other channels, our guide to measuring influencer marketing ROI covers the attribution trade-offs in full.

Frequently asked questions

How much does it cost to sponsor a small podcast? Most independent shows under 5,000 downloads per episode quote a flat $150–$500 per episode rather than a strict CPM, because the effort of running the ad is fixed. That is often the best value in the medium if the audience is a genuine fit.

Is a host-read ad worth the premium over a produced spot? Yes. The endorsement is what you're buying; a produced spot is just a radio ad in a podcast. If budget forces a choice, buy fewer host-read episodes rather than more produced ones.

What CPM should I negotiate down to? Direct deals with independent shows commonly settle 10–30% below published network rates, especially for multi-episode flights booked in one go. Offering four episodes up front is a stronger lever than haggling on a single spot.

How do I know if a podcast's audience numbers are real? Ask for IAB v2 certified download figures over a 30-day window per episode, plus geographic breakdown. Vague answers, "monthly listeners", or subscriber counts instead of downloads all mean the same thing: price the deal conservatively or walk.