Podcast news roundup: 3 September 2026

Sixty-four per cent. Keep that figure in view for the rest of this edition, because it is the share of a YouTube podcast "view" that still means what the word meant before 24 August. Podtrac published the number on 2 September, the first hard measurement of how YouTube's redefined view count lands on podcasts specifically, and it changes how every public figure on your channel should be read. Around it, the week of 27 August to 2 September 2026 produced a British commercial partner founded by three of the best-connected people in UK audio, an awards finalist switched off by Spotify for ten weeks over music it had licensed, two hosting companies fixing the unglamorous end of distribution, and a diary that runs from Terminal 5 in New York to Kings Place in London. The edit desk takes on the cold open, which YouTube has just turned from a matter of taste into a measured one.

What does YouTube's new view count do to your podcast numbers?

Start with the rule change, effective 24 August and carried by both Podnews and vidIQ. A public YouTube view is now counted the moment a video starts to play, from the first frame, and the rule applies worldwide to Shorts, long-form video, podcasts and livestreams. The thing YouTube used to call a view has not gone; it has been renamed an "engaged view", defined as how many times viewers stayed to watch past the first frame and the initial seconds. Creators can still see engaged views in YouTube Analytics. Nobody else can: they are absent from public pages and from YouTube's public Data API, which is the door most third-party rankers and analytics firms walk through. YouTube says earnings and Partner Program eligibility are untouched, because both continue to run on engaged views and watch time.

Then the measurement. Podtrac sampled 84.2 million views of videos posted to YouTube Podcast Channels over 27 and 28 August. For podcast episodes, engaged views came to 64% of total views, a difference of minus 36%. For the other videos on those same channels, engaged views were 58% of the total, minus 42%. Podtrac reads the six-point gap as consistent with more intentional viewing of podcasts than of the other material on those channels, and it has said it will use engaged views for its YouTube podcast rankings from here.

Now the arithmetic every host needs. If engaged views are 64% of views, the public figure is a little over one and a half times the engaged one. Your channel's public counts are about to look substantially larger with not a single extra person listening. Three things not to do. Do not celebrate the jump; it is a relabelling, not an audience. Do not compare this month's public number with last month's, because they were counted under different definitions and the comparison is meaningless. And do not let a sponsor quote the public figure back to you in a deck, because a media buyer who later reads the fine print will feel misled, and they will not blame YouTube.

Two things to do this week. Export or screenshot your engaged views now, per video and per month, so you hold a like-for-like baseline before the old definition fades from memory. And in any sponsor conversation, quote engaged views. It is the honest number, it is the one that survives scrutiny, and if a rival show is quoting the inflated one, you would rather be the show whose figures still hold up in the second meeting.

The structural worry sits underneath all of this. The metric that means something has just gone private, so every public leaderboard built on view counts gets less useful at a stroke. That is precisely why Podtrac switching its rankings to engaged views matters: it is the first major ranker choosing the meaningful number over the visible one, and the rest should follow it.

Is Everywhere a bet worth watching?

A new UK commercial partner called Everywhere launched on 2 September, reported by Podnews and PodcastingToday, and for now the founders are the story. Andy Goldsmith built Adelicious into one of the UK's leading independent podcast networks before Audioboom acquired it in 2024, then served as Audioboom's UK President. Craig Eastwood worked alongside him at both companies, most recently leading Audioboom's international expansion, with earlier years at RadioWorks planning and buying radio, digital audio and podcasts. Mike Wooller led content development at the BBC, Acast and Global, and was latterly Head of Strategy and Platforms, Podcasts at DMG.

The model: Everywhere acts as the exclusive commercial partner to a curated collective of creators and content publishers, selling their audiences across audio, video, social and live events as one proposition rather than as separate budgets. Advertisers get a single point of access to a multi-platform creator audience. Goldsmith's framing: "Brands have always wanted to be where their audiences are and, increasingly, those audiences are choosing to spend their time with creators." The launch material leans on a Goldman Sachs forecast that the creator economy could reach roughly $480 billion globally by 2027. No creators have been named; the roster is promised in the coming weeks.

My reading. This is a British bet that the podcast is no longer the product; the creator's whole audience is. The pedigree is real and the thesis is sound, because the budgets genuinely are converging. But be honest about what has not been shown: no roster, no numbers, no case study, and "exclusive commercial partner" contains the word exclusive, which is a serious thing to sign and a harder thing to unwind. The useful point for an ordinary show is sharper still. A cross-platform sale only pays if your audience actually exists on more than one platform, and for most shows the second and third platforms are an afterthought with a logo on it, not an audience. Until that changes, Everywhere is a story to watch rather than one to act on, unless they are already courting you, in which case ask to see the roster before you talk terms.

Why did Spotify switch off a licensed podcast for ten weeks?

Reported by Podnews on 1 September: Cassette Culture, an Australian show from East Coast Studio, was removed from Spotify for ten weeks. One episode contained music, Spotify's automated music-detection systems flagged it, and the whole show came down. The team held proper licensing, and it made no difference, because Spotify's terms reserve the right to delist an entire show where musical content appears questionable, licence or no licence. At the time, the show was a finalist for the Independent Podcast Awards. The same report carries Martin Franklin's caution against depending on a platform and his warning about which embedded player you choose to build on.

Here is the point worth making plainly. Licensing protected them legally and did nothing for them operationally, because the enforcement was a machine and the appeal took ten weeks. Ten weeks is most of a quarter. For a show building momentum toward an awards night, it is worse than the calendar suggests, because momentum is the thing that does not come back when the feed does. This is the strongest argument for owning your RSS feed that anyone has made this year, and acting on it costs nothing.

Three prescriptions. If your show uses music beyond a short sting, keep the licence paperwork where you can produce it inside an hour, not inside a fortnight of emails. Never make a platform's embedded player the only route to your audience. And make sure your own site or show page carries a player that no platform can switch off; when the feed is yours and the player is yours, a delisting is a lost storefront, not a lost business.

Distribution plumbing: one click, one upload, one feed

Two hosting companies improved the dull end of publishing this week, and the dull end is where most shows quietly leak audience.

Buzzsprout's redesigned directory flow, shipped on 13 August according to its What's New page, submits a show to 16 audio directories with one click, and its rebuilt directories page shows where a podcast is already listed, where video is enabled and what still needs attention. The news, reported by Podnews on 1 September, is what the feature revealed: Buzzsprout found that over 70% of active podcasts are not in all the directories they could be, and more than 5,200 Buzzsprout podcasts added at least one new directory listing within two weeks of the launch. That 70% is the most immediately useful number in this edition. It describes a free, unglamorous, ten-minute job that most shows have simply never done. Go and check your listings this week, whoever hosts you.

Blubrry, announced on 2 September via PR Newswire and Podnews, now works with Apple Podcasts to carry HLS video for shows published through Blubrry Publisher. One MP4 upload becomes a video version and an audio version through a single RSS feed: Blubrry builds the HLS video for Apple Podcasts, keeps the original video file, and generates an MP3 for apps that cannot play HLS video, while listeners in Apple Podcasts switch between watching and listening with a tap. It is for Blubrry hosting customers on Blubrry Publisher, and not for shows published through the PowerPress plugin.

Put the two together and the "video is a separate pipeline" excuse has collapsed a little further. One upload, one feed, every surface. The cost of being in more places keeps falling; the cost of being good in those places has not moved an inch, and if you are still deciding whether video belongs in your workflow at all, the questions in video podcast versus audio podcast remain the right ones.

Diary: Flightcast, New York and Kings Place

LiSTNR, the audio arm of SCA, Australia's largest commercial radio company, is moving more than 150 shows to Flightcast for distribution, reported by Podnews on 31 August. It is Flightcast's biggest customer to date; the company launched in October 2025, founded by Steven Bartlett and Rox Codes, and LiSTNR is leaving Triton's Omny Studio and adding video distribution in the move. Worth a moment's thought: Bartlett's infrastructure company is now landing enterprise contracts, which is a different business from The Diary of a CEO and arguably the more durable one. Shows have a lifespan; infrastructure contracts get renewed.

Podcast Movement announced its New York keynotes on 28 August. Tom Webster will preview The Podcast Landscape 2026 research from Sounds Profitable, and TINK Media's Lauren Passell and Arielle Nissenblatt will present "WTF is a podcast in 2026". The week runs 14 to 18 September 2026 at Terminal 5, per the event's own schedule: the Sounds Profitable Business Summit on 14 and 15 September, the IAB Podcast Upfront after party on 16 September, and Podcast Movement itself on 17 and 18 September.

Closer to home, PodcastingToday reports the London Podcast Festival returning to Kings Place in September with live and streamed shows across comedy, science, music, politics and audio drama, including Clara & Munroe Are Into It with Clara Amfo and Munroe Bergdorf, and the return of Audio Drama Day. If you have never watched a podcast recorded in front of a room, go; it teaches you more about pacing than any dashboard.

The numbers

The second row is the one to remember: on a podcast channel, roughly a third of what YouTube now calls a view never made it past the opening seconds, and that third is exactly what the next section is about.

From the edit desk: what makes a cold open actually work?

Read the lead story again with an editor's eyes. YouTube counts a view at the first frame and an engaged view only if someone stays past the first frame and the initial seconds. The cold open is the thing that converts one into the other. It is the single highest-leverage twenty seconds in an episode, and across well over a thousand edits it is the twenty seconds that has had the least attention paid to it, because it feels like packaging rather than content.

The commonest failure is starting with logistics and courtesy. The welcome, the thanks for coming on, the "before we get into it, tell everyone where they can find you". Meanwhile the best twenty seconds of the episode already exist, forty minutes in, where the guest said the thing that made the host sit up, and nobody moved them to the front. That move is an edit decision, not a recording decision, so every show that has already recorded can make it. Find the moment, lift fifteen to twenty-five seconds of it, and put it before anything else.

The other failures I see repeatedly:

• Summarising the episode instead of opening it. A cold open should raise a question the episode answers; a summary tells the listener they no longer need to stay.
• Music and a branded sting doing the job of an opening. They are not an open, they are a delay, and under the new definition a delay is measured.
• Credentials before curiosity. Nobody cares that the guest ran a division at anything until they have heard the guest say something worth caring about. Hook first, biography second.

A checklist you can run on the timeline before export:

1. Does the first sentence contain a specific claim, a number, or a moment? "We lost the account on a Tuesday" passes. "Today we're talking about client retention" fails.
2. Is the first voice the most interesting voice? Usually that is the guest, not the host.
3. Could the first twenty seconds stand alone as a short-form clip? If not, it is not strong enough to open with.
4. Have you cut everything before the first interesting word? Everything.

And a line to start the episode proper once the cold open has landed, so the host does not feel the welcome has been lost: "That was [guest], and we'll get to how that happened. First, who they are." Credentials land better when the listener already wants them. Building this open properly is most of what a retention intro is.

Which of this week's jobs actually takes ten minutes?

Two of them, and one that takes a little longer. Export your YouTube engaged views today, before the old baseline is gone, and quote that number, never the public one, to anyone who asks. Check your directory listings, since Buzzsprout has just shown that most shows are missing from somewhere, and confirm your own site carries a player no platform can switch off. Then the longer job: open your most recent episode and move the best twenty seconds to the front. If you would rather have someone else do that and everything after it, a full episode edit starts at £280, billed per four-week cycle with no minimum, and we revise until you are happy; get in touch whenever suits. Back next Thursday.