How do I read to prepare for a meeting, a presentation, or a decision?

Read toward a declared use. Name what you have to do in the room, ask, defend or decide, then write the questions you expect and read only far enough to answer them. Finish with a two-minute recall of the three facts you would defend. Most prep time goes on re-reading context you already had.

The order matters more than the length. Name the use in one sentence, write the three questions you expect, read against those questions, and end by recalling the facts you would defend. That last move is the one almost nobody runs, and the only one that shows what is missing while the document is still open.

The timebox is the other half of it. One commenter in a Reddit thread on meeting prep described the smallest version that works: note who is in the meeting, spend fifteen minutes writing down your thinking on the topic. That is a bounded task with an artifact at the end, not reading until the call starts.

Why does most of my meeting prep turn into re-reading things I already saw?

Because re-reading feels like work and costs nothing to start, so it fills the timebox. You reopen last week's thread and the same three tabs, call that context recovery, and arrive with a head full of state and nothing retrievable. Encoding for a use is a different task, and it is the one that survives the meeting.

One poster in that thread calculated their meeting prep time and got twelve hours across two weeks. The top reply renamed the behaviour: "Most of what I called prep was actually me reconstructing context I should've already had." The poster agreed, and called roughly three of the twelve hours unavoidable.

That is one person's accounting, not a measurement of how prep is spent, but the shape is familiar: the same documents reopened and nothing you could say out loud without the page. The cost lands after the meeting, as a lost record of what was agreed.

Why do I miss things in a presentation when I haven't read the material first?

Because building the map and using it draw on the same limited capacity. If the material is new in the room, you spend the meeting assembling the basics, and little is left for the question you meant to ask. A prior pass is what buys you the room to think on your feet.

One experienced developer described the mechanics plainly: "If I'm hearing something fresh for the first time I'm pretty much going to miss things. This is because I'm taking notes on questions I have and thinking about what you said thus not fully paying attention." A base knowledge of the topic is what lets them work off the cuff at all.

Steven Pan and Shana Carpenter reviewed the prequestioning literature in 2023 and found that attempting questions before you have learned the material can often improve later learning, as long as the answers follow. Read the prequestioning review. The attempt appears to change how you then read, which is why it beats a passive second pass.

The earlier experiments were smaller and blunter. Richland, Kornell and Kao had readers attempt prequestions before reading a passage and found a benefit for later recall even when the attempts were wrong. Read the 2009 pretesting experiments. Both carry a caveat: the size of the benefit depends on how the questions are written and how learning is checked. The direction is reliable, the magnitude is not.

How do I read a report so I can defend its numbers?

Work backwards from the number you would be asked to defend. Find it in the document, trace where it comes from, and check the one assumption that would change it. Say your answer out loud once. If you cannot state it without the page open, you have read it and not yet learned it.

For a decision, the target is narrower than the report: you are reading it to answer the question you would get if the decision goes wrong, which means knowing which facts are load-bearing.

The organisational version is a pre-read nobody opens. One project manager sent a document to 60-plus participants with a read receipt attached and reported that roughly 10% had opened it. That is one person's number, not a statistic. What follows is a meeting treated as the source of record.

This is also where preparation turns into deciding on information you only half understood.

How much prep is enough before a meeting?

Enough to answer the questions you expect without opening the document. Fifteen focused minutes that produce a written artifact beat two hours of re-reading, because the artifact is the output. Test readiness with one sentence: what am I being asked to do in this room, and what are the three facts I would defend?

Two details make it work. If the agenda lists questions, read for those specifically. And if you freeze when asked to speak, write your points the night before: retrieval in advance works for the same reason retrieval practice does.

If you cannot tell whether the weak link is your reading, your listening or your interpreting, stop guessing and measure it. See Absorb on the App Store runs three short rounds, one each for reading, listening and analyzing, inside a 10 to 12 minute session, with the scores kept on your device.

You do not need to remember the whole document. You need the questions you can answer and the difference between preparing for a room and re-reading for one.