Guide
Cut where the sentence ends. Pace is a consequence, not a setting
There is no correct number of cuts, because a cut is not a unit of time, it is a punctuation mark. Cut on a full stop and nobody notices, cut in the middle of a clause and it reads as a mistake even to someone who could not explain why. What controls how fast a short video feels is three separate rates, and the cuts are only one of them.
- Reference spacing for a non-cut transition
- 10 s
- Reference spacing for a punch-in
- 6 s
- How long fast cutting hides a weak script
- 4 s
The sentence tells you where the cut goes, and it is right more often than the clock
Speech already has structure. There is a boundary at the end of every clause and a bigger one at the end of every sentence, and a viewer feels both whether or not they are paying attention.
Cut on one of those and the edit disappears. Cut across one and the viewer registers an interruption, which costs a fraction of a second of attention on a video that only has a few to spend.
This is why removing pauses and false starts is invisible while a metronome edit is not. One follows the structure of the speech and the other imposes a rhythm on top of it.
It also means your pace is mostly decided before you edit anything. Somebody who speaks in short sentences produces a fast cut and somebody who rambles produces a slow one, and no amount of editing changes that without changing the read.
Three rates decide how fast a video feels, and only one is the cutting
The first is how often the picture changes: cuts and cutaways together. This is what people mean by pace and it is the one everybody adjusts.
The second is how often something moves within a shot, mostly punch-ins. We space these about six seconds apart at the default, because a scale change every couple of seconds stops reading as emphasis and starts reading as a nervous camera operator.
The third is how often you use an effect rather than a plain cut. Whips, punches, glitches and sweeps have a budget, and the reference spacing is about one every ten seconds. Our packs actually run at twice their declared spacing, because in testing even the reference cadence felt busier than anyone wanted.
Those three add up. A video with frequent cuts, frequent punch-ins and a transition every few seconds is not fast, it is exhausting, and viewers leave exhausting videos faster than they leave slow ones.
Reference spacing for the things that are not cuts
Both are floors on the interval, not targets. Going under them is what makes an edit feel frantic rather than energetic.
- Punch-inone per 6 s
- Non-cut transitionone per 10 s
- What our packs actually runone per 20 s
This is the whole editor
Highlight a phrase and a clip lands on those exact words. No timeline, no keyframes, no layers.
Fast cutting buys you about four seconds, then it stops working
Rapid cutting genuinely holds attention at the start of a video. It signals that something is happening and it buys a viewer's patience for a few seconds.
After that the effect inverts. If the cuts are fast and nothing is being said, the viewer works out that the movement is a substitute for content, and they leave faster than they would have from a slow video with a good first line.
This is the diagnostic worth remembering: if you are reaching for a faster edit, check whether the problem is actually the script. A cut cannot make a sentence more interesting.
The reverse case is real too. A slow edit on a compelling read is not a problem, and plenty of the best performing direct response video is one person talking with barely any cutting at all.
What a faster edit does and does not fix
Two of these are worth cutting for. The other two get worse when you speed the edit up.
| The actual problem | Does faster cutting help? | |
|---|---|---|
| A visually flat stretch mid-read | Nothing to look at | Yes, that is what it is for |
| Pauses and false starts | Dead air | Yes, remove them entirely |
| A weak opening line | Nobody is listening | For four seconds, then no |
| No clear argument | Script problem | No, it makes it obvious |
The practical version, in the order to do it
Remove everything that is not speech first. Pauses, filler, the two false starts before the sentence landed. That alone usually takes a third off the runtime and raises the pace without a single stylistic decision.
Then place cutaways only where you named something worth seeing or where the read goes flat. Keep the total under about half the runtime so the speaker stays the ad.
Then add effects last, sparingly, and count them. If you have more than two or three non-cut transitions in a thirty second ad, take some out.
Cutroom does the first step automatically and applies the spacing rules to the rest, and you can change the pace and have the whole thing recut. What it will not do is invent a rhythm for music-led video, where the edit follows the track rather than the read, and a timeline editor is the right tool for that.
Questions people ask
- How many cuts should a 30 second ad have?
- However many sentence boundaries it contains. That is usually somewhere between eight and twenty, and the number is a result of how the person speaks rather than a target to hit.
- Do fast cuts improve retention?
- At the start, for a few seconds. Sustained across a whole video with nothing being said, they reduce it, because the viewer reads the movement as compensation.
- Should every cut have a transition effect?
- No. A plain cut is invisible and that is its virtue. Effects are punctuation and they should be roughly as frequent as an exclamation mark in decent writing.
- Can I control the pace in Cutroom?
- Yes. Pace is a setting and changing it recuts the whole thing, including how much cutaway coverage is allowed. You can also trim, delete a line and change how each clip enters.
Cut on the sentence, budget your effects, and check the script before you reach for a faster edit. Changing the pace here recuts everything at once, which makes trying the slow version cost about a minute.