Average Podcast Episode Length: Statistics & Completion
The median podcast episode runs 38 minutes and 42 seconds, based on Pacific Content's analysis of 10 million episodes. The mean is higher, around 43 minutes, because a handful of multi-hour shows drag the average up. Half of all episodes land between 20 and 40 minutes (Buzzsprout). That is the answer to "how long is normal." The harder question is how long yours should be, and for that you need completion data, not averages.
This page does both. First, the length numbers, each tagged to a named source. Then the completion evidence, including one widely repeated stat that turns out to have no primary source, which is worth knowing before you cite it.
What is the average podcast episode length?
The average podcast episode is 38 to 43 minutes, depending on whether you measure the median or the mean. Pacific Content's 10-million-episode study found a median of 38m42s and a mean of 43m24s. The gap between the two matters: the median tells you what a typical episode looks like; the mean is pulled upward by outliers like a single 11-hour episode.
Always ask which one a source quotes. A blog citing "the average podcast is 43 minutes" is technically right and practically misleading, most shows are shorter than that, and a few giants inflate it. The median is the number to plan around.
Pacific Content's follow-up the next year widened the sample to 18.8 million episodes across 637,793 distinct shows and found the mean had dropped to 41m31s and the median to 36m34s (Pacific Content, via The Podcast Host). Same researcher, bigger sample, one clear signal: episodes were getting shorter.
| Measure | Figure | Source (sample) |
|---|---|---|
| Median, all shows | 38m42s | Pacific Content (10M episodes) |
| Mean, all shows | 43m24s | Pacific Content (10M episodes) |
| Median, larger sample | 36m34s | Pacific Content (18.8M episodes) |
| Mean, larger sample | 41m31s | Pacific Content (18.8M episodes) |
| Most common band | 20–40 min (31% of episodes) | Buzzsprout (May 2026, ~114k shows) |
How are episodes distributed by length?
Most episodes cluster in a narrow band. Buzzsprout's May 2026 data across roughly 114,000 active shows puts 31% of episodes in the 20–40 minute range, the single largest group, with the rest spread thin above and below (Buzzsprout stats). That 31% has held steady on the same band across Buzzsprout's last several monthly updates, so it is a stable shape, not a one-month blip.
A useful contrast: when The Podcast Host asked 3,400+ aspiring podcasters to plan a show, 52% picked the 20–40 minute band as their intended length (The Podcast Host). That is a planning intention, not a measured episode, so it is not directly comparable, but the fact that both real episodes and stated intentions converge on 20–40 minutes tells you the norm is deeply baked in.
So when people say "20 to 40 minutes is the sweet spot," they are describing where the herd already is, not where performance peaks. Those are different claims, and the difference is the whole point of this page.
Do chart-topping shows run longer?
Yes, popular shows skew noticeably longer than the field. In Pacific Content's data, the Top 400 charting shows had a median of 47m30s and a mean of 53m04s, versus 38m42s and 43m24s for all episodes (Pacific Content). The chart-toppers run roughly nine to ten minutes longer at the median.
Resist the obvious read. The researcher who ran the numbers, Dan Misener, said it plainly: those shows chart because they are high quality and well marketed, "not because they're longer than average." Length correlates with success here; it does not cause it. A 90-minute episode of a thin show is still a thin show, just a longer one.
What is a good podcast completion rate?
A completion rate above 70% is generally treated as good, with 90%+ considered excellent and 50–69% average (PodRewind, citing Spotify and industry benchmarks). Podcasts finish at much higher rates than scrollable video because listening is a committed, background activity, not something you swipe past, but no platform publishes a single market-wide completion benchmark, so treat any tidy number with suspicion.
Here is where honesty matters more than a clean number. The most repeated podcast stat online is "72% of listeners finish an episode in its entirety, the highest of any media format," almost always credited to Edison Research. I went looking for it in Edison's primary reports and could not find the exact 72% figure in their published materials, Edison's Podcast Metrics measures actual listening behavior across platforms, not a stated "finish in entirety" percentage, and the 72% line shows up only in second-hand marketing posts that cite Edison without linking a primary report. The "70%+ finish most or all" framing is well supported across analytics platforms; the precise "72%, highest of any format" line is a secondary-source claim that has detached from a verifiable origin. Treat it as directional, not gospel.
| Completion rate | Rating | Note |
|---|---|---|
| 90%+ | Excellent | True crime and fiction reportedly reach this (analytics summaries citing Spotify) |
| 70–89% | Good | The benchmark most shows aim for |
| 50–69% | Average | Common for longer or interview-heavy formats |
| Below 50% | Needs work | Check your first 30–60 seconds first |
Does a shorter episode get finished more often?
Probably, but the clean cross-length completion numbers you will see online are weaker than they look. The intuitive claim, short episodes complete at higher rates, long ones at lower, is directionally sound and supported by analytics-platform guidance (PodRewind). The widely circulated specific pairing ("85% completion at 20–40 minutes, dropping to 60% past an hour") does not trace to a primary study, so I will not present it as a sourced fact.
What is well established: completion rate and total minutes consumed pull in opposite directions. A tight 25-minute episode finished by 80% of listeners delivers 20 minutes of attention each. A 70-minute episode finished by 55% still delivers ~38 minutes. Neither is "better" in the abstract, it depends on whether you are optimizing for reach (completion) or depth (minutes). Decide which one your show is for, then pick a length that serves it.
A length decision rule you can actually use
Stop guessing from averages. Use this instead:
- Pull your own retention curve. Spotify for Creators, Apple Podcasts Connect, and YouTube all show where listeners drop off. Your real data beats any benchmark.
- Find your cliff. Most shows have one obvious drop-off point. If half your audience is gone by minute 30, a 50-minute episode is mostly talking to no one.
- Cut to just past the cliff, not to a number you read online. If your audience holds to minute 35, a 30–40 minute episode fits them. If it holds to minute 70, run long, you have earned it.
- Re-check after every five episodes. The cliff moves as your audience and format change.
The median is 38m42s. Your audience's cliff is the only length statistic that should change what you record next week.
Why length data and clips belong in the same conversation
The episode-length debate assumes one mode of consumption: someone presses play and either finishes or bails. That is no longer how most people meet a show. A growing share of new listeners now discover podcasts by watching a short clip first, a 30-to-90-second moment lifted from a full episode, and only then decide whether to commit 40 minutes (video podcast statistics).
That reframes the length question. The clip does the recruiting; the episode does the relationship. A 60-minute episode is not "too long" if the clip earned the listener's trust before they hit play. You can run the length your content deserves and use clips to lower the cost of someone trying you out. For the wider numbers on how that discovery layer works, see our breakdown of the podcast clipping industry and how the clipping economy actually works.
FAQ
What is the ideal podcast episode length? There is no single ideal. The median episode is 38m42s (Pacific Content), and most shows run 20–40 minutes (Buzzsprout), but the right length is wherever your own retention curve says listeners stay. Match the length to the content, then verify with your drop-off data, not to a benchmark.
Why is the average podcast longer than the median? Because the mean is sensitive to extremes. A few multi-hour episodes (Pacific Content noted an 11-hour outlier) pull the average up, while the median, the middle value, is not distorted by them. The median is the better number for planning a typical episode.
Do longer podcasts get more downloads? There is no reliable evidence that length drives downloads. Chart-topping shows run longer on average (median 47m30s vs 38m42s, per Pacific Content), but the researcher behind that data attributes their success to quality and marketing, not duration. Length correlates with success; it does not cause it.
Is a 70% completion rate good? Yes. Above 70% is generally rated good and 90%+ excellent, with 50–69% average (PodRewind). Podcasts complete far better than short-form video, where ~30% completion is a typical benchmark, because listening is a committed, background activity rather than something you swipe past.
Should I shorten my episodes to improve completion? Only if your retention data shows a clear early drop-off. Shorter episodes tend to complete at higher rates, but longer ones can deliver more total minutes consumed. Cut to just past where your audience actually drops, not to a number you read in a roundup.
Caveats and method notes
- Length figures skew indie. Buzzsprout and similar hosts represent a single-digit share of all podcasts; Spotify and Apple publish no comparable distribution. The 20–40 minute concentration is real but reflects mostly independent shows.
- The completion landscape is fragmented. No platform shares a market-wide completion benchmark. Genre figures (true crime and fiction at 85%+) come from analytics-platform summaries citing Spotify, not a single primary dataset.
- The "72%" stat is unverified at the source. It is widely attributed to Edison Research but does not appear in their primary published reports that I could locate. Cite "70%+ finish most or all" instead, which is broadly supported.
Every figure here names its source and its sample. For the full picture of the industry these numbers sit inside, start with our 2026 podcast statistics roundup, then how many podcasts are there and how many people listen.