"How long should my episodes be?" is one of the first questions every new podcaster asks, and it gets answered badly almost every time — usually with a confident number and no reasoning.
Here's what the data actually shows, where it contradicts itself, and why length is probably not the thing you should be optimising.
The Numbers
The average episode is around 38 minutes. Buzzsprout data puts the typical podcast at roughly 38 minutes; broader estimates cluster between 30 and 42 minutes. Among the most popular shows, the average sits at about 38 minutes 42 seconds.
The distribution is wide. Roughly 31% of podcasts fall in the 20–40 minute band — the single biggest group, but nowhere near a majority. About 55% of podcasts run over 30 minutes.
Successful shows skew longer. This is the finding that undercuts most advice: around 78% of successful shows average 40 minutes or more.
The retention "sweet spot" is quoted as 25–40 minutes — long enough for depth, short enough to finish in one commute, which does genuinely improve completion rates.
The Contradiction Nobody Points Out
Read those last two together:
- The retention sweet spot is said to be 25–40 minutes.
- But 78% of successful shows run 40+ minutes.
These can't both be straightforward advice. The resolution is that completion rate and success are different things, and optimising for the first can cost you the second.
A 25-minute episode is easier to finish. But shows that succeed tend to be ones people want more of — and that want usually comes from depth, which takes time. Acquired's multi-hour episodes have terrible completion rates by any normal standard and an extraordinarily devoted audience.
There's also a survivorship issue running the other way: longer shows may succeed because hosts who can sustain 60 interesting minutes are simply better, not because length itself causes success.
So: length correlates with outcomes. It almost certainly doesn't cause them.
The Actual Rule
Length should be an output of your format, not an input.
| Format | Natural length |
|---|---|
| Daily news | 10–25 min |
| Solo commentary / explainer | 15–30 min |
| Conversational interview | 45–75 min |
| Deep-dive research / narrative | 60 min+ |
| Panel / co-hosted chat | 45–60 min |
| Fiction / audio drama | 20–40 min per episode |
Pick the format that suits your material, then let the episode run as long as it's genuinely good. The failure mode isn't "too long" — it's padding, which is what happens when you have a target length and insufficient material.
The honest test: would you cut this section if nobody had told you episodes should be 40 minutes? If yes, cut it.
What Actually Drives Retention
Here's the finding that matters more than any duration number: consistency beats length optimisation. A predictable schedule reduces the decision cost that makes listeners skip or defer episodes — and deferred episodes become the backlog people never clear.
Ranked by actual impact on whether people keep listening:
- Publishing consistently — the single strongest driver, and the one 85% of shows fail at.
- A strong first three minutes. Most drop-off happens early. If your show opens with 90 seconds of housekeeping, that's your retention problem, not your runtime.
- Whether every minute earns its place. A tight 70 minutes beats a padded 30.
- Chapters. They let listeners navigate long episodes instead of abandoning them, and they improve YouTube retention. (Guide here.)
- Length. Last, and some distance behind the others.
Practical Guidance
If you're starting out: aim for 30–45 minutes, but treat it as a loose frame rather than a target. Under 20 usually means you haven't developed the idea; over 90 in your first ten episodes usually means you haven't learned to edit yet.
If you're established: stop thinking about length entirely and check your first three minutes instead. That's where the recoverable audience is.
If episodes are ballooning: the problem is almost always structure, not discipline. Plan three to five segments before recording and the runtime tends to sort itself out.
If you're inconsistent: fix that before anything else. A reliable 25-minute show beats a brilliant 60-minute one that appears unpredictably.
The Production Reality
One genuine argument for shorter episodes that nobody makes: long episodes cost more to publish. A 90-minute episode means more editing, longer show notes, more chapters, more clip-hunting. That's the load that kills most shows.
If long-form suits your material, protect it by cutting the post-production cost instead of the runtime — DriftNote's creator tools generate show notes, descriptions, and timestamped summaries that double as a clip shortlist, so a 90-minute episode doesn't cost you a whole evening to ship.
The One-Line Answer
As long as it's good, and not one minute longer.
Unsatisfying, but it's the only answer the data supports. Every specific number you'll see quoted is an average across wildly different formats, and your show is one format.
Where to Go From Here
- Try DriftNote's creator tools
- How to start a podcast in 2026
- How to grow a podcast in 2026
- Podcast statistics 2026: 40+ numbers that matter
Figures reflect 2026 industry reporting and vary by source and methodology — treat them as direction, not precision.