Script Writing for Video That Respects Runtime

How many words actually fit in a given runtime, how to cut a script without losing the argument, and why writing to length is a structural discipline.

Almost every commercial video script arrives too long, and the gap is not marginal. A client who has been asked for a sixty second film routinely supplies four hundred words, which is close to three minutes of narration. Resolving that gap early is a conversation. Resolving it late produces either a rushed read that sounds anxious or a film that overruns and gets rejected by the platform it was made for.

The arithmetic is simple and worth internalising. A comfortable commercial voiceover runs at roughly one hundred and forty to one hundred and sixty words per minute. That gives about one hundred and fifty words for sixty seconds, seventy five for thirty, and around forty for a fifteen second social cut. Those figures assume a normal pace with room to breathe. Pushing to two hundred words in a minute is possible and it sounds hurried, which undermines whatever authority the film was meant to convey.

The figures also assume narration is the only thing happening, which it rarely is. A film needs silence for a product reveal to land, for a title to be read, for a demonstration to be understood and for the music to resolve. Those pauses are not wasted seconds, they are where the film does its emotional work, and they have to be budgeted. A realistic working figure is around one hundred and twenty words of narration in a sixty second film that also has to show something.

Cutting is easier when it follows a principle rather than proceeding line by line. The most reliable order is to remove the introduction first, since most scripts spend their opening explaining what they are about to explain. Then remove qualifications and hedging, which add length and reduce conviction simultaneously. Then remove the second example wherever two are given. Then compress descriptions of things the film is already showing, which is the most common form of redundancy in video writing.

That last point deserves emphasis because it is specific to the medium. Narration that describes what is on screen is wasted, since the audience already has the information visually. The narration's job is to supply what the picture cannot: the reason, the implication, the number, the consequence. A script written without reference to the storyboard almost always duplicates the visuals, and reading the two together is the fastest way to find twenty percent of the length.

Writing to length from the beginning produces better results than writing freely and cutting, because the constraint forces structural decisions early. That means deciding the single most important idea before writing, allocating a rough word budget to each section, and accepting that a sixty second film contains one argument with two supporting points rather than five ideas compressed. Scripts that try to include everything communicate proportionally less of each thing.

Attention research supports front loading rather than building. Yu et al. (2025) found that visual attention to advertising messages within video stories is distributed unevenly rather than remaining constant, which means material placed late is seen by fewer people and with less attention. The practical consequence is that the conventional structure of context, then argument, then conclusion is exactly wrong for short commercial video. The conclusion should arrive early and the support should follow.

Sentence construction matters more in spoken script than in written copy because the listener cannot re-read. Short sentences, one idea each, active voice, and concrete nouns are not stylistic preferences here, they are comprehension requirements. Subordinate clauses that work on a page cause a listener to lose the thread, and a sentence a performer cannot say in one breath is a sentence that will be re-recorded.

Reading aloud is the only reliable test and it is skipped surprisingly often. A script that reads well silently frequently contains phrases that are awkward to say, alliteration that becomes noticeable, and sentences that run out of air. Timing the read aloud, with a stopwatch, against the intended runtime, gives a real number rather than an estimate, and doing this before the client approves the script prevents approving a runtime nobody has verified.

For multilingual delivery the length problem multiplies, because the same content occupies different durations in different languages. A script cut precisely to sixty seconds in English will overrun in Bahasa Malaysia. Writing the primary language slightly short, and building the edit with small timing tolerances at scene boundaries, is what allows the other versions to fit without re-editing the picture. This costs nothing at script stage and saves an entire round of work later.

References

Yu, W.-Y., Wang, Z. J., & Tao, C.-C. (2025). The dynamics of visual attention to advertising messages in video stories. Journal of Advertising, 54(5), 713–731. https://doi.org/10.1080/00913367.2025.2524837