There are exactly three ways to use podcasts in marketing, and they cost wildly different amounts. You can host your own show, appear as a guest on other people’s shows, or buy advertising. Guesting is cheapest and fastest, advertising is the only one that scales on demand, and hosting your own show is the most expensive and slowest but the only one that builds an asset you control.
The measurement problem is worth stating up front. Podcast delivery is a file download, so there is no click between hearing the ad and arriving on your site. Every attribution method for podcasts is lossy. Plan for that rather than being surprised by it.
The three routes compared
| Route | Cash cost to start | Hours per month | Time to first result | What you own afterward |
|---|---|---|---|---|
| Guesting on other shows | $0 to $300 | 6 to 10 | 6 to 10 weeks from first pitch | Clips, backlinks, relationships |
| Advertising on other shows | $500 per placement and up | 3 to 5 | 2 to 6 weeks from booking | Nothing, you rent attention |
| Hosting your own show | $800 to $4,000 for gear and hosting | 20 to 40 | 6 to 12 months | A back catalog, a subscriber base, endless raw material |
Most teams should start in the top row and stay there for a quarter. Guesting tells you whether podcast audiences respond to your message at all, and it costs almost nothing to find out.
Route one: get booked as a guest
Build the target list first. For each candidate show record the host name, the audience it actually serves, episode length, publishing cadence, whether it books guests at all, the three most recent guests, and one specific episode you have listened to. That last column is the one that gets you booked.
Then pitch the segment, not yourself. Hosts receive a constant stream of messages describing someone’s job title. What they are short of is a specific 20 minute conversation they can imagine recording. Here is the structure that works.
Subject: Segment idea after your episode with [previous guest]
Hi [name],
Your episode with [previous guest] made the case that [specific claim].
I run [role] at [company] and we hit the opposite result when we
[specific situation], which I think would make a good counterpoint.
Three things I could walk through:
1. [Concrete thing with a number in it]
2. [A mistake you made and what it cost]
3. [A process listeners could copy on Monday]
Happy to send a two minute voice memo so you can hear how I sound.
[Name], [link to one relevant page]Send 40 of these over four weeks. A response rate in the low tens of percent is normal for a well targeted list, and a handful of bookings from 40 pitches is a good outcome. Track the list in the same place you track everything else; our guide to building a content calendar works fine for outreach with a status column swapped in.
Route two: buy advertising, but calculate breakeven first
Podcast inventory is sold on CPM, meaning cost per thousand downloads. Before you agree to anything, understand what a download is. The IAB Tech Lab Podcast Measurement Technical Guidelines define the industry standard, and certified hosts count a download only when a threshold portion of the file is delivered, deduplicated per listener per episode within a 24 hour window. Ask any show you are considering whether their numbers are IAB certified. If they are not, treat the figure as a raw server hit count and discount it heavily.
Now the arithmetic. Work backward from what an order is worth to you.
| Input | Value | Source |
|---|---|---|
| Downloads per episode | 25,000 | Show’s media kit, IAB certified |
| CPM for a host read midroll | $25 | Quoted rate |
| Placement cost | $625 | 25 thousands times $25 |
| Gross margin per order | $22 | Your accounting, not revenue |
| Orders needed to break even | 29 | $625 divided by $22 |
| Required response rate | 0.12% | 29 divided by 25,000 |
Now the only question that matters: do you believe roughly one listener in 850 will buy from a single midroll mention? For a product priced near an impulse threshold, aimed squarely at that show’s audience, with a memorable code, that is a defensible bet. For a $9,000 annual contract with a six week sales cycle, it is not, and you should be measuring pipeline instead of orders.
Buy three episodes minimum on any show, never one. Podcast response builds through repetition, and a single insertion mostly measures how many people happened to be in market that week. Ask for host read rather than a produced spot, and send the host three talking points instead of a script.
Route three: host your own show
Do not launch until you can answer three questions. Who is the specific listener, what will keep them subscribed after episode three, and where does the audio get reused? If the answer to the third is “nowhere”, the economics rarely work.
The realistic starting kit is a decent dynamic microphone per host, a headphone set, a remote recording service that records each participant locally, and a hosting platform that generates an RSS feed and gives you IAB style analytics. Editing is the hidden cost: budget two to three hours of editing per finished hour if you do it yourself, or roughly the cost of a nice dinner per episode if you contract it out.
Commit to 20 episodes before you judge anything. Discovery in podcast apps rewards catalogs, and most shows that quit do so at around episode eight, right before the back catalog starts working.
Repurposing: where podcast marketing actually pays
Whether you are guesting or hosting, the audio is raw material. One 45 minute conversation reliably yields the following.
| Output | Count | Effort | Where it goes |
|---|---|---|---|
| Vertical clips with captions | 5 to 8 | 20 min each | Shorts, Reels, TikTok, LinkedIn |
| Article built from the transcript | 1 | 3 to 4 hours | Your blog, with the episode embedded |
| Quote cards | 3 | 45 min total | Social, sales decks |
| Newsletter section | 1 | 30 min | Email, linking the article |
Always ask the host for permission before publishing clips of their show, and credit them with a link. Most are delighted, but a hashtag or a download does not grant you a license, which is the same principle covered in our guide on using user generated content. Fold the outputs into the repurposing map described in integrating social media with your marketing strategy so nothing sits unused in a folder.
Measuring something this lossy
Use three instruments at once and accept that none is complete. A vanity URL per show captures direct navigation. A unique promo code per show captures purchase intent, but only from people who remember it. A “how did you hear about us” field on your signup or checkout form catches everyone else and is usually the highest volume signal of the three.
Also watch branded search volume and direct traffic in the two weeks after each episode airs. A lift there with no other campaign running is real evidence even without a click path. For the broader framework, see our piece on measuring marketing ROI.
Frequently asked questions
How many listeners does a show need before advertising on it is worth it?
Audience fit matters more than size. A show with 3,000 downloads of exactly your buyer often beats one with 60,000 of a general audience, because your breakeven response rate is a percentage. Run the breakeven calculation above for both and compare the required response rates.
What is the difference between preroll, midroll and postroll?
Preroll runs before the content, midroll inside it, and postroll at the end. Midroll usually costs the most because listeners who reach it are engaged, and postroll costs least because many people stop before it. Buy midroll unless budget forces otherwise.
How do I get on podcasts with no audience of my own?
Pitch a specific segment idea tied to an episode the host recently published, and offer something concrete such as a number, a mistake you made, or a process listeners can copy. Hosts book interesting conversations, not follower counts. Start with shows in the low thousands of listeners and work upward.
Should my company podcast be about our product?
No. A show about your product has an audience of people who already bought it. Make it about the problem your buyers have, and let the product appear as evidence in stories rather than as the subject.
Do promo codes undercount podcast conversions?
Substantially. Listeners hear a code while driving, forget it, and search your brand later. That is why the “how did you hear about us” field matters: treat the promo code as a floor on performance, not a measurement of it.
The bottom line
Guest first, because it costs a few hours a week and tells you whether the channel works for your message. Advertise second, only after you have written down the breakeven response rate and decided you believe it. Host your own show last, and only if you already have a repurposing habit that will use the audio.
In every case the real return comes from the derivative assets and the relationships, not from the audio itself. Book the appearance, then spend the following week turning it into clips, an article and a newsletter section. That is the part most teams skip, and it is the part that compounds.
About this article: GeekBlog covers U.S. technology news, AI, phones, smartwatches and gaming. Every story is written and checked under our Editorial Policy. Spotted a mistake or have a story tip? Contact our editors.
