How many clips can one episode give you?

Every host asks this question with a number already in mind, usually a large one. Ten clips per episode. One a day. Enough to never think about social again. It is an appealing arithmetic: you recorded an hour, a clip is a minute, therefore an hour contains sixty clips.

It does not. An hour of conversation contains a handful of moments that survive being taken out of it, and the number varies episode to episode in ways you cannot plan around. Here is what actually determines the count, based on cut sheets from real shows rather than on what a repurposing calculator would like to be true.

How many clips can one episode actually produce?

Three to five for most shows, most weeks. That is the range we work to across our clients, and it holds because it is what a normal hour of talking supports without scraping.

Eight is achievable, but the reason is structural rather than editorial. On one two-host business episode we cut, the editor's brief planned exactly eight shorts, split four per host, each with its own hook line and its own timestamp range. Eight was possible because there were two people in the room who were both worth clipping, and the episode gave each of them four moments where they said something complete. A single host interviewing a nervous guest does not produce that, and pushing it to eight anyway just means four good clips and four that dilute them.

One is also a legitimate answer. If an episode contained exactly one moment that genuinely stands alone, publish that one. A feed of one strong clip a week beats a feed of five where two are strong and three are people saying "yeah, exactly".

The trap runs in the other direction too. One long-running show we reviewed had sixty-six episode folders in its archive, and across all of them only three genuinely vertical short-form clips had ever been finished. There was no shortage of intent: dozens of files were labelled "for shorts", but they were raw multi-camera source, between 655MB and nearly 2GB each, sitting exactly where they were exported and never cut into anything. Gathering material is not producing clips. That gap, between the raw stuff marked for later and the finished thing on a feed, is where most repurposing plans quietly die.

What makes a moment clippable?

An editor watching for clips is really asking three questions, and a moment has to pass all three.

Does it stand alone? A stranger arrives with no idea who is speaking or what came before. If understanding the point requires the previous four minutes, it is not a clip, however good the point is. This is the test most moments fail, and it is why the smartest thing in an episode is often unclippable while a smaller, self-contained observation travels for weeks.

Does it start hot? The first two seconds decide everything. A clip that opens with "so, I suppose the way I'd frame that is" has already lost. Editors either find a moment that begins on the substance or trim into it so it does. This is also the mechanical reason automatic clipping tools disappoint: they cut at sentence boundaries in a transcript, which means they routinely open on the throat-clearing sentence instead of the one worth hearing.

Is there a reason to still be there at second three? Tension, a number, a claim you disagree with, a question left hanging. Something that makes stopping feel like a decision rather than an accident.

Length follows from the moment rather than the other way around. A hook clip that lives or dies on one line runs twenty to forty seconds. A segment clip, where the value is a full explanation, can run much longer: one recent clip we cut came in around 106 seconds and works precisely because the whole argument is there. Platforms tolerate both. What they do not tolerate is a forty-second clip containing twenty seconds of material.

Where do the best clips hide?

Not usually where the host expects.

The recap. One interview show we work with ends every episode with a short analysis exchange between the host and a colleague, unpacking the conversation that just happened. Going through that show's transcripts, the recap consistently produces the most polished, most liftable lines in the entire episode. It makes sense once you see it: the speakers are summarising a point they have just spent an hour arriving at, so they say it better the second time. Almost no host thinks to clip their own outro.

The tangent. The bit where someone says "this is slightly off topic, but" is very often the most human ninety seconds in the file, and it is self-contained by definition.

The answer that nearly got cut. When a guest hesitates and then says the honest version, that is the clip. It rarely looks like the clip in a transcript, which is another reason a human watching the episode finds things a transcript scan cannot.

Clips also do not have to respect episode boundaries. On one show, a single finished short was assembled from material spanning two different episodes, because the same theme came up twice and the two halves of the point lived in different weeks. If a subject recurs, the archive is fair game.

If you want the wider picture of everything an episode can become beyond short-form, we set out the full inventory in one episode, fifteen assets. This post is deliberately narrower: it is only about the clips.

How should you space clips out?

Drip them. Three to five clips released across a week does more than the same clips dumped on release day, because a feed rewards showing up repeatedly and because your audience is not all online on Tuesday morning.

A simple rhythm that works: the strongest clip on the day the episode drops, one mid-week, one at the weekend when feeds are quieter and competition is lower. Anything left over goes into a bank for a week when recording did not happen. Every show has those weeks, and a small bank is the difference between a gap and a normal week.

Two practical notes on volume and cost, since they interact. Clips are priced individually with us, £35 for a standard cut and £65 for an advanced one with motion captions, B-roll and sound design, so you are choosing volume consciously rather than buying a bundle you may not fill. That matters more than it sounds: a fixed monthly quota is exactly the pressure that produces filler. You can see what separates the two builds, with the same moment cut both ways, on the short-form clips page.

And flag your own favourites. Clients tell us the timestamps they loved while recording, and those jump the queue. You were in the room. You know which moment made the other person sit up, and no transcript records that.

Where do you go from here?

Take your last episode and try to find three moments that pass the standalone test. If you find three easily, your show supports five. If you struggle to find one, the problem is upstream in how the conversation is structured, and no amount of clipping will fix it.

Then decide the count deliberately rather than by default. If you would like us to cut them, spec the number on the configurator and change it whenever the show changes. And if you are cutting them yourself, keep the standard high and the quota low. Nobody ever grew a show by publishing the fourth-best minute of the week.