If your clipper needs timestamps, you hired an editor

The tell is on the intake form. Somewhere between the episode link and the caption colour there is a field for timestamps, and if it is required, the vendor is cutting on your marks rather than choosing them. That is editing. Clipping is the choosing, and a vendor who needs you to do it has handed the hard half of the job back to you.

That one field tells you more about what you are buying than the pricing page does.

Why a clipper who needs timestamps is an editor

There are two jobs hiding under the word clipping, and vendors are not always careful to say which one they sell.

The first job is execution. Take a start mark and an end mark, reframe the section to vertical, track the speaker, caption it, apply the style, export. This is editing. It is skilled work, it takes hours, and it is the part a machine is getting steadily better at.

The second job is selection. Watch the episode. Decide which moments deserve to stand alone, where each one starts, where it ends, which ones say something the audience has not heard from this person before, which ones would embarrass the guest, and how many the episode can honestly support. This is clipping. It is the reason you went looking for a vendor in the first place, because the person who cannot find the hours to rewatch their own episode is usually the person who hosts it.

If the vendor needs timestamps from you, they are doing the first job. You are doing the second. That can be a fine arrangement, and we come back to when it is. The problem is that most people who send timestamps thought they were paying for both.

What selection involves

Selection starts with the full episode, at normal speed, watched the way a viewer would watch it. A transcript skim does not count, and neither does a keyword search for the topic the host remembers going well.

The sentence that carries the point is rarely where the thought starts. The start is usually a beat earlier, in the setup. Cut on the sentence and the clip opens in the middle of something. The viewer does not work out what went wrong. They leave. How we choose clips from a podcast goes through the mechanics.

It means knowing the client. A founder who sells to enterprise buyers and a founder who sells a course have different moments worth cutting from the same conversation. The story that is funny at a dinner is not always the one that should sit under the company name for as long as the clip is up. Someone has to hold that judgement, and it has to be someone who knows what the client is for.

It means deciding the count from the footage. A long episode with a strong guest carries far more usable moments than the same length with a flat one. A vendor who has promised a number per episode has to fill it, and the filler is easy to spot once it is posted. How many clips an episode can support is its own question, but the honest answer always starts with watching.

And it means saying when the episode is weak. This is the part vendors least want to do, because it means a smaller delivery and a harder conversation. It is also the part that separates a clipper from a contractor.

Executing a brief versus watching the episode

A brief is a fine input. It is the client doing selection and passing the result along. Editors work from briefs all day and the output is good.

The cost is where it lands. Selection takes at least the running time of the episode, because there is no shortcut past watching it, and it has to come from someone who knows the show. If that person is the host, you have turned the vendor into a way of spending more of the host's time. If that person is a producer who already logs the show, the timestamps exist anyway and you should pay editor rates for the execution, which are lower.

There is a quality argument too, and it is the one people resist. Hosts are unreliable judges of their own episodes. They remember what felt good in the room, which is not the same as what reads cold to a stranger in the opening of a vertical clip. The moment the host is proudest of is often the one with the long run-up. The moment that works is often an aside they had forgotten. A second pair of eyes with no memory of the recording session is the point of the exercise.

What to ask a vendor

Across 17 competitor clipping-agency FAQ pages we read on 10 September 2026, revisions were answered by one, ownership of source files by three, and what the reporting looks like by two. The questions that decide whether you are buying selection are mostly not answered in public at all. So ask them directly, and listen for the shape of the answer.

The last row is one of the gaps from those FAQ pages. Ask about revisions and reporting as well, whether or not the vendor is doing selection, because silence on a public page usually means the answer is set per client, and you want yours set before you pay.

The delivery note question is the quickest test. A vendor who picked the moment can say why in a sentence. A vendor who cut on your marks cannot, because the reason was yours.

One more question is worth asking, because the answer is often a tool: what does the vendor use to find candidate moments? Transcription and first-pass candidate surfacing are reasonable things to automate. What you are checking is whether a person then watches, or whether the candidates go straight to the edit.

When timestamps are the right call

If you have a producer who already marks the show during recording, or a host who genuinely enjoys the rewatch and is good at it, sending timestamps is efficient. You have done the selection in-house and you should buy execution at execution rates. The in-house versus agency comparison covers what that costs in hours.

The mistake is paying agency rates, sending timestamps, and assuming the vendor is adding judgement on top. If they asked for the marks, they are cutting on the marks.

How we run it

We take the episode link. Someone watches the whole thing, picks the moments, and cuts them in the client's style rather than ours. Each clip comes back with a line on why it was chosen. When the episode is flat, we say so and deliver fewer, because a padded delivery costs the client more in the feed than a short one costs us on the invoice.

The next time a vendor sends an intake form, find the timestamp field before you find the price. If it is required, you now know which job you are buying, and you can pay for that job and not the other one.