Why do I zone out in meetings and miss the one decision that mattered?

Usually not because you were careless. Attention to a talk declines measurably with time on task, and meetings put their decisions in the second half. A decision spoken out loud and never written down leaves you holding the shape of it while the specifics go.

The cleanest demonstration is a lecture study, which is a meeting where nobody asks you anything. Participants watched a 40 minute recorded talk, rated their attentiveness every five minutes, then were tested on what they kept. Both attention and retention fell as time on task went up (Farley, Risko and Kingstone 2013). Forty minutes is a short meeting; a 60 minute call gives the decline twenty more minutes.

Two processes eat that second half: vigilance decrement, the drop in your ability to keep monitoring something dull, and mind wandering, attention simply leaving. A 2023 sustained-attention experiment found the two dissociate: raising task demands changed the vigilance decline, while brain stimulation changed the mind-wandering rate (Martínez-Pérez and colleagues 2023).

Why can I be present for the whole meeting and still not encode the decision?

Because presence is not encoding. Sitting in a room does not write anything to memory. Encoding happens when you do something with the words while they are still available, which is why the person summarizing, arguing or writing the notes is the one who keeps the decision.

Mind wandering is the base state, not the exception. The best known estimate, from Killingsworth and Gilbert in 2010, put it at 46.9 percent of waking hours, and the wider literature lands between 30 and 50 percent (PubMed record). An engineering forum thread on forgetting what was said in meetings is mostly people discovering memory was never the plan.

Relevance gating does the rest. Meetings are fragmented by whose area is on the table, so attention switches off during other people's segments. In a career forum thread about two-hour meetings, one person put it exactly: people discuss things unrelated to my field, I switch off, and days later a manager asks what I think of a decision made in that stretch.

Why do I remember the gist and get the numbers wrong?

Because memory keeps the summary and drops the exact parts. What survives is the direction of the decision. What goes first is the number, the date, the owner and the wording. Two people can both remember that something was agreed and still be describing different things.

Psychologists call the meaning of what happened gist memory, and the exact words and figures verbatim memory. Gist is what holds up over time, and it holds by letting the rest go. Test it: you can name the architectural call your team made two years ago, and not the deadline attached to it. That asymmetry is why "I thought we decided X" is usually a memory artifact rather than a lie.

Why does the decision always get made at the worst possible moment?

Because the meeting rarely ends on its agenda. Side threads open, the item people actually came to settle gets pushed, and the close gets compressed into the final minutes. That is the least attentive point of the hour, and nobody is writing anything down.

The failure mode is public. In a managers' forum thread, one poster described a four hour meeting that decided none of the items it existed for, and the deal was gone six weeks later. Then the wording problem: close-out statements are often empty. "We'll move forward with the direction" gives you nothing concrete to encode, so even a listener who was fully present has no object to remember. Nothing here says endings are specially forgettable. The end just gets the least attention and the vaguest language.

Why did the cost only land a week later?

Because in the room you had the flow of the conversation and a feeling of knowing. The bill arrives when someone states their version of the decision and you have nothing to check it against. Then "I thought we decided X" becomes an argument, because you have a gist and they have a gist.

The social cost lands there too. In an actuary thread, one person described being asked days later to weigh in on a stretch they had switched off from, with no honest way to say "I spaced out". In an ADHD thread, another described feeling stupid asking a coworker for a reminder, because the request makes it look like they were not listening even when they were.

What should I write before I leave the room?

Three lines: what was decided, who owns it, and by when. Write them while you are still in the room, in the same minutes the decision is being made. The capture act is what gives future you something to hook onto.

The best line from a managers' thread was simple: if you can't write down what was decided, the meeting didn't end, it just paused. Whoever writes the record decides the version that survives, and a decision you never encoded cannot be recovered later by trying harder.

The AI notetaker changes the shape of the problem without solving it. The transcript now exists, and you still hold only the gist. Keep the assistant for the record and still write your own three lines.

The skill underneath is being asked what a conversation meant and finding out whether you had it right, which is what Absorb scores across reading, listening and analysis. If you want that check, see Absorb on the App Store.