Best place to advertise a podcast: inside another show, or on social? An October study says three to one.

On 9 October The Podcast Study 2026 put out one line that answers a question most show owners ask about their promotion budget.

Listeners are more than three times as likely to name social media as a way they found a new podcast than an ad inside another podcast, according to that release, which I read on 19 October.

So the best place to advertise a podcast is social. Case closed?

Not quite. The number is real. It just measures something narrower than the headline suggests, and the people quoting it have a reason to like it.

Step 1: Read what the study measured

The Podcast Study 2026 is a survey of 1,205 podcast listeners and viewers aged 18 to 54, run by Strategic Solutions Research and Point-To-Point Marketing, with Crooked Media, NPR and PRX named as co-sponsors on the study's own site.

Respondents were asked how they discover new shows. About half named social media, per the 9 October release. About one in six named an ad in another podcast.

That is a recall question. It tells you which route people remember. It does not tell you which route produced a listener who stayed for ten episodes.

And recall is biased toward the thing you saw most recently and most often. Social feeds are seen dozens of times a day. A midroll in someone else's show is heard once, in the car, between two segments you were waiting for.

So the gap is partly a frequency effect. Still a gap, though.

If you want the other big survey from October, Wondery's, our post on how people find new podcasts reads who it asked.

Step 2: Notice who is saying it

Here is the complication.

Point-To-Point Marketing co-runs the study, and its CEO Tim Bronsil is quoted in the release saying social media "can typically be bought at a significantly lower CPM than podcast advertising." Point-To-Point sells audience campaigns.

That does not make the finding wrong. It means the cost claim is a vendor's claim with no figure attached, and the release prints no CPM for either channel.

The discovery ranking is survey data. The "and it is cheaper" part is a sales line. Keep them apart when you take this to whoever signs off the budget.

The same release is fair to the other side, which helps its credibility. It says podcast ads beat ads in music streams for podcast discovery, and that the in-show ad share rises among superfans.

Step 3: Price both options honestly

An in-show ad has a clear unit. You buy a slot in a show whose listeners already listen to podcasts, which is the hardest habit to create.

A swap costs no cash. But it costs you a slot in your own episode, and the same study says that slot is getting expensive in a different currency. About half of its respondents said podcasts already carry too many ads, per the 9 October release, and fewer than one in ten said they would accept more to keep shows free.

Social has a less obvious unit. Clips cost editing time or an agency fee, and the reach is uneven. Most clips will die.

That is fine. The point of a clip is that the winners are found by the platform, not by you, and one clip that travels can outrun a month of midrolls. At VALORAE Arc we have delivered 500+ videos and generated 10M+ views across clients, and that total is carried by a minority of clips. We cannot predict which ones in advance. Nobody can.

Step 4: Put clips first, for a specific reason

Clips go first because they are the only channel here that lets a stranger sample the show without committing to it.

An ad in another podcast asks a listener to remember a name, open an app and search. A clip is the show. The viewer hears the host, the guest and the pace in under a minute and decides on the spot.

That is the job Arc does for podcasts and founders: take the long episode, find the moments that work for someone who has never heard of you, cut them vertical with captions, and post them. Retention matters more than raw views here, so it is the number we read first on every Arc report.

One clip cut once still needs adjusting per platform. Media's guide to posting one video across five platforms covers what to change and what to leave alone.

Step 5: Then test one in-show ad, properly

Do not skip in-show ads because a survey ranked them second. Run one, and run it so it can be read.

Pick a show whose listeners would plausibly like yours. Ask for a host-read slot, because the 9 October release says host-read ads outperform prerecorded ones on getting listeners to try something.

Give the ad its own link or vanity URL. Nothing else should point there.

Then compare cost per new follower from that ad against cost per new follower from a month of clips. That comparison is the answer for your show. A survey of 1,205 strangers is a starting point, nothing more.

Step 6: Treat the swap as a trade, not a freebie

Swaps feel free. They are not.

You give up an ad slot in your own episode, which is either revenue or goodwill with your listeners. If your show already runs several ads, a swap adds to the load that half of those survey respondents say is too high.

So limit swaps to one per episode at most, keep them host-read, and only swap with shows of a similar size. A swap with a much bigger show is a gift you will not be offered twice. A swap with a much smaller one costs you more than it returns.

I would not run a swap and a paid in-show ad in the same month. You will not be able to tell which one worked.

Step 7: Decide what counts as a win before you spend

Write the success measure down first.

For clips: new followers on the show's own channels, and the new versus returning split on the full episode in YouTube Studio. For an in-show ad: hits on its own link, and follows in the week after it airs.

Pick a window, say one month, and do not move it. If the clip programme cannot show new people arriving within that window, change the clips before you change the channel.

We did not expect the ranking in this study to be that lopsided. It might narrow next year. Measure on your own show and you will not need to care.

The short version

• Read the three-to-one figure as recall, not conversion.
• Separate the survey finding from the vendor's cost claim.
• Start with clips, because they let strangers sample the show.
• Test one host-read ad in another show, with its own link.
• Keep swaps to one slot, with shows your size.
• Set the success measure and the window before spending.

Clips cut from your own episodes and posted for strangers is what we run at valoraearc.com.