How many clips should you cut from one episode?

The common pattern is one clip per episode, posted the week it comes out, and then the episode is done forever. That is almost always leaving something on the table.

The expensive part is finding a good episode

Recording, guest-wrangling, editing and publishing a two-hour episode is days of work. Finding the second clip inside an episode you have already recorded is minutes.

So the marginal cost of clip two, three and four is close to nothing, while the marginal cost of a whole new episode is enormous. If you only cut one clip per episode, you are paying the large cost repeatedly to avoid the small one.

How many are actually in there

A good two-hour conversation usually contains somewhere between four and eight passages that genuinely stand on their own. Not thirty. The claim that any episode contains twenty viral clips is a claim made by people selling clip volume.

Below that number you are leaving good moments unused. Above it you start posting excerpts that need context, and those perform badly enough to drag down how your account gets shown to people.

Four to eight from a strong episode. One from a weak one. Zero is a legitimate answer.

Not every episode is a clipping episode

Some episodes are good listens and contain nothing self-contained — a slow build, a long technical walkthrough, a conversation that only works cumulatively. Forcing three clips out of it produces three bad clips.

Being willing to cut nothing from an episode is what lets you cut six from the next one without it feeling inconsistent.

Spread them out

There is no reason a clip has to be posted the week its episode came out. Nobody is checking. An episode from four months ago is new to almost everyone who sees the clip, and a back catalogue is a supply of posts that costs nothing more to produce.

If you are starting a clipping habit, the best first move is usually to go back through the archive rather than wait for the next recording.

What this means for how we price

This reasoning is why Keptbits charges per clip rather than per minute of video. If the price depended on source length, you would be penalised for exactly the behaviour that works — going back to a long episode and taking several clips out of it.

One clip, one token, whatever the source. A four-hour episode and a ten-minute one cost the same for the same five clips.