Why Meetings Go Wrong, Quietly

The bad meetings I remember weren't the ones with shouting. The ones that did real damage were polite. Someone said a number that wasn't quite right. Someone changed the subject when a question got close to a soft spot. Everyone nodded. The meeting ended on time. A decision got made on the back of a thing that wasn't true.

This is the thing about why meetings go wrong. They almost never go wrong loudly. They go wrong in four small places, and the social cost of pointing at any of them in the moment is higher than most people are willing to pay.

Unverified math

Numbers get spoken aloud and become facts. A founder says, "We're growing 30% on a $4M base, so we'll be at $6M by year-end." The room moves on. Nobody pulls out a calculator, because pulling out a calculator in a pitch is rude.

This kind of thing is rampant. Round-number inflation, mismatched timeframes, ARR confused with bookings, gross confused with net. Once a number has been stated confidently and not challenged, it becomes the floor for the rest of the conversation.

Contradictions

The CFO said, in March, that the gross margin was 62%. In June, in a different meeting, with mostly different people, she said it was 71%. Both numbers are now sitting in different decks. Nobody in either room remembered the other meeting well enough to ask.

Contradictions get past us because human memory for spoken claims is, frankly, terrible. We remember vibes. We remember the conclusion. We don't remember the supporting figure two months later, and so the same question gets answered three different ways across a quarter, and nobody notices.

Dodged questions

A board member asks about the timeline for the migration. The CTO answers, fluently, about the architecture. Forty seconds of impressive-sounding detail. The board member, satisfied that something has been said, moves on.

The timeline never got answered.

This is the most common failure mode in any meeting with status differential. The senior person asks a question. The junior person answers a related question. The senior person, not wanting to seem to have missed the response, doesn't ask again. The original question, which was the actual question, dies.

Unsupported claims

"Our retention is best-in-class." "We're seeing strong pull from enterprise." "The new hire is working out great."

None of these are statements. They're vibes wearing the clothes of statements. In a healthy room someone would ask, "Best in class measured how, against whom?" In most rooms, nobody asks, because asking signals distrust, and distrust is socially expensive.

The reason all four of these get past everyone, every time, is the same reason. Calling out a math error in real time makes you look like a pedant. Pointing at a contradiction makes you look like you're keeping score. Re-asking a dodged question makes you look like you weren't paying attention the first time. Asking for evidence behind a casual claim makes you look adversarial.

So nobody does it. The meeting ends. The decision gets made. Six months later, when the numbers come in differently than the room expected, no one can quite reconstruct where the gap opened up.

It opened up in the meeting. Quietly.