Podcast news roundup: 1 October 2026

This edition runs from 17 to 25 September 2026, a longer stretch than usual, and it takes in Podcast Week in New York, the Made on YouTube event on 23 September, and the reports that followed both. The theme: platforms are starting to describe your show on your behalf. YouTube wants to test your first minute, YouTube Music wants to pitch your podcast in a synthetic voice, and the largest ad buyers have agreed on what a YouTube view is worth. Acast, meanwhile, has data showing audiences treat audio and video as one programme, which carries through to the edit desk: how do you record usable audio in a room that is working against you?

Should you be testing three versions of your opening?

YouTube's Made on YouTube announcements on 23 September included video A/B testing, described as coming soon, which will let creators test up to three video cuts to see which hook performs best. It extends the title and thumbnail testing YouTube already offers, which it says creators have used for more than 40 million experiments. In the same post, voice likeness detection is being integrated with facial detection to improve match accuracy, accessible in the YouTube mobile app later this year. And YouTube says 72% of US video posters aged 14 to 44 have used AI to help create or edit in the past year.

My reading of the A/B tool: three cuts that differ in the hook are, in practice, three cold opens. The body of the episode stays the same and the first minute changes, so what is being measured is which moment goes first. That is an editorial decision, and the one we spend longest on in a retention intro. A test will tell you which of three opens won. It will not find the fourth one you should have cut.

Also on 23 September, the YouTube Music app is getting Your Podcast Lineup, a personalised AI-audio guide that gives a quick weekly spoken preview of recommended shows, why you would like them, and then the suggested episodes. YouTube calls it eyes-optional and hands-free; it is coming soon, with no date, for YouTube Music Premium and YouTube Premium subscribers worldwide. Alongside it is Ask Music, a conversational experience across podcasts and over 300 million songs. TechCrunch covered both. The practical point: a machine is about to describe your show in its own words, and its only raw material is your show description and your episode titles. If those are an afterthought, the pitch will be too. SEO titles and show notes exists for that job.

The buyers have decided what a YouTube view is

Earlier editions covered YouTube's 24 August change to what counts as a public view, and Podscribe moving to engaged views. The follow-through came at the Sounds Profitable Business Summit in New York. Per a LinkedIn write-up dated 18 September, three large podcast ad buyers, Glenn Rubenstein of ADOPTER Media, Dan Granger of Oxford Road and Krystina Rubino of Right Side Up, said they will use YouTube Engaged Views as their definition of a YouTube impression for 2027 buying; together, the write-up says, they represent roughly half a billion dollars of buying. Rubenstein put it plainly: "We're using Engaged Views. That's our definition of a YouTube impression next year." The same write-up notes that public counts can run up to 2.5 times engaged views. Podnews reported on 21 September that YMH Studios made a similar commitment, and quoted the IAB's Jason Adamski on the PodBiz podcast: "I don't know if we have a standard definition of a play."

Opinion: for anyone selling sponsorship, this settles it. The public number on your video page is a vanity figure, and the people with the budgets have now said so out loud. Put engaged views in your media kit and nothing else.

Are your listeners switching between audio and video mid-episode?

Acast's Podcast Pulse 2026, released on 23 September, is a Differentology survey from August 2026 of 4,300 respondents aged 16 to 64 across 13 markets. 88% consume both audio and video podcasts, up 10 points on 2025; 10% are audio-only and 2% video-only. 63% have started an episode in one format and finished it in another. 42% trust brand recommendations from podcasters, up from 33% in 2025 according to Acast's report page. 55% co-view video podcasts on a connected TV, and 63% of podcast clip viewers go on to consume the full podcast at least occasionally. Acast CEO Greg Glenday: "The industry spent years arguing over formats, but audiences moved on long ago."

The switching figure is the one I would pin up. If nearly two thirds of an audience might start with the picture and finish without it, the audio and the video are not two products; they are one edit that has to work with the screen off. In practice that means no "as you can see", no pointing at a slide without saying what is on it, and no joke that only lands if you saw the face. We already treat video and audio editing as a single job for that reason, and the clip figure is the case for short-form clips cut to send people somewhere, rather than to exist.

Netflix is now ahead of Apple for finding podcasts, in the US

Tom Webster's Sounds Profitable article "Tu-dum", published on 23 September and drawing on The Podcast Landscape 2026, is US data, so read it as such. Asked where they encountered podcasts in the past year, respondents put YouTube first, Spotify second and Netflix third, ahead of Apple Podcasts. Netflix reached 21% of podcast consumers, and 5% name it their primary platform. Podcast-native apps as the most-used app fell from 19% to 15%.

Webster is not celebrating. As Podnews picked out on 24 September: "We are potentially making a deal with the devil when we put our programming on Netflix, because ultimately we don't matter to Netflix." His concerns are gatekeeping (you cannot upload yourself; only established shows get picked), cancellation on viewing metrics, and the loss of discovery. Podnews on 18 September quoted his comparison of the Netflix podcast shelf to "the CD range in a Walmart". The UK has had a taste already: Goalhanger's The Rest Is Football ran as a daily Netflix programme during the 2026 World Cup, as The Independent reported in June. My view: for a show with a following, Netflix is a broadcast slot rather than a home, and broadcast slots get withdrawn. Keep the feed you own.

Spotify pays in more places, Apple promises paid video

Spotify's Partner Program expanded to more than 35 new markets on 17 September, including Italy, Spain, Brazil, Mexico, Poland and much of Latin America and the Caribbean. Spotify says total monthly payouts to enrolled shows have increased by more than a third, and that video shows in the programme have seen more than 45% average growth in consumption hours since its launch. The UK is not on the list.

Apple said on 22 September that video is coming to Apple Podcasts Subscriptions soon, with no date. Creators will be able to offer subscriber-exclusive, ad-free, early-access and bonus video, available to audiences in over 170 countries and regions, and hosting providers will share details in the coming months. The counterpoint ran in Podnews the same day: creator Courtney Elmer spent six weeks and more than 50 emails trying to offer her weekly show two days early through Apple Podcasts subscriptions, and found the subscribe button only shows when there are episodes waiting. Adding video to a subscription product whose subscribe button is conditional is the wrong order of work, and that is opinion. The same Podnews issue covered Podtastic, an ad-skipping app at $9.99 a month that removes ads on the device, with app store fees going to Apple and Google and nothing to creators; Podnews says it reported the app to Apple on 22 July and neither company has acted.

Also via Podnews, on 25 September: Apple Podcasts Connect quietly corrected consumption figures after a data processing error. On Podnews' own show, a 9 July episode went from 176% average consumption to 85%, a 6 July one from 166% to 83%, and larger shows were affected too. If you have screenshotted Apple numbers into a sponsor deck, recheck them.

Sheffield's £6.65 million, and where Goalhanger's money comes from

Crossed Wires, the Sheffield podcast festival, said on 23 September that an independent assessment found its 2026 festival generated £6.65 million in economic impact. Attendance doubled to 50,000 in the festival's third year; 52% came from outside South Yorkshire and 40% from outside the north of England and the Midlands. South Yorkshire mayor Oliver Coppard: "Tens of thousands of people have come to our region, filling venues, supporting local businesses and bringing millions of pounds into our economy."

Podnews read Goalhanger's 2025 accounts at Companies House on 25 September. Of roughly £37 million in revenue, 67% was advertising, 25% memberships, 4% events and 2% commissions, with 54 staff; 38% of revenue came from the UK and 46% from the US. The Independent reported in June that 2025 sales were £37.9 million with around 321% average annual growth over three years, top of the Sunday Times 100. What I take from it: a quarter of Goalhanger's income is people paying for the shows directly, which is the strongest argument for a paid tier I have seen this year, and events, for all the noise around them, are 4%.

Two briefs. Podnews also noted on 25 September that a counterfeit "The Mel Robbins Podcast", credited to "Mel Robinss" and hosted on Spreaker, sat at number 28 in Apple Podcasts UK's self-improvement chart while the real one was at number 2; it has since been removed. Search your own show name now and then. And Podcast Week 2027 is scheduled for 13 to 17 September, per Podnews on 21 September.

The numbers

From the edit desk: how do you get usable audio when the only room you have is a bad one?

Of everything that lands on my timeline, the fault I can do least about is a room. Noise I can reduce: fridge hum, a fan, traffic outside, all come down with tools and mostly stay down. Echo does not. Reverb off hard walls and a bare floor is baked into every syllable, and every plugin that promises to remove it swaps the room for a hollow, underwater artefact that is worse than the echo was. So the fix happens before anyone presses record, and most of it costs nothing.

1. Pick the smallest, most cluttered room you have. A bedroom with a bed and a wardrobe beats a kitchen or an empty office, because soft things swallow reflections and hard things bounce them back.
2. Soft near the mouth beats foam on far walls. A duvet or a thick coat hung behind the screen, opposite the person talking, does more than panels three metres away.
3. Get close: about a hand-width from the mic. Distance is where the room gets in.
4. In a bad room, a dynamic mic over a condenser, which hears everything including the wall behind you. Our podcast microphones guide covers which is which.
5. Turn off the fridge, the fan and any heating that ticks. Phone on silent and away from the mic.
6. Headphones on, so nothing bleeds back into the mic.
7. Record ten seconds of room tone before you speak. It is what the gaps get filled with.
8. Do a 30-second test and listen back on headphones before the real start. If you can hear the room, move the duvet and go again.

For guests, send this the day before, word for word if you like: "Could you record from the smallest room you have, with something soft behind your laptop, headphones in, and the mic about a hand's width from your mouth? A 30-second test before we start would be brilliant. It is the difference between sounding like a podcast and sounding like a phone call."

Why now: Acast's 63% who switch between audio and video mid-episode means the audio carries the episode even when the picture looks fine. A tidy shot in an echoing room still sounds like an echoing room the moment the phone goes into a pocket. Tools help with noise; they do not help with a room. Only the room helps with the room.

Which of these will you do before the next episode goes out?

Three, in order of effort. Export your engaged views and put that number, and only that number, in front of any sponsor, because the buyers have just told you it is the one they will count. Rewrite your show description and your episode titles as if a machine will read them aloud to a stranger, because YouTube Music is about to. Then listen to your last episode with the screen off and note every moment that only made sense with the picture; that is the list for your next recording, along with the duvet. If you would rather hand the edit to a person, a full episode edit starts at £280, billed per four-week cycle, with no minimum and revisions until you are happy. Tell us about your show and we will take it from there. Back next Thursday.