These are the podcast figures we saw quoted most often on other sites. We followed each one back to the survey or report it came from. For each you get what the source says, who it measured and when, and a line you can quote instead. We also counted a few things ourselves from the Podcast Index database of podcast feeds and from one fetch of a random sample of those feeds, and show the query behind each count.
Last checked September 27, 2026. 22 quoted figures: 11 check out, 4 are wrong or misread, and 7 have no primary source we could find. Plus 5 counts of our own.
Our own counts
These come from the Podcast Index database snapshot we hold and from one fetch of a random sample of its feeds. Each shows its query or procedure and when it was measured, so you can run it again.
Our own count: 10.3% of feeds listed in our Podcast Index snapshot have a newest episode dated in the 90 days before it
The Podcast Index database snapshot we hold lists 4,603,628 podcast feeds. Its latest update is 2026-02-01 07:49:55 UTC. Of those Podcast Index feeds, 474,698 (10.3%) have a newest episode dated in the 90 days before that moment.
Query, run on the SQLite snapshot:
WITH s AS (SELECT MAX(lastUpdate) AS t FROM podcasts) SELECT COUNT(*), SUM(newestItemPubdate > s.t - 90*86400 AND newestItemPubdate <= s.t), SUM(newestItemPubdate > s.t - 90*86400 AND newestItemPubdate <= s.t AND lastHttpStatus <> 200) FROM podcasts, s;
The date is the feed’s own newest-episode date. We counted feeds whether or not Podcast Index’s last fetch succeeded: 4,889 of them returned something other than HTTP 200 on that fetch. Podcast Index lists every feed it has ever seen, including shows that stopped years ago, so this is a share of listed feeds, not of shows people listen to.
Where feeds with a recent episode are served from
Of the 474,698 feeds in the Podcast Index snapshot with a new episode in the prior 90 days, 23.2% have a feed URL on anchor.fm, 11.5% on buzzsprout.com and 10.7% on spreaker.com. The next two in the same Podcast Index count are podbean.com at 5.6% and rss.com at 4.1%.
Query or procedure (Snapshot 2026-02-01 07:49:55 UTC):
WITH s AS (SELECT MAX(lastUpdate) AS t FROM podcasts)
SELECT lower(host) AS host, COUNT(*) AS feeds,
ROUND(100.0 * COUNT(*) / SUM(COUNT(*)) OVER (), 1) AS pct
FROM podcasts, s
WHERE newestItemPubdate > s.t - 90*86400 AND newestItemPubdate <= s.t
GROUP BY lower(host) ORDER BY feeds DESC LIMIT 5;
Domain = Podcast Index's own `host` field for the feed (the registered domain of the feed URL, e.g. anchor.fm for anchor.fm/s/…/podcast/rss), lower-cased so Podbean.com and podbean.com count once. It is the domain the feed is served from, which for a custom-domain feed is not the hosting company. One feed has an empty host and counts in the denominator.
Episode counts of feeds with a recent episode
The median feed in the Podcast Index snapshot with a new episode in the prior 90 days lists 45 episodes, across all 474,698 such feeds. In the same Podcast Index count, 21.8% list fewer than 10 episodes, 30.2% list 10 to 49, 14.3% list 50 to 99, 19.8% list 100 to 249, 12.1% list 250 to 999, and 1.7% list 1,000 or more.
Query or procedure (Snapshot 2026-02-01 07:49:55 UTC):
WITH s AS (SELECT MAX(lastUpdate) AS t FROM podcasts),
a AS (SELECT episodeCount AS n FROM podcasts, s
WHERE newestItemPubdate > s.t - 90*86400 AND newestItemPubdate <= s.t)
SELECT (SELECT AVG(n) FROM (SELECT n FROM a ORDER BY n
LIMIT 2 - (SELECT COUNT(*) FROM a) % 2
OFFSET (SELECT (COUNT(*) - 1) / 2 FROM a))) AS median,
COUNT(*) AS feeds, SUM(n < 10) AS under_10,
SUM(n BETWEEN 10 AND 49) AS n10_49, SUM(n BETWEEN 50 AND 99) AS n50_99,
SUM(n BETWEEN 100 AND 249) AS n100_249, SUM(n BETWEEN 250 AND 999) AS n250_999,
SUM(n >= 1000) AS n1000_plus
FROM a;
Episodes = Podcast Index's episodeCount field for the feed. The snapshot's highest value is 1,000 and 7,977 feeds sit exactly there, so the top bucket reads "1,000 or more"; the median is unaffected (source).
Sampled recent feeds with a transcript tag
We fetched a random sample of 2,000 of the 474,698 feeds that the Podcast Index snapshot lists with a new episode in the 90 days to February 1, 2026. Of the 1,881 sampled Podcast Index feeds we could fetch on September 28, 2026, 325 (17.3%, ±1.7 points (95% binomial sampling margin for the fetched feeds; doesn't cover the 119 feeds that failed to fetch)) carry a <podcast:transcript> tag.
Query or procedure (Sample drawn from the snapshot of 2026-02-01 07:49:55 UTC; feeds fetched 2026-09-28 03:04 to 03:32 UTC):
Sample: the ids of all 474,698 feeds with a newest episode in the 90 days
before the snapshot, sorted ascending, then a seeded
partial Fisher-Yates shuffle (mulberry32, seed 20260927); the first 2,000.
Fetch: each feed URL once, 2 at a time, 10 s timeout, robots.txt respected.
Success: HTTP 200 and an RSS document.
Transcript: the document contains a <podcast:transcript> tag.
Video: the enclosure on the item with the newest pubDate has a video/* MIME type,
or, when its type is missing or not audio/* or video/*, a .mp4 .m4v .mov .webm .mkv or .avi file.
119 of the 2,000 sampled feeds could not be fetched (HTTP errors, robots.txt refusals, DNS failures or non-RSS responses) and are left out, so the share is of the 1,881 that answered (source). The margin of error is the normal approximation, 1.96 × √(p(1−p)/n) (source). The tag marks a transcript file linked from an episode; a show can publish transcripts elsewhere without it.
Sampled recent feeds whose newest episode is a video file
In the same sample of Podcast Index feeds, 14 of the 1,845 feeds with an enclosure on their newest episode (0.8%, ±0.4 points (95% binomial sampling margin for the fetched feeds; doesn't cover the 119 feeds that failed to fetch)) carried a video file rather than audio.
Query or procedure (Sample drawn from the snapshot of 2026-02-01 07:49:55 UTC; feeds fetched 2026-09-28 03:04 to 03:32 UTC):
Sample: the ids of all 474,698 feeds with a newest episode in the 90 days
before the snapshot, sorted ascending, then a seeded
partial Fisher-Yates shuffle (mulberry32, seed 20260927); the first 2,000.
Fetch: each feed URL once, 2 at a time, 10 s timeout, robots.txt respected.
Success: HTTP 200 and an RSS document.
Transcript: the document contains a <podcast:transcript> tag.
Video: the enclosure on the item with the newest pubDate has a video/* MIME type,
or, when its type is missing or not audio/* or video/*, a .mp4 .m4v .mov .webm .mkv or .avi file.
Same sample and fetch as the transcript count. 36 fetched feeds had no enclosure on their newest item and are left out (source). This counts video files inside the RSS feed only: a show that posts video to YouTube and audio to its feed counts as audio here.
Quoted as“73% of Americans have ever consumed a podcast”
What the source says
"73% of those ages 12+ have ever consumed a podcast" (Source: Edison Research, The Podcast Consumer 2025, July 23, 2025). The base is Americans age 12 and older, not all Americans.
Measured: U.S. population age 12+.
Edison's newer Infinite Dial 2026 puts the "ever" measure at 80% (see below). Quote the year with the figure.
Quoted as“55% of Americans are monthly podcast listeners”
What the source says
"55% have consumed a podcast in the last month," among Americans age 12+. Edison counts people who consumed a podcast, which includes watching, so it is not a count of listeners (Source: Edison Research, The Podcast Consumer 2025, July 23, 2025).
Measured: U.S. population age 12+.
The 2026 figure for the same measure is 58%, from The Infinite Dial 2026.
Quoted as“40% of Americans consume podcasts weekly”
What the source says
"40% have consumed a podcast in the last week," among Americans age 12+ (Source: Edison Research, The Podcast Consumer 2025, July 23, 2025).
Measured: U.S. population age 12+.
Quoted as“YouTube has 1 billion monthly podcast users”
What the source says
The headline says "users", but the post itself says "more than 1 billion monthly active viewers of podcast content on YouTube", as of January 2025 (Source: YouTube Official Blog, Drop the mic: Celebrating 1 billion monthly podcast users on YouTube, February 26, 2025). That counts anyone who watched podcast content, not podcast listeners or dedicated podcast users.
Measured: Monthly active viewers of podcast content on YouTube, worldwide, as of January 2025.
It is YouTube's own figure, worldwide, and it counts viewers of podcast content on YouTube only.
Quoted as“51% of Americans have watched a video podcast”
What the source says
"Over half (51%) of the U.S. 12+ population has ever watched a video podcast" (Source: Edison Research, The Podcast Consumer 2025, July 23, 2025).
Measured: U.S. population age 12+.
Keep "ever" and "age 12+". This is not a weekly or monthly figure.
Quoted as“584.1 million people worldwide will listen to podcasts in 2025”
We couldn't trace this to a primary source
Roundups credit this to eMarketer, but none links to an eMarketer page or report with the figure, its definition or its method, and we found none.
Delete it. We have no sourced replacement for a worldwide listener count.
Quoted as“80% of listeners finish all or most of an episode”
We couldn't trace this to a primary source
We found no survey or platform study behind this figure, with a date, a population and a question wording.
Delete it. Completion rates depend on the show and how "finish" is measured.
Quoted as“167 million U.S. monthly podcast consumers (58% of age 12+)”
What the source says
"58% of Americans age 12+ (167 million) have consumed a podcast in the last month" (Source: Edison Research, The Infinite Dial 2026, March 12, 2026).
Measured: U.S. population age 12+ (national survey of 2,050 people) (source).
Quoted as“The global podcast market is worth $39.63 billion in 2026”
We couldn't trace this to a primary source
Roundups credit $39.63 billion to Grand View Research. Grand View's current page gives $50.8 billion for 2026; we couldn't find $39.63 billion in any source we checked.
Market-size figures are one firm's estimate, not a measured industry total, and firms disagree widely. Name the firm and the year.
Quoted as“80% of Americans have consumed a podcast”
What the source says
"80% of Americans age 12+ (230 million) have ever listened to or watched a podcast" (Source: Edison Research, The Infinite Dial 2026, March 12, 2026).
Measured: U.S. population age 12+ (national survey of 2,050 people) (source).
This is the "ever" measure. It is not 80% monthly, and it is not the 2025 figure (73%, from The Podcast Consumer 2025).
Quoted as“400 million hours of podcasts are watched on TVs every month”
What the source says
YouTube's December 2025 update says viewers "watched over 700 million hours of podcasts on living room devices in October 2025, up from 400 million hours from just a year prior" (Source: YouTube Official Blog, How podcasts took over the living room in 2025, December 18, 2025). That dates the 400 million hours to October 2024 (source).
Measured: YouTube viewing on living-room devices, worldwide.
Quoted as current, 400 million hours is a year out of date.
Quoted as“YouTube has 39% of podcast listening, Spotify 20% and Apple Podcasts 11%”
What the source says
"% Using [Service] Most Often For Podcasts", Q4 2025, among U.S. weekly podcast consumers age 13+: YouTube 39%, Spotify 20%, Apple Podcasts 11% (Source: Edison Research, The Infinite Dial 2026 (presentation), March 12, 2026).
Measured: U.S. weekly podcast consumers age 13+, Q4 2025.
It is the service people use most often, not a share of listening. The same Infinite Dial 2026 presentation gives the share of daily podcast time by source for Americans age 13+ in Q4 2025: YouTube 32%, Spotify 25%, Apple 20%.
Quoted as“50.6% of podcasts publish full video episodes on YouTube”
Podtrac's figure describes the biggest shows. It says nothing about the share of all podcasts on YouTube.
Quoted as“59% of Gen Z consume podcasts on YouTube”
We couldn't trace this to a primary source
We found no survey that reports this. The March 2025 version of Transistor's Gen Z podcast survey (102 respondents aged 13 to 29) had a 59% figure, but it was the share who most often listen at home on a phone or tablet, not on YouTube.
Delete it. Transistor updated its report on August 13, 2025 with Coleman Insights data and no longer shows the 59% figure; the March 2025 version survives only in the archive Transistor links to.
Quoted as“80% of U.S. podcast consumers use both audio and video”
What the source says
Of the monthly podcast audience age 18+: listen only 13%, watch only 7%, "Watch & Listen" 80% (Source: Triton Digital, 2025 U.S. Podcast Report, February 2026).
Measured: Monthly U.S. podcast listeners age 18+, Triton Podcast Metrics Demos+ survey, Q2 to Q3 2025, 6,000+ respondents (source).
The base is monthly podcast listeners age 18+, not all Americans.
Quoted as“U.S. podcast ad revenue reached $2.86 billion in 2025, up 17.6%”
Quoted as“Podcast listening time rose 355% in ten years”
What the source says
Citing its Share of Ear study, Edison says Americans spent "a total of 170 million weekly podcast listening hours" ten years ago and now "spend a total of 773 million hours per week—an increase of 355%" (Source: Edison Research, How new consumers are shaping podcasting's next chapter, January 27, 2026). The Podcast Consumer 2025 gives the base as Americans age 13+ and the start year as 2015.
Measured: Americans age 13+, total weekly podcast listening hours, Edison's Share of Ear study.
It is total weekly listening across the population, not how long each listener spends.
Quoted as“Podcasts now account for 40% of spoken-word listening”
What the source says
"As of Q4 2025, 40% of time spent listening to spoken-word is now spent with podcasts and 39% of time is spent with AM/FM radio" (Source: Edison Research, Podcasts lead AM/FM in spoken-word listening, marking a first, February 24, 2026).
Measured: Americans age 13+, share of time spent with spoken-word audio, Q4 2025.
The denominator is spoken-word audio only, not all audio.
Quoted as“54% of U.S. adults listened to a podcast in the past year”
What the source says
"Just over half of U.S. adults (54%) say they've listened to a podcast in the past 12 months" (Source: Pew Research Center, Podcasts and News Fact Sheet, September 25, 2025).
Measured: U.S. adults, survey conducted August 18 to 24, 2025.
Pew surveys adults 18+ and asks about the past 12 months, so it is not comparable with Edison's 12+ monthly figures.
Quoted as“42% of weekly podcast consumers use YouTube most”
What the source says
Asked which platform they use most for podcasts, "42% indicate YouTube, up from 15% in July 2019" (Source: Cumulus Media and Signal Hill Insights, Podcast Download, Fall 2025 Report, November 17, 2025). The survey covered 603 weekly podcast consumers, fielded October 8 to 21, 2025.
Measured: 603 U.S. weekly podcast consumers, surveyed October 8 to 21, 2025.
A different survey from Edison's 39% in The Infinite Dial 2026. Don't combine the two.
Quoted as“There were 464.7 million podcast listeners worldwide in 2023”
We couldn't trace this to a primary source
Every page we found repeats it from another roundup. None names the organisation that produced it or how "listener" was defined.
Delete it. We have no sourced replacement for a worldwide listener count.
Quoted as“The podcast market will reach $104.97 billion by 2028”
We couldn't trace this to a primary source
The trail ends at a 2024 Patreon post that mentions the figure without naming the forecast or its publisher. We found no market report that contains it.
Delete it, or replace it with a named firm's current estimate, such as Grand View Research's.