Is a podcast still worth starting?

Should I start a podcast for my business? In my view, yes, if two things are true. There is a specific group of people whose trust you want, and you have something to tell them they can't easily get elsewhere. Don't start one for a big audience or for ad money: ad buyers usually look for far bigger audiences than a typical show has. And it isn't too late. As of October 2026, the listeners are still there in steady numbers, and the field is less crowded than the millions of listed shows make it sound.

Narrated by a synthetic version of Lam Hoang's own voice.

The question comes up because the numbers sound like a rush. In Sounds Profitable's release on The Podcast Landscape 2026 ↗︎, published on 1 October 2026, 34% of American podcast consumers said they began consuming podcasts within the past year, up from 28% in 2025. Yet the share of adults listening barely moved. The release's own reading: "Although overall reach has remained steady, the podcast audience continues to refresh." So a third starting in a year is turnover, not fast growth.

Behind the numbers. The survey reached 5,473 American adults in August 2026, weighted to the Census, as the report page ↗︎ describes it. The start date is self-reported, and "podcast consumers" counts watching as well as listening. The pages we read do not say whether the base is everyone who has ever consumed a podcast or only monthly consumers, so how many people stopped in the same year cannot be worked out from them. Sounds Profitable is the podcast industry's trade association, funded by audio companies.

Is anyone still listening?

Yes, in steady numbers. In Sounds Profitable's summary of the core numbers ↗︎, 55% of American adults had consumed a podcast in the past month, 39% in the past week and 74% ever. All three are within a point of 2025. The survey's author, Tom Webster, calls the audience "large and stable", and adds that after two decades of growth, stable is itself news. It is a plateau, not a boom, so a new show can't count on a rising tide.

What is changing is who is listening. In my view, an audience that keeps refreshing is good news for a new show: it isn't only chasing people whose habits are already set. That is a reading, not a finding. Nobody has measured whether those newcomers try new or small shows rather than the biggest ones.

An outside check

Pew Research Center sells nothing to podcasters. Its Podcasts and News Fact Sheet ↗︎, published in September 2025 from a survey of US adults on its probability panel in August 2025, found that 54% had listened to a podcast in the past 12 months, "a slight increase since we first asked this in 2022 (49%)." That is a year's window, not a month's, so its level does not compare with the trade figures; only the direction does. Its wording is "listened", and the data is more than a year old as this episode goes out.

Why the record highs are left out

Edison Research's Infinite Dial 2026 ↗︎, from a national survey of 2,050 Americans aged 12 and older in January 2026, put monthly podcast consumption at 58%, "also an all-time high". In 2025, Edison widened its question ↗︎ to count video viewers as podcast consumers, so only the change from 55% in 2025 ↗︎ to 58% in 2026 compares like for like. The study was released with support from SiriusXM Media, which sells audio advertising. The trackers agree on the size of the audience and differ on growth: Edison finds a small rise among people 12 and older, Sounds Profitable finds adults flat. So the episode says large and steady, never booming.

How crowded is it?

Less than it sounds. Millions of podcasts are listed, but far fewer are still publishing. Listen Notes' podcast statistics ↗︎, tables updated on 10 October 2026, counted at least 3,821,938 podcasts worldwide. Of those, 567,678 had their latest episode in 2026, about one in seven (14.9%, our arithmetic). A single episode was enough to count.

Behind the numbers. Listen Notes is a podcast search engine that sells podcast data. Its count ↗︎ leaves out deleted feeds, empty or test feeds and AI-generated junk, which it filters with automated scripts and human moderators. The figures are worldwide and in every language. A show last active in 2025 may be seasonal or finished rather than dead, and listed is not listened to.

Fewer new shows

Fewer podcasts are starting, too. Podchaser, a podcast-data company, counted 153,767 new podcasts worldwide in the first half of 2026, "the lowest January-to-June total in the past eight years", down 14% on a year earlier, as Inside Radio reported on 2 July 2026 ↗︎. We read the relay, not Podchaser's report. Podchaser's own reading: "The people launching now are, increasingly, doing it on purpose." The same relay says 41.7% of the half year's new shows had already stopped publishing by the end of June.

Listen Notes' yearly counts point the same way: 1,020,468 new podcasts in 2020, the pandemic peak, then 211,038 in 2025, about a fifth of the peak. 2020 was a surge, so that says far below the peak, not still falling. Past years' counts rise a little as late feeds are indexed.

Machine-made feeds

Some new shows aren't made by people at all. Listen Notes removed 3,040 AI-generated fake podcasts in September 2026, against 8,848 new podcasts made by people: about one in four of that month's new feeds (25.6%, our arithmetic). That is its own classification, with no published accuracy figure. Other counts measure something else. In one 24-hour window in April 2026, the Podcast Index's own classifier marked 45.7% of new shows as potentially produced by AI, as Podnews reported ↗︎, and Podchaser's half-year estimate was at least 2.5% automated content mills. None of them measure listening: they count feeds.

Using AI isn't the problem. Putting in less work because you used it is. A show with real direction is easier to tell apart. This show is made with the help of AI, and I listen to every episode before it goes out. The episode on making a podcast with AI covers what directing the work looks like.

What a new show is up against

Not millions of shows. In the Sounds Profitable release ↗︎, "The typical monthly consumer regularly follows two or three podcasts." So a new show competes for one of a few spots, and that means being someone's best show on something specific they care about. The figure is self-reported, and the competing-for-a-spot reading is mine.

A made-up company that repairs commercial kitchen equipment shows the difference. A general business show would join one of the biggest categories listed: business is the third-largest genre in the Listen Notes index, with 472,666 shows, 12.4% of the total, by the genre each show chose. A show for restaurant owners about keeping a busy kitchen running has far fewer rivals for their time. That is an illustration of the reasoning, not a measured comparison.

What about ad money?

Don't count on it. The money is real: US podcast advertising revenue was $2.9 billion in 2025, up 17.6%, in IAB's Internet Advertising Revenue Report ↗︎, conducted by PwC and released on 16 April 2026. That is about 1% of digital advertising. IAB represents companies that sell digital advertising, and a market total says nothing about what a new show will earn.

What ad buyers look for

Acast, a podcast hosting and advertising company, set out the usual bars in a guide published on 2 July 2026 ↗︎. Ad marketplaces, Acast's among them, start accepting shows at around 1,000 monthly listeners, with automatically inserted ads. Advertising networks typically look for 1,000 to 5,000 downloads an episode, counted within the first 30 days. Direct sponsors usually want 5,000 to 10,000 for a general-interest show. The guide adds: "Niche B2B, finance, or professional audiences can land sponsors far earlier." These are an ad seller's stated thresholds, with no data behind the ranges.

What a typical show gets

Buzzsprout's podcast statistics ↗︎ for September 2026 rank every kind of show the hosting company carries by first-week downloads. The top 5% of shows started at 1,007 downloads an episode in the first 7 days, the top 10% at 412, and the median show got 27. There is no business-only split.

A week isn't a month, so setting Buzzsprout's 7-day figures against a 30-day threshold is our comparison, and a loose one. A thousand downloads in 30 days may sit nearer Buzzsprout's top tenth than its top twentieth. Either way, outside the niche exception, ad money needs an audience far above what a typical show gets. Plan the show around a different job. Any sponsor is a bonus.

What's it good for?

Depth with the right people. The episode on starting a podcast for your business put it plainly: a business doesn't need a crowd.

Jay Acunzo, who makes podcasts and teaches show-making, said in a 2018 Q&A with Managing Editor ↗︎ that podcasts were "better suited for depth and resonance, not reach and net new audience." His summary: "A podcast is intimacy that scales." He sells a show-premise workshop, a course, speaking and advice.

The podcasting coach Cliff Ravenscraft asked in an episode in May 2026 ↗︎ whether the world really needs another podcast. His answer "is not always yes": it doesn't need another generic interview show. His better question asks whether a specific group of people would be better served if you consistently shared what you have learned. He says "A podcast can build trust before the first sales conversation. It can help referrals understand how you think." He sells a podcasting course, coaching and a mastermind.

Other teachers set the same conditions, with different emphasis. Eric Nuzum, in a 2022 interview with Podcast Marketing Magic ↗︎, named FOMO as the worst reason to make one: "There needs to be a need in the world that only you can fulfill." Dave Jackson of School of Podcasting, writing in October 2025 ↗︎, answered the too-late question: "The secret isn't timing. The secret is focus." Jackson's post ends in an offer for his paid membership.

Confidence, not measurement

How much is that trust worth? Here, confidence runs ahead of measurement. Outcomes Rocket, a podcast production company, published a benchmark report in July 2026 ↗︎ from a May 2026 survey of 861 marketing and PR professionals, recruited through an online research panel and mostly based in Europe: 31% in Europe outside the UK, 30.8% in the UK and 18.8% in the US. In it, 47.3% of podcast marketers rated their overall return as strong. The report itself says that rating "often rests on engagement signals rather than pipeline or revenue data". Formal return on investment was tracked by 16.8%, and direct sales attribution by 20.4%.

A production company found that few marketers measure, which cuts against its own interest and lends the finding some weight. So treat the confidence you hear about business podcasts as a feeling, not a measurement. We found no published independent measure of what a business show pays back.

If you start one, take your own measurement from the first inquiry. Niall Mackay, who runs a small podcast production company, suggests in a July 2026 episode of his show Smarter Podcasting ↗︎ asking everyone who gets in touch how they heard about you, and keeping a sheet with three columns: who it was, how they found you, and whether the show came up. After about three months, he says, you can see whether the show is helping. Later episodes dig into what a show returns in clients and money.

Check your corner first

Timing matters less than the two conditions, so don't wait for a better moment. Before you decide, search the subject you would cover in a podcast app and play a few of the shows that come up. Check which ones have published in the last few months. Then ask what you could tell their listeners that they don't.

If a show of your own isn't for you yet, the next episode asks whether to make a podcast, be a guest on others' shows, or advertise.

What to do this week

1. Write down the specific group whose trust you want, in one sentence.
2. Write what you can tell them that they can't easily get elsewhere. If you can't, wait.
3. Search your subject in Apple Podcasts and Spotify, and play three of the shows that come up.
4. Note which of them have published in the last few months, and what none of them covers.
5. Decide what the show is for in your business, other than ad money, and how you will judge it. Start the three-column sheet with your first inquiry.
6. If you would rather hand the making to someone else, look at what podcast production for a business covers before you commit.

FAQ

Is it too late to start a podcast?

No. In Sounds Profitable's August 2026 survey, 55% of American adults had consumed a podcast in the past month, within a point of the year before, and 34% of podcast consumers said they started in the past year. Launches also fell to an eight-year low for a first half in 2026, by Podchaser's count.

How many podcasts are there?

Listen Notes counted at least 3,821,938 podcasts worldwide in October 2026, but only 567,678 of them, about one in seven, had published an episode in 2026, and a single episode was enough to count. Listed is not the same as active, or as listened to.

Can a new podcast make money from ads?

Rarely at first. Ad networks typically look for 1,000 to 5,000 downloads an episode in its first 30 days, by Acast's guide, while the median show on Buzzsprout got 27 downloads in its first week in September 2026. Niche business audiences can land sponsors earlier, but plan the show around another job.