You circulate a 12-page design pre-read 24 hours before a 50-minute decision meeting with nine attendees. Roughly how much of it will have been read when the meeting starts, and what does that number change about the way you run the first ten minutes?
Show the full answer Hide the answer
The assumptions, stated
A page of dense technical prose with a diagram on it carries roughly 450 to 550 words. Easy prose is read at about 250 words a minute; a design document with numbers, a schema and a cost table is read at 150 to 200 words a minute, because the reader stops to check things. Nine attendees each have a working day between receiving it and the meeting.
The arithmetic
12 pages × 500 words = about 6,000 words. At 175 words a minute that is roughly 34 minutes of attention, and it is uninterrupted attention, which nobody has between two meetings.
The realistic split for a group of nine with a day's notice: one or two people read it fully (usually the author's closest collaborator and the person who already disagrees), three or four read the first two pages and the headings, and the rest open it when the meeting starts. So the room arrives holding somewhere between 15% and 30% of the document, and almost none of the part you actually need a decision on, because the ask is conventionally at the end.
Which assumption dominates the error
Not the reading speed. It is the notice period. Twenty-four hours is one working day, and a working day with four meetings in it contains perhaps two 20-minute gaps. Moving the circulation to three working days changes the full-read fraction far more than cutting two pages does. The second-largest term is where the ask sits: a document whose first page states the decision being requested gets its decision read even by the people who read nothing else.
What the number rules in
A pre-read must fit the reading time you can actually command, and the only reading time you can command is the time inside the meeting. Two workable shapes:
- Shrink to fit a 15-minute in-room silent read: 15 minutes × 175 words a minute is about 2,600 words, so five or six pages. You spend 15 of your 50 minutes and get a room that has all read the same thing. The remaining 35 minutes is more decision time than a 50-minute meeting where the first 20 are spent reconstructing the problem aloud.
- Keep 12 pages but make pages 7 to 12 an appendix that the decision does not depend on. The budget then applies to the first six.
The failure the budget prevents is specific: when nobody has read it, the first substantive objection surfaces after the halfway mark, and an objection raised with 20 minutes left produces "let us take this offline" rather than a decision.
When this is the wrong answer
When the audience is two people who own the decision and work on the system daily, the pre-read is overhead; send a three-paragraph message with the ask in the first line. When the document exists for the record rather than for the meeting - a compliance submission, an architecture decision record - length is not the constraint and the in-room read is wasted time. And when the decision is genuinely reversible within a day, skip the meeting: circulate the choice with a deadline and let silence be assent.
Common weak answers
- "Ask everyone to read it beforehand." That is the plan that already failed. The instruction does not create the 34 minutes.
- "Summarise it at the start." A verbal summary of a document people have not read puts them in the worst position: they cannot check your summary, and they will not argue with a claim they cannot see.