Speaking time calculator

Find the rate that is actually yours

Every duration on this page comes out of one number: words divided by a speaking rate. Change the rate and everything downstream changes with it. A 1,300-word script is exactly ten minutes at 130 words per minute and a little over thirteen at 100, and no amount of careful drafting rescues a rate that was guessed.

You only have to establish it once. The measurement is more reliable if you use material like the material you will deliver, because technical vocabulary, proper nouns and figures all slow people down. A passage of light prose will flatter you.

• Pick a section from the middle of your own script, not the opening. Openings get rehearsed more than anything else and they run faster than the rest of the talk.
• Stand up. Sitting changes your breathing, and most people read a little quicker seated than they speak on their feet.
• Start a stopwatch, read aloud for three minutes at the pace you intend to use, and mark where you stopped.
• Count the words from the start to the mark, divide by three, and enter the result as your custom rate.
• If you can rehearse in the actual room, measure there. A large space and a microphone both tend to slow speakers down.

Why the rate moves between rehearsal and the room

A speaking rate is not a fixed property of a person. It is what one person did, once, under particular conditions — and several of those conditions change on the day.

Adrenaline is the largest single factor. Standing in front of an audience compresses your sense of time: pauses feel longer than they are, silence feels like a mistake in progress, and the instinctive correction is to speed up. Speakers who have never been timed on stage tend to be surprised by how much of the script they got through and how little of it landed.

Rehearsal pushes in the same direction. The tenth pass through an argument is quicker than the third, because you are no longer deciding what comes next. A talk that timed perfectly at home can finish two minutes early for that reason alone: the practised version was faster than the version you measured.

Two things push the other way. Unfamiliar material costs time, and so does a room that makes you work — projecting to the back of a hall, waiting for a laugh, or taking a question mid-sentence.

The practical response is to treat your rate as a band rather than a point. Check the duration at your measured rate, then check it again ten percent faster. If the script fits at both ends you are safe. If it only fits at one end, the script is doing the gambling on your behalf.

A talk is longer than its script

The estimate describes speech with nothing in between it. A presentation is speech with a great deal in between, and the difference lands at the end of your slot rather than being spread evenly across it.

Totalled honestly, that overhead is what turns a ten-minute slot into roughly nine minutes of actual speech. Setting the target duration to 85 or 90 percent of the time you were given is not caution. It is arithmetic.

Here is where the time tends to go in a short conference talk. The figures are illustrative, not measured, but they are the right order of magnitude — and building your own version of this list is worth more than the numbers themselves.

When the slot includes questions

"Twenty minutes" and "twenty minutes including Q&A" are different briefs, and the second is what most conference emails actually mean. Decide the split before you write a word. The split determines the script, and the script is what you will be trying to rescue if you get this wrong.

Reading a script and speaking from notes are different speeds

A full script read aloud runs faster than the same content delivered from bullet points, and the reason is not confidence. Reading removes the small decisions — which word next, how to phrase this — and those decisions are taken in the gaps between phrases. Remove them and the gaps close up.

Speaking from notes costs time a second way: you say things more than once. Unscripted delivery circles back, restates, and adds the clause you forgot the first time. None of that was ever in your word count. Ten minutes from notes covers noticeably less ground than ten minutes from a page.

That has a direct consequence for the estimate. If you intend to deliver from notes, your script length is not what you will say — it is a floor. Write the talk out to get a count, then time a rehearsal delivered the way you actually mean to deliver it, and trust that duration instead.

The hybrid is worth knowing about. Script the first ninety seconds and the last ninety seconds word for word and work from notes in between. The opening and the ending are where fumbling costs most, and they are short enough to time exactly.

Pacing, and why the ending is what gets destroyed

People do not run over evenly. They run over at the end, because the sense of time pressure arrives late and the reflex it triggers is acceleration.

The pattern repeats often enough to plan around. The opening goes out at rehearsal pace or slower. The middle swells: an example earns an extra sentence, a slide needs more explanation than it did at home, someone asks something. By the time the speaker notices the clock there are four minutes of material left and ninety seconds to put them in. The ending — the only part anyone will still be carrying an hour later — gets delivered at double speed or abandoned.

A word count cannot fix that by itself. Time-marks can.

• Split the script into three or four blocks and give each block a clock time. Paste one block on its own and the tool gives you its duration.
• Write the mark in the margin of your notes — "6:00 by here" — and look at it once, at the halfway point.
• Rehearse the final two minutes on their own, aloud, at least twice. It is the section that gets the least practice and carries the most weight.
• If you are behind at halfway, cut from the middle. Losing an example costs the audience one illustration. Losing the ending costs them the point.