How many words per minute? Timing a script before you record it

Count roughly 140 to 160 words per minute for an English script read at a natural pace. That is the most useful working range in a booth, and it should be taken for what it is: a range that moves, never a standard. Depending on the intention you ask for and the channel the piece plays on, the same page can land at a minute forty or run past two and a half minutes.

Which end of that range you land on is where the whole job sits.

Why a reading time calculator gives you the wrong answer

Most people arrive at script timing through a reading time calculator, and most of those are built on silent reading. A meta-analysis of 190 studies covering 18,573 participants puts the average silent reading rate for adults in English at 238 words per minute for non-fiction and 260 for fiction (Brysbaert, 2019). The same review puts oral reading, in test conditions, at 183 words per minute.

Neither is the average words per minute speaking rate for a directed voice over, which sits lower again because it has to land first time, through a broadcast channel, for a listener doing something else. Run a 300-word script through a generic words per minute reading tool and it will promise about seventy-five seconds. In the booth it runs closer to two minutes. Over-length scripts usually start in exactly that gap.

Benchmarks by format, as a starting point

The question “how many words per minute” has one answer per format, because a radio tag and a training module are not doing the same thing to a listener.

Three bands cover most professional reads. Call them measured, at roughly 110 to 130 words per minute; natural, at 140 to 160; brisk, at 170 to 190.

Format and duration	Typical pace	Rough text length
Radio spot, 15 seconds	brisk	35 to 45 words
Commercial, 30 seconds	brisk to natural	70 to 85 words
Corporate video, 2 minutes	natural	270 to 300 words
Training module, 5 minutes	measured	550 to 650 words
Documentary narration, 10 minutes	measured to slow	1,000 to 1,200 words
Audiobook, 1 hour	natural and even	8,000 to 9,000 words

These are booth benchmarks, refined across fifteen years of voice over sessions, and they sit deliberately below the raw arithmetic. The gap is where the breaths go, along with the brand name and any legal tag. Use them to frame a first draft and to spot an overrun early. They do not validate a delivery: in practice, only a stopwatch on the take does that.

What actually moves the pace

Intention comes before the stopwatch

A message that has to reassure is not delivered at the speed of a message that has to provoke. An executive address, a safety reminder: it settles, it breathes, it drops under 130 words per minute without effort. A limited offer, the call to action at the end of a spot: it climbs, and it passes 180 quite happily. Pace is a consequence of meaning, not a technical setting.

That is also why a constant rate from first word to last sounds wrong. Where a machine holds a steady speed, an actor slows on what matters and moves through what merely connects. The final average looks like your starting benchmark, and comes out of a very uneven journey.

The channel sets the clock

Radio and television buy calibrated time. A thirty is a thirty, and an overrun is not something the edit will quietly absorb for you: it is the hardest constraint on any advertising script. Online looks more forgiving and is much less so than people assume: a pre-roll, a vertical cut and a story each have their own window of attention. A corporate video shown in a conference room can breathe. Nobody there is going to skip.

Density weighs more than length

Two 300-word scripts do not last the same time. Simple ideas, in short sentences, with everyday vocabulary, go quickly and cleanly. A script stacking technical data, unfamiliar concepts and figures forces the pace down, or loses the listener on the way. In e-learning the relationship is close to mechanical: the newer the notion is to the learner, the further the delivery drops.

Proper nouns and figures cost far more than expected

This is the line item scripts underestimate most. A phone number, a dictated web address, a full company name, an acronym: each is said more slowly than the copy around it, because it has to land first time and cannot be recovered from context. A foreign brand name not prepared with the actor can cost a second on its own. On a fifteen, a second is a whole line.

Silences are part of the duration

A well-placed silence carries meaning. It lands a payoff, and it stops two ideas bleeding into each other. Shaving the breaths buys a few seconds and costs comprehension, the one thing you cannot buy back at the mix.

A detail in the research makes the point neatly: the cross-language study cited below measured its durations after discarding pauses longer than 150 ms. Laboratory speech rates exclude the silences. Broadcast timing pays for them, so they belong in the first estimate rather than the last-minute adjustment.

The same script does not last the same time in English, French or Spanish

This is the trap that breaks campaigns adapted across markets. A spot approved at thirty seconds in English does not mechanically fit thirty seconds in French or Spanish, and the reason is measurable.

Researchers recorded twenty texts translated into seven languages, read by fifty-nine native speakers, and found syllabic rates ranging from about 5.2 syllables per second in Mandarin to 7.8 in Spanish and Japanese, with English around 6.2 and French around 7.2. The faster languages also carry less information per syllable: taking Vietnamese as the external reference, information density per syllable came out at 0.91 for English against 0.74 for French and 0.63 for Spanish (Pellegrino, Coupé and Marsico, 2011).

Read from a producer’s chair, that says something specific. English packs more meaning into each syllable than French or Spanish do, so an English master adapted into either language usually needs more syllables to carry the same content. The adapted versions overrun, and they overrun by different amounts. Which is why a multilingual adaptation gets rewritten to the target duration: word-for-word translation is what produces the overrun in the first place. Across our six native languages, timing is set at the brief, language by language, and never discovered at the mix.

Your script runs long: what to cut, and in what order

“We will fix it in the edit, we will just push the pace a little.” That works once, on a second or two. Past that the compression is audible: the voice tightens, sentence endings get swallowed, and the message loses the quality you came to a human voice for. Cut on the page instead.

Order of cuts, from the easiest to the most painful:

1. Restated ideas, where the same point is made twice in different words. 2. Intensifiers that carry no information (“really”, “particularly”, “truly unique”). 3. Explicit transitions, which an inflection carries better than a phrase. 4. Subordinate clauses. One long sentence split in two is often faster to say and easier to follow. 5. As a last resort only, a secondary product benefit.

What never gets cut:

• The silence before the payoff. It is what makes the payoff audible. Removing it saves half a second and cancels the effect.
• The brand name, and the time to place it. A name rushed at the end throws away the twenty-eight seconds that paid for it.
• The mandatory legal mention. It cannot be negotiated, so it gets budgeted in seconds from the first draft.

The final duration decides the format, and the format decides the rights

You think you are counting words. You are calibrating a media format, and behind that format, a rights period.

A spot that overruns is no longer the slot the media plan reserved. Either the script goes back for another pass or the buy moves, and either way the chain shifts behind it. The link runs deeper than scheduling. A thirty on national television, a six-month online pre-roll and a training module running internally for three years do not represent the same use, so they do not fall under the same license. The duration of a read script sets the first parameter of the broadcast agreement you will sign.

Which is why script timing belongs in the brief, with the studio, and never at delivery.

The working habit fits in a line: start from the target duration and write backwards. Read the script out loud, stopwatch in hand, before anyone approves it. And when two versions are close, keep the one that leaves room for the silences.

Have you got a script and an imposed duration that do not match? Tell us the format, the channel and the intention you are after, and we will tell you what holds and what needs rewriting before the booth. Let’s talk about your project.