The meeting that was supposed to take thirty minutes
Someone blocked thirty minutes. It was on the calendar in clean digital ink: 2:00 to 2:30. At 2:31 the host said, "Let's keep going, this is important." At 2:48 three people had silently opened laptops to triage the meetings they were now late for. At 2:55 the actual decision—the one the meeting existed to make—got crammed into the last four minutes and deferred to a follow-up nobody wanted.
If this feels familiar, you are not bad at scheduling. You are running into one of the most reliably documented biases in all of behavioral science. Understanding it won't just make you feel better. It will change how you put time on a calendar.
The planning fallacy, named
In 1979, Daniel Kahneman and Amos Tversky described a pattern they called the planning fallacy: the tendency to underestimate how long a task will take, even when we have direct experience of similar tasks taking longer. The strange part isn't that we're wrong. It's that we're wrong in a consistent direction. We almost never overestimate. We almost always run over.
Kahneman later illustrated it with a story about his own team writing a textbook. Asked how long it would take, they guessed around two years. He then asked a colleague how long comparable projects had taken others. The answer: seven to ten years, and many were never finished. The team had all that information and still produced an optimistic guess, because they built the estimate from the inside—imagining the smooth path through their specific plan—rather than from the outside, looking at how similar efforts actually unfolded.
Meetings are the planning fallacy in miniature, repeated dozens of times a week. When you block thirty minutes, you are picturing the ideal version: everyone shows up prepared, the conversation goes straight to the point, the decision lands cleanly. You are not picturing the five minutes of "can everyone see my screen," the tangent about the unrelated launch, or the colleague who needs the entire context re-explained because they skimmed the doc.
Why the inside view feels so convincing
The inside view is seductive because it's built from real, specific detail. You can genuinely see the agenda in your head. Each item looks small. The problem is that visualizing a plan activates the parts of it you can articulate, and the things that blow up timelines are precisely the things you can't picture in advance—the interruptions, the misunderstandings, the question that reveals two people thought they'd agreed on different things.
There's a related mechanism at work, too. Researchers distinguish the planning fallacy from simple optimism by noting it's anchored to a specific plan. The more concrete your mental script for the meeting, the more confident—and the more wrong—your estimate becomes. Detail breeds certainty, not accuracy.
Add to this a social pressure unique to meetings. Booking sixty minutes feels greedy; booking thirty feels respectful and efficient. So we systematically under-book to signal consideration, then systematically run over in practice, which is the least considerate outcome of all because it steals time people didn't agree to give.
The outside view: borrow from your own past
The fix Kahneman and Tversky pointed to is deceptively plain: take the outside view. Instead of estimating from the plan, estimate from the reference class—the actual track record of similar meetings.
This is sometimes called reference class forecasting, and for meetings it's almost embarrassingly available, because you have a near-perfect dataset: your own calendar history and your own memory. Before you book the next recurring sync, ask the outside-view question. Not "how long should this take if it goes well?" but "how long has this kind of conversation actually taken the last several times?"
If your weekly status meeting has run to fifty minutes for a month, the honest estimate is fifty minutes, not the thirty you keep optimistically rebooking. The recurring overrun is not a series of unlucky exceptions. It is the reference class. The thirty-minute slot is the fiction.
Make the estimate harder to fool
A few concrete practices turn the outside view into a habit.
Estimate from history, then add a margin. Look at what the last three comparable meetings actually consumed, including the ragged ends. Use that number. The discomfort you feel booking the larger block is the planning fallacy resisting correction—that discomfort is the signal you're finally estimating honestly.
Name the decision the meeting must produce. A meeting with one explicit decision has a finish line. "Discuss the roadmap" can expand forever because nothing tells it to stop. "Decide whether to cut feature X from the release" ends when the decision is made. Open-ended purpose is a primary fuel for overrun.
Do a premortem on time. Before the meeting, spend thirty seconds imagining it has already run fifteen minutes over, and ask why. The answers come quickly: someone wasn't briefed, two items secretly needed their own meeting, the demo broke. A premortem deliberately shifts you from the optimistic inside view to a prospective outside view, and it surfaces the time sinks while you can still cut them.
Review the gap afterward. The only way the outside view improves is if you compare what you booked against what it took. A one-line note—"booked 30, ran 52, lost ten minutes to setup"—is a data point. Collect a few and your estimates stop being wishes and start being forecasts.
What honesty about time actually buys you
Notice what all of this protects. It isn't just the meeting in question. Every meeting that runs over is a small theft from whatever came next, and the cost compounds across a team's whole day. The person who under-books to seem efficient creates the cascade of lateness that makes everyone feel perpetually behind. Accurate time estimates aren't a productivity flourish; they're a form of respect, paid in advance, in the only currency a calendar understands.
The deeper shift is from optimism to evidence. You stop asking your imagination how long a meeting will take and start asking your history. Your imagination is a flatterer. Your history is a witness.
Bringing the witness to the table
This is the gap MeetingMortem was built to close. After a meeting, it helps you run a quick honest review—what you planned for, what it actually took, where the time went—so the next estimate is grounded in your real reference class instead of your best-case guess. Over a few weeks, that record becomes the outside view the planning fallacy keeps trying to hide from you, and your thirty-minute blocks either become honest or become something else.
If your meetings keep quietly stealing the rest of your day, you can start building that witness here: https://meetingmortem.lumenlabs.works. The science is free and yours to use whether or not you ever sign up—but the calendar believes evidence, not intentions.