Does Podcast Advertising Work and What the ROI Data Shows

Reel Paper, a paper goods brand, saw podcasts become its top new-customer channel in under two months, ahead of Meta and Google, according to its agency ADOPTER Media¹. Over its first six months, monthly podcast-attributed revenue rose 343%¹.

So yes, podcast advertising works for a lot of brands. It doesn't work for everyone, and it rarely works on the first three shows you try. This page lays out the response data, the market growth, two published cases, the ways podcast ads fail and a cheap way to test before you commit real money.

Facts and prices checked November 11, 2026. See something wrong? Tell us.

Key takeaways

• Magellan AI found a 2.23% median podcast ad response rate in H1 2026, and host-read ads beat produced ads.
• Among listeners who reached the advertiser's site, the median purchase conversion rate was 3.98% and the median lead conversion rate 4.65% (Magellan AI, H1 2026).
• US podcast ad revenue reached $2.862 billion in 2025, up 17.6%, from $105.7 million in 2015 (IAB and PwC).
• Reel Paper's podcasts became its top new-customer channel in under two months, with a 343% rise in monthly podcast-attributed revenue over six months, per ADOPTER Media.
• A fair test is 5 to 10 shows, three or four episodes each over seven to ten weeks, judged at 30 days or later.

The short answer

Podcast advertising works when three things line up. The show's audience matches your buyer. The host reads your ad in their own voice. And you measure long enough to see the result. When one of those breaks, podcasts look like they don't work.

The data supports the upside. Brands keep spending more. Magellan AI reports US podcast ad spend up 26% year over year in Q3 2025² and up 23% in Q2 2026³. New brands keep arriving, with 1,297 brands advertising on podcasts for the first time in Q2 2026 alone³. Brands don't keep raising budgets on a channel that loses money for all of them.

The audience is big too. Edison Research's Infinite Dial 2026 found 58% of Americans age 12 and up listen to a podcast monthly, about 167 million people, and 45% weekly⁴.

The catch is measurement. Half of brand decision makers in an Oxford Road survey said they hold back podcast spend because of limits in performance data⁵. Podcast ads often work better than they look, because the tools miss sales. That's covered below.

What the response data shows

The best public benchmark on podcast ad response is Magellan AI's 2026 report covering campaigns from January 1 to June 30⁶. Response rate here means the share of unique reached listeners who later visited the advertiser's website, with a 30-day lookback⁶.

• Median response rate across ads was 2.23%⁶.
• Host-read ads had a higher median than produced ads, as the chart shows⁶.
• Among visitors, the median lead conversion rate was 4.65% and the median purchase rate 3.98%⁶.
• Video podcasts had a 2.05% response rate against 1.88% for audio-only feeds, and a 4.50% lead conversion rate against 2.57%⁶.

What that means in dollars

Take a show with 20,000 downloads an episode at a $25 CPM. Influencer Marketing Hub prices that at about $500 a placement⁷. Using Magellan's medians, 20,000 reached listeners at 2.23% is about 446 site visits, and 3.98% of those is about 18 purchases⁶. That's roughly $28 per purchase. This is our arithmetic on medians, and downloads aren't the same as reached listeners, so treat it as a rough guide. If your first order is worth more than $28 in margin, that's a show worth testing.

A host read can improve that math, since host reads had the higher median response⁶. Same price, same show, a different ad type.

Market growth says brands keep buying

Market size isn't proof an ad works for you. But steady growth over ten years means many advertisers renew.

US podcast ad revenue was $105.7 million in 2015⁸. It passed $1 billion in 2021 at $1.449 billion⁹, reached $2.433 billion in 2024¹⁰ and hit $2.862 billion in 2025, up 17.6%⁸. Podcasts posted double-digit growth in four of the past five years⁸.

Spend is split between goals. In Q3 2025, 56% of podcast ad spend went to brand awareness, 41% to direct response and 3% to tune-in promotion². So a big share of buyers judge podcasts on sales, not just reach. For the full set of numbers, see our podcast advertising statistics.

Two published brand cases

Both cases below are published by the agency or tool involved. Neither is independently audited. They are still the clearest public examples of what a working podcast program looks like.

Reel Paper

Reel Paper sells sustainable paper goods direct to consumers. After iOS 14 hurt its digital ads, a lean in-house team struggled to pick shows and measure results¹. ADOPTER Media moved it to 60 second host reads on shows including Pod Save America, Pod Save the World and What a Day¹. It also ran a $50,000 buy across top US shows to support Reel's 2022 launch in Target¹.

Per ADOPTER, podcasts ranked as Reel's top new-customer channel in a post-purchase survey within two months, ahead of Meta, Google and influencers¹. Over the first six months, monthly podcast-attributed revenue rose 343% and cost per acquisition fell, which ADOPTER reports as a 276% decrease¹.

Fum

Fum, a wellness brand, ran host-read podcast ads tracked with promo codes and vanity URLs. Only about 20% to 30% of listeners used them, leaving roughly 70% to 80% of conversions unattributed¹¹. After adding pixel attribution, Podscribe says those conversions were accounted for, and it headlines the case as an 830% scale-up without explaining the metric¹¹.

What the two have in common

Both used host reads. Both measured beyond codes, with a survey for Reel Paper and pixels for Fum. And both scaled after the measurement showed podcasts were working. That's the pattern to copy.

Video podcasts and brand effects

Podcast ads don't only pay back through codes and clicks. Two trends make the case stronger than direct sales alone suggest.

More shows now come with video

About 71% of US podcast creators make video versions of their shows, per Sounds Profitable research cited in the IAB report⁸. Edison Research found the share of Americans who have both listened to and watched podcasts rose from 48% to 57% in a year¹². For an advertiser, that means one host read can reach people on audio apps and on YouTube.

The video audience seems to act on ads more. In Magellan AI's H1 2026 data, video podcasts beat audio-only feeds on response rate, 2.05% against 1.88%, and on lead conversion, 4.50% against 2.57%⁶. Magellan's buyer guide also says shows with a YouTube simulcast have an 8% higher advertiser renewal rate¹³. Renewals are a decent signal that ads on those shows paid back.

Awareness shows up later

More than half of podcast ad spend in Q3 2025 went to brand awareness, at 56%². Those buyers judge podcasts on recall and search, not codes. Even direct response buyers get some of this effect. A listener who hears your brand three times over six weeks may search your name instead of typing the code. That sale lands in your branded search or direct traffic, not under the podcast. It's one more reason podcast ads often look weaker in a dashboard than they are.

What this means for your test

Ask each show if it publishes video, and where. Watch branded search and direct traffic during the test, not just code sales. And when you compare shows, compare like with like. A video show and an audio-only show of the same size may not deliver the same results.

When podcast ads fail

Most failed podcast tests fail for reasons you can see coming. Here are the common ones, with what the data says.

Codes are the only measure

Acast says codes and vanity URLs undercount, because listeners often search the brand later¹⁴. Fum's codes caught only about a quarter of conversions¹¹. If you judge on codes alone, you'll cut shows that work.

The test is too small

Three shows and one episode each is not a test. Influencer Marketing Hub says to start with at least 5 to 10 podcasts⁷. Magellan's buyer guide suggests three or four episodes per show over seven to ten weeks¹³.

Judging too soon

Only about 41% of site visits arrive in the first seven days, and nearly 76% within 30 days⁶. A day 7 review will make every show look weak.

Over-scripted reads

Brands hurt performance by over-scripting hosts with language that doesn't fit the show⁷. Produced spots, the most scripted ads of all, trail host reads on response⁶.

Paying for the famous name

Oxford Road, an audio agency, says brands systematically overpay for household-name shows, and its own top 15 by ad performance left out some of the biggest shows⁵. That ranking isn't externally checked, but the point stands. Big downloads don't guarantee a fit.

Can't find enough shows

The quiet failure is supply. To test 5 to 10 shows and swap out losers, you need 20 or 30 candidates and a way to reach each host. Many brands stop at the three shows they already know, and a three-show test rarely gives a clear answer.

Who podcast advertising works best for

Spend data shows which kinds of brands keep coming back. In September 2026, Magellan AI's top estimated podcast spenders included Quince, BetterHelp, T-Mobile, Toyota, DraftKings and Shopify¹⁵. In Q3 2025, financial services was the largest category at $89.6 million, followed by consumer software at $75 million and business services at $69.7 million².

From the cases and benchmarks, podcasts tend to suit brands with these traits.

• A product a host can use and talk about. Host reads beat produced ads on response⁶, and a real story needs a real product.
• A clear first offer. A discount or trial gives the listener a reason to act and a code to type.
• A niche audience you can name. Reel Paper targeted shows whose audience matched its core buyer¹.
• Patience for a 30 to 60 day read. Responses nearly double between day 7 and day 30⁶.
• Margin to cover a learning period. First-time podcast advertisers spent an average of $33,900 in Q3 2025², though a small test can cost far less.

Podcasts are a weaker fit if you need clicks today, can't make a code or landing page, or sell something a host can't honestly describe. If you're weighing podcasts against creator video, see podcast ads vs YouTube sponsorships.

How to test podcast ads cheaply

You don't need a $50,000 launch to find out. Influencer Marketing Hub puts small business and startup podcast budgets at $2,000 to $10,000⁷. Here's how to make that work.

Go smaller than you think

CastFox puts typical flat fees at $25 to $220 per placement for shows with 1,000 to 5,000 downloads an episode, and $100 to $800 for 5,000 to 20,000¹⁶. Ten shows in the 5,000 to 20,000 band, three episodes each, would run $3,000 to $24,000. Mix in some 1,000 to 5,000 download shows and you can fit inside a $10,000 budget.

Copy the Libsyn shape

Libsyn Ads suggests 1 to 3 larger shows in your niche plus 10 to 20 smaller ones to test ROI¹⁷. The small shows teach you which audiences buy. The bigger ones tell you if it scales.

Set up measurement on day one

One code and URL per show, a post-purchase survey and a target cost per sale. Our podcast attribution guide covers the setup.

Give it time

Magellan's guide says to let ads mature, citing research that a single ad can take about six weeks to reach full effect¹³. Run three or four episodes per show¹³ and judge at 30 days or later⁶.

A simple 5 to 10 show test plan

1. Pick 20 to 30 candidate shows in your category, country and language.
2. Email each host for rates and ask for average downloads, audience and one past sponsor result.
3. Book 5 to 10 shows that fit your budget⁷, mixing a few larger ones with smaller ones¹⁷.
4. Send talking points and the product so each host can speak from real use. See our podcast advertising guide for formats.
5. Run three or four episodes per show over seven to ten weeks¹³.
6. Read results at 30 and 60 days using codes, URLs and survey answers together⁶.
7. Renew the winners bigger and replace the rest from your candidate list. Libsyn notes reorders often cover a quarter or half a year¹⁷.

Step 1 and step 7 depend on the same thing. A long list of shows that fit, with a way to reach each one. That's what keeps a test alive after the first few shows miss.

Run your podcast test with the right shows

A podcast test is only as good as the shows on it. Media List helps you build that list and reach every host on it.

• Find shows. Search 1.2 million podcasts with contact emails by category, country and language.
• See who to contact. Email the host or producer directly for rates and open dates. Emails unlock from your plan's allowance, no bot checks.
• Save a shortlist. Keep your 20 to 30 candidates in one saved list, plus backups.
• Send from built-in campaigns on Professional. Send one rate request to the whole list, follow up, and track replies.

Media List doesn't have rates or download numbers. You get those from each show when you ask. If you're new to the first email, read our guide to writing a pitch that gets a reply.

Basic at $99 a month covers search, saved lists and exports. Built-in email campaigns come with Professional at $399 a month. Start with a 3-day trial, $0 today, and cancel anytime. See plans and pricing.

Questions, answered

Do podcast ads actually work?

For many brands, yes. Magellan AI found a 2.23% median response rate in H1 2026, and published cases like Reel Paper show podcasts becoming a top new-customer channel. Results depend on audience fit, host reads and measuring for at least 30 days.

Is podcast advertising worth it for small businesses?

It can be, if you test small shows first. CastFox puts flat fees at $25 to $220 per placement for shows with 1,000 to 5,000 downloads an episode. Influencer Marketing Hub puts small business budgets at $2,000 to $10,000.

What is the ROI of podcast advertising?

There's no single audited number. Vendors publish figures like a 4.9x average return on ad spend in one study of Acast campaigns, but those are vendor claims. Your ROI depends on your margin, the show's price and how much your tracking misses.

Are host-read ads better than produced ads?

On response, yes. In Magellan AI's H1 2026 data, host-read ads had a higher median response rate than produced ads. Host reads also let the host share real use of the product.

Why do podcast ads seem not to work?

Usually because of measurement, test size or timing. Codes miss many buyers, three shows is too few to compare and nearly 76% of site visits come within 30 days, not the first week. Over-scripted reads and poor audience fit also hurt.

How long does it take for podcast ads to work?

Plan for 30 to 60 days. Magellan AI found about 41% of site visits come within seven days and nearly 76% within 30 days. Its buyer guide suggests three or four episodes per show over seven to ten weeks.

How many people listen to podcasts?

Edison Research's Infinite Dial 2026 found 58% of Americans age 12 and up listen monthly, about 167 million people, and 45% listen weekly.

Related guides

• Podcast Advertising Guide for Brands Buying Direct in 2026
• Podcast Advertising Rates and CPM Benchmarks in 2026
• How Much Does It Cost to Advertise on a Podcast
• Podcast Database and Contact List for Advertisers
• Podcast Ad Agency vs Marketplace vs Buying Direct From Hosts

Sources

1. ADOPTER Media, Real results for Reel Paper (case study), checked 2026-11-11
2. Radio Online, Magellan AI finds podcast ad spending up 26% in Q3 (Nov 4, 2025), checked 2026-11-11
3. Magellan AI, Podcast advertising benchmarks Q2 2026, checked 2026-11-11
4. PPC Land, 58% of Americans now listen to podcasts monthly (Edison Research Infinite Dial 2026, Mar 15, 2026), checked 2026-11-11
5. Tubefilter, Podcasts are selling products, but some brands are still holding back (Oct 31, 2025), checked 2026-11-11
6. Radio Online, Magellan AI host-read podcast ads lift response (Sept 10, 2026), checked 2026-11-11
7. Influencer Marketing Hub, How to buy podcast ads (updated May 27, 2026), checked 2026-11-11
8. Radio Ink, IAB digital audio grew 10% in 2025 as podcasts near $3B (Apr 16, 2026), checked 2026-11-11
9. IAB and PwC, US Podcast Advertising Revenue Study FY2023 (May 2024), checked 2026-11-11
10. IAB and PwC, Internet Advertising Revenue Report, Full Year 2024 (Apr 2025), checked 2026-11-11
11. Podscribe, How Fum scaled podcast ads (case study), checked 2026-11-11
12. National Public Media, Audio continues to break listening records, 2026 Infinite Dial report (Apr 2026), checked 2026-11-11
13. Magellan AI, How to find profitable podcast sponsorships (Aug 2021), checked 2026-11-11
14. Acast, Podcast attribution tools, what's available and how to choose (Jul 6, 2026), checked 2026-11-11
15. Magellan AI and Podnews, Top podcast spenders (September 2026), checked 2026-11-11
16. CastFox, Podcast sponsorship rates by audience size 2026 (Oct 14, 2026), checked 2026-11-11
17. Libsyn Ads, Learn how to run a successful podcast advertising campaign, checked 2026-11-11