Podcasts don't grow on episodes — they grow on clips, and video ate the format

You put out 40 episodes. Each one took six hours to record and edit. The show sits at 200 downloads per episode. It has been there for eight months. You see people with ten episodes at 5,000 downloads each. You look for the difference. Someone tells you it is about consistency. You are consistent. Someone else says audio quality. You bought the mic.

No one says clips. No one says the thing your show does not have is a second product. The episode is not the thing that spreads. The clip is.

Here is what happened while you were making episodes. Video podcasts overtook audio. YouTube became the place more people go to listen than any other service. And 89% of podcast listeners now watch clips on social platforms. A third of them convert. The episode is still the asset. But the clip is the distribution. The format changed under everyone. Where the clip lands matters as well, since Instagram scores a reel, a carousel and a static post on different predictions.

We went looking for the numbers behind that. We read audience studies from Edison Research. They have tracked digital listening in the US since 1998. We read what YouTube and Spotify published about their own platforms. And we read a study on how clips work, from a firm called Sounds Profitable. The study was done with a research partner. It was called Signal Hill Insights. It measured 5,061 US adults.

Every claim below is linked, dated and quoted. One section reports where podcast discovery actually happens. It is YouTube. Not Instagram.

Video overtook audio in 2025

Start with the format that everyone thought podcasting was. Audio. People downloaded MP3 files or streamed them. They listened in the car or on a run. That was the whole picture for fifteen years.

It is not the picture now. Edison Research has run a study every year since the format existed. It is called The Podcast Consumer. The 2026 report carries a sentence no earlier report could print:

82% of weekly podcast consumers actively watching video compared to 78% listening to audio only — Edison Research, The Podcast Consumer 2026

Video is ahead. Not by much. But ahead. The crossover happened in the third quarter of 2025. Before that the two formats had been running close. Go back three years and the gap was the other way. In 2023 it was 89% audio against 73% video. Audio was still the default. By 2025 the default had flipped.

That is among weekly podcast consumers. People who listen or watch at least once a week. It is not everyone. But it is the audience that matters. The people who come back.

The Infinite Dial is a different Edison study. It runs every January. It has tracked digital media use since 1998. The 2025 report put a number on how many Americans have ever watched a podcast. Fifty-one percent. Just over half the country. For listening the number is 70%. So more people have tried audio than video. But the ones who stuck around now watch more than they listen.

A format does not flip like that by accident. Two things caused it. The first is that the platforms with the most reach all added video. The second is that video shows up in more places than audio does. A video podcast can be a YouTube video. It can also be a reel. Or a TikTok. Or a clip someone sends in a message. Audio does not move that way. You cannot post an MP3 to Instagram. You can post a clip of someone talking.

YouTube is where people listen now

The platform that won is not the one the format started on. Podcasts were born on RSS feeds. They moved to apps like Apple Podcasts and Spotify. Those apps were built for audio. YouTube was built for video. But YouTube is where the audience went.

Edison Research measured which service weekly podcast listeners use most often. The 2025 Infinite Dial report names the leader:

33% of U.S. weekly podcast listeners using the service — Edison Research, The Infinite Dial 2025, describing YouTube

One in three. That is the share of people who say YouTube is the service they use most for podcasts. Not sometimes. Most often. The next closest platform was Spotify. Among Gen Z weekly podcast listeners the two ran almost even. Thirty-six percent used YouTube most often. Thirty-five percent used Spotify.

YouTube itself put a bigger number on the table. On 26 February 2025 it announced it had passed one billion monthly active podcast users. That is its own count. Edison's number is independent. It comes from a survey of 5,020 people age 12 and up. Conducted in January 2025. The sample was weighted to match the US population. A mix of phone and online interviews. In English and Spanish.

YouTube also said viewers watched over 400 million hours of podcasts every month on living room devices. TVs. That is a format no audio-only app can touch. You do not listen to a podcast on your TV. You watch one.

Spotify has video podcasts too. It announced the feature's growth on 21 May 2026. Its number:

More than 500 million users have streamed a video podcast — Spotify, Investor Day announcement, 21 May 2026

That was up nearly 50% year over year. Spotify also said those users engage more and spend more time across the platform. Video is not a side feature anymore. It is how the format works now.

Eighty-nine percent of listeners watch clips on social platforms

The episode is an hour long. The clip is 90 seconds. The episode goes on YouTube or Spotify or Apple Podcasts. The clip goes everywhere else. TikTok. Instagram Reels. Facebook. X. YouTube Shorts. That is where people find the show.

Sounds Profitable is a podcast industry research site. It ran a study with Signal Hill Insights. Published 1 July 2026. They called it The Podcast Atlas. It measured 5,061 US adults age 18 and up. All of them had used at least one creator or media platform in the past 30 days. That was the screen. People who were already in range of seeing a podcast clip.

The study mapped five parts of how podcasting works now. Audio. Video. Clips. Social platforms. Newsletters. It put the creator at the center. Not any one platform. Then it asked how those five parts connect. The part that drives discovery is clips. The study printed the share of podcast listeners who watch them:

89% of podcast listeners watch clips on at least one social platform at least sometimes — Sounds Profitable, The Podcast Atlas, 1 July 2026

Eighty-nine percent. That is not early adopters. That is the whole listener base. Watching clips is what podcast listeners do now. Not instead of listening to full episodes. On top of it.

A follow-up piece ran on 15 July 2026. It broke the number down further. Seventy-two percent watch clips often or always. Then it listed the platforms. YouTube led. Sixty-seven percent of people who watch podcast clips watch them on YouTube often or always. TikTok was 63%. Facebook 62%. Instagram Reels 58%. X was last at 48%.

That order matters for anyone making clips. YouTube is not just where full episodes live. It is also where clips get the most watch time. On a platform built to serve more of what you just watched.

A third convert from clips to episodes

A clip gets watched. Does that turn into a listener? Sometimes. The Podcast Atlas measured conversion. It asked clip viewers what they did next. The answers:

Thirty percent said they often or always watch a specific episode after seeing a clip. Eighty-one percent said they do it at least sometimes. Thirty-three percent often or always become regular listeners after discovering a podcast through clips. Eighty-four percent do it at least sometimes.

So about one in three clip viewers becomes something. Either they watch the full episode the clip came from. Or they become a regular listener. The study describes clips as a discovery and conversion tool. Not a replacement for episodes. A way to recruit new people and bring back old ones.

One other number sits with those. Seventy-four percent of clip viewers give clips their full or near-full attention. For comparison, full audio episodes get 77% full attention. Clips are not background noise. People watch them the way they listen to episodes. That came from the same 15 July piece.

YouTube dominates discovery, and Instagram does not

Podcast discovery does not run on Instagram. It runs on YouTube, and two separate measurements say so.

Go back to the 2025 Infinite Dial. It asked Gen Z listeners where they discover podcasts most often. Twenty-eight percent said searching on YouTube. Twenty-six percent said social media posts. Social media as a whole category only just matched YouTube on its own. And YouTube was both the top listening platform and the top discovery platform for that group.

The numbers say YouTube is where podcasts grow. Instagram Reels took 58% of clip viewers. YouTube took 67%. YouTube also serves full episodes. And it is where people search when they want to find a new show. Instagram is part of the system. It is not the engine.

That order is worth holding on to when someone sells you a podcast growth plan that puts Instagram first. Ask them why the data says YouTube.

What the shift means for anyone sitting on long-form

You have 40 episodes. That is 40 hours of material. If the show is not growing, the mistake is usually not the episodes. It is that no one outside your current audience knows they exist. The episode is the asset. The clip is the distribution.

The format became a video medium. That does not mean you need a studio. It means people expect to see the person talking. Phone cameras work. The Sounds Profitable study calls video the deepening layer. It builds trust faster than audio. But clips are the discovery layer. They are what shows up in someone's feed when they were not looking for you.

The data says this: make clips. Put them on YouTube first. Then put them everywhere else. Measure which clips convert. Not which clips get views. Views do not matter if no one comes back. Thirty percent conversion is the benchmark from a sample of 5,061 people. If your clips are doing better, you have something. If they are doing worse, the clips are not working yet.

One other thing falls out of this. The episode length does not matter as much as people think. A two-hour episode and a 20-minute episode both produce clips. The long episode gives you more to pull from. The format is not dying because attention spans are shrinking. It is growing because clips made it visible.

What we could not determine

Three things we tried to measure and could not.

How long a clip should be. The Podcast Atlas study and the follow-up piece give no clip length. They say 89% of listeners watch clips. They do not say which lengths convert better. Ninety seconds is common. Sixty seconds fits most platform limits. Three minutes is long for a clip but short for an episode. We found no published data comparing them.

The shelf life of a clip versus a full episode. YouTube and Spotify both publish hours watched. Neither publishes a decay curve. How fast does a clip stop getting views? Do clips have a longer tail than episodes because the algorithm can serve them in more places? We could not find that measured anywhere.

Whether video podcasts perform better than audio-only podcasts in search. YouTube is a search engine. Audio-only podcasts are on YouTube too. As static images with audio. Do those get recommended as often as video podcasts? Do they rank as well in search? No published data. YouTube's 26 February 2025 announcement talks about over 1 billion users. It does not break that into video versus audio-only uploads.

How we did this

Scope: this post reads published research and first-party platform announcements. No data collection of our own. No experiment.

How we read them: every source was opened directly on 17 August 2026. Quotations are verbatim. Numbers come from the documents as published. Where a percentage is stated we report the percentage. Where a sample size is stated we report the sample size. Where a methodology is described we include it.

Limits: three of the seven sources are first-party. YouTube's announcement about its own user count. Spotify's announcement about its own video podcast users. Both are claims. Not audited figures. The Edison Research studies are independent. The Podcast Atlas is industry-funded. Its sponsors were SiriusXM Podcast Network, ESPN Podcasts, BetterHelp, NPR and American Public Media. Vendor research can carry bias toward the outcome the sponsor wants. We cite it because the sample size and method are disclosed. And because no other study we found measured clip conversion at all.

Sources

1. The Podcast Consumer 2026. Edison Research, 2026. Source of the 82% video versus 78% audio statistic, and the Q3 2025 crossover. Also the 2023 comparison showing 89% audio-only versus 73% video. No single publication date or sample size stated on the page accessed. Combines data from multiple Edison studies: The Infinite Dial, Share of Ear, Edison Podcast Metrics and Sports Poll by SSRS. Retrieved 17 August 2026.

1. The Infinite Dial 2025. Edison Research, January 2025. Sample: 5,020 individuals age 12 and up. Method: 1,006 telephone surveys and 4,014 online interviews, conducted in English and Spanish, weighted to US 12+ population. Source of the 33% YouTube usage figure, the 51% who have watched a podcast, the 70% who have listened, and the 73% who have consumed in either format. Also the Gen Z discovery data: 28% discover on YouTube versus 26% on social media posts. Retrieved 17 August 2026.

1. Top 10 Findings of 2025 from Edison Research. Edison Research, 2025. Source of the Gen Z platform preference data: 36% use YouTube most often, 35% use Spotify most often. Retrieved 17 August 2026.

1. YouTube: Over 1 billion monthly active podcast users. YouTube Official Blog, 26 February 2025. Source of the 1 billion monthly active podcast users claim and the 400 million hours watched monthly on living room devices. Retrieved 17 August 2026.

1. Investor Day: Podcast features and updates. Spotify Newsroom, 21 May 2026. Source of the 500 million users who have streamed a video podcast, and the nearly 50% year-over-year growth figure. Retrieved 17 August 2026.

1. The Podcast Atlas. Sounds Profitable, 1 July 2026. Research conducted by Signal Hill Insights. Sample: 5,061 US adults age 18 and up who used at least one ad-supported creator or media platform in the past 30 days. Sponsored by SiriusXM Podcast Network, ESPN Podcasts, BetterHelp, NPR and American Public Media. Source of the 89% clip viewing figure, the 74% attention rate for clips versus 77% for full episodes, and the statement that about a third of clip viewers convert to full episodes or regular listeners. Retrieved 17 August 2026.

1. The Front Door: How Podcast Clips Drive Discovery. Sounds Profitable, 15 July 2026. Author: Tom Webster. Based on The Podcast Atlas study above. Source of the 72% who watch clips often or always, the platform-specific percentages (YouTube 67%, TikTok 63%, Facebook 62%, Instagram Reels 58%, X 48%), and the conversion rates: 30% often/always watch a specific episode after a clip, 33% often/always become regular listeners. Retrieved 17 August 2026.

Changelog

• 2026-08-18: First version drafted.