01
Is there a shortcut?
Two that help. Look for places where a question changes the answer, since those are where something actually happens. And ignore the opening and closing minutes, which are almost always throat-clearing on both ends. Beyond that, the search itself is the expensive part, which is why handing it over changes how long the job takes rather than how good the result is.
02
Questions to ask of each candidate
Would a stranger understand this without anything that came before? Does something change between the first sentence and the last, a surprise, a reversal, a new piece of information? Would someone want to send it to a friend? A moment that passes all three is worth cutting. A moment that passes only the first is usually a nice passage rather than a clip.
Be honest about the second question especially. Many passages are interesting but static: the speaker says one thing, elaborates on it, and ends where they started. Clips that travel almost always move somewhere.
Speed matters too: decide on each candidate in one viewing where you can.
03
Using the audience you already have
If the long version is public, its comments and any timestamps viewers shared are a free guide to the moments people cared about. A timestamp several people mention independently is a strong candidate before you have watched anything yourself.
Co-hosts and guests are another source. Ask each of them which moment they remember most from the recording. Their answers rarely match exactly, and the overlap usually points straight at the best clip.
Your own reactions count too.
04
Setting a limit before you start
Decide how many clips you want before you begin looking. Without a target, it is easy to either stop after two or keep going until every passage looks like a candidate. A fixed number forces you to compare moments against each other and keep only the strongest.
If you finish with more good candidates than your target, keep the extras in reserve rather than lowering your standard for the week.
Stick to it.
05
Keeping notes for next time
After the clips are published, note which ones performed and what they had in common. Over a few months those notes describe what your audience responds to more precisely than any general advice.
Share that description with anyone who helps choose moments, so the whole team looks for the same qualities.
06
Hand over the searching
- Skip the first and last few minutes. Openings and wrap-ups rarely survive being lifted out.
- Look for turns, not highlights. Where the conversation changes direction is where the clips are.
- Keep, adjust, or override. On paid plans you can supply your own timestamps instead.
Try it on one long recording and see what it puts forward.
Questions people ask
- Can I overrule the choices?
- On any paid plan, yes, by entering timestamps yourself.
- Does it cut at fixed intervals?
- No. Equal slices are what produce clips that start mid-sentence.
- What if I disagree with all of them?
- Set your own timestamps and the rest of the process still applies.
- Is there a limit on the recording length?
- Ten hours, on every plan. Thirty minutes is what the single free credit covers.
- Is a two-hour recording more expensive to run?
- On Lite 30 it is one credit, the same as a short one, because a credit covers two hours there. On Lite 10 it is two credits - and those two bring twenty included clips instead of ten.
- How many usable moments does an hour usually hold?
- Fewer than people expect. An hour of talking rarely carries more than a handful of passages that stand on their own, and stretching the count is what makes a batch feel weak.