The meeting that felt too easy

There is a particular kind of meeting that feels wonderful while it is happening and turns out to have been a disaster. Everyone nods. The energy is warm. Someone floats a plan, a few people build on it, nobody frowns, and within twenty minutes you have a decision and a shared sense that you are a sharp, aligned team. You leave a little proud. Weeks later, when the plan quietly falls apart, the strangest part is that no one can remember anyone objecting. Because no one did.

That smoothness is worth being suspicious of. The psychologist Irving Janis spent years studying how intelligent, experienced groups talked themselves into decisions that any one of them, alone, would have questioned. He coined a name for the pattern in 1972: groupthink. His central claim is counterintuitive and worth sitting with. The problem is not that the people are foolish or the information is bad. The problem is that the group is getting along too well, and the desire to keep it that way quietly overrides the desire to be right.

What groupthink actually is

Janis studied famous fiascos—the Bay of Pigs invasion, the escalation in Vietnam—and found the same texture in the rooms where they were decided. Cohesive teams, under pressure, drifting toward consensus not because they had examined the alternatives but because examining them felt like friction nobody wanted to introduce.

He described a cluster of symptoms. An illusion of unanimity: silence gets read as agreement, so everyone assumes everyone else is on board. Self-censorship: individuals privately doubt the plan but keep it to themselves, reasoning that if they were right, surely someone else would have spoken. Direct pressure on the occasional dissenter, usually gentle—a joke, a slight cooling, a "let's not rabbit-hole." And what Janis memorably called mindguards: members who take it upon themselves to shield the group from inconvenient information, the way a bodyguard shields a body.

Underneath all of it runs a simple social calculation most of us make without noticing. Voicing a doubt is costly. It slows things down, it risks looking negative, and it sets you against the visible momentum of the room. Staying quiet is free. When enough people run that calculation at once, the group manufactures an agreement that none of its members fully hold.

The illusion of agreement

The engine here is a second phenomenon researchers call pluralistic ignorance—a situation where most people in a group privately reject an idea but wrongly assume the majority accepts it, so everyone publicly goes along. Each person is looking around the table for permission to object, and finding only other people looking around the table. The result is a room full of private reservations and public consent.

This is why groupthink is so hard to feel from the inside. Your own hesitation seems like a personal quirk rather than a signal. You assume the others know something you don't, that their confidence is evidence. But their confidence may be nothing more than a performance of the same assumption you are making. Confidence, in a group, is contagious in a way that doubt is not—doubt has to be spoken to spread, and speaking it is exactly what the dynamic suppresses.

Cohesion makes it worse, which is the part managers least want to hear. We are told to build tight, trusting teams, and tight, trusting teams are wonderful for morale and execution. But Janis found that the same cohesion that makes a group pleasant to belong to also raises the price of dissent. The more you like these people and want their regard, the more a contrary opinion feels like a small betrayal. High-performing, friendly teams are not immune to groupthink. They are its natural habitat.

Why smart people don't fix it by being smart

The tempting response is to assume it won't happen to your team because your people are rigorous and independent-minded. Intelligence is not the safeguard it feels like. Groupthink operates on the social layer, not the analytical one. A room full of brilliant individuals is still a room, subject to the same reading of silence, the same reluctance to be the one who stalls the meeting, the same drift toward the emerging consensus.

There is a related trap called the shared information bias: groups spend most of their discussion time on facts everyone already holds in common and neglect the crucial thing that only one person knows. The quiet engineer who has seen this exact plan fail before doesn't mention it, partly because it cuts against the mood, and partly because no one asks. The decision gets made on the pooled common knowledge of the group, which is precisely the knowledge that made everyone agree in the first place. The dissenting fact never enters the record.

So the failure is not a failure of thinking. It is a failure of surfacing. The information that would have changed the decision existed in the room the whole time. It just never got said out loud.

Building in the doubt

What is striking about Janis's work is that his remedies are almost all structural. He did not ask people to be braver or more honest by force of will. He asked leaders to change the shape of the meeting so that dissent stopped being socially expensive.

The most durable of these is assigning a devil's advocate—not hoping someone will object, but making it a named role someone is obligated to fill. When disagreement is a job rather than a personal stance, the person voicing it pays no social cost. They are not being difficult; they are doing what was asked. The objection gets its hearing, and the group is forced to answer it rather than absorb it into the general nod.

Janis also urged leaders to withhold their own opinion until others have spoken. A leader who opens with their preference has effectively ended the meeting; everything after is people arranging themselves around that anchor. State the problem, not your answer, and let the room fill the silence before you fill it for them. He suggested breaking larger groups into independent subgroups that examine the same question separately, so their conclusions can be compared instead of merged. And he proposed the second-chance meeting: after a decision feels settled, reconvene once more with the explicit purpose of airing every remaining doubt before the decision hardens. Sleep on it, then try to talk yourselves out of it.

None of this requires heroism. It requires designing the meeting so that the quiet reservation has somewhere to go. The goal is not to manufacture conflict—it is to make sure the disagreement that already exists, silently, in people's heads actually reaches the table while it can still matter.

The record is the thing

The deeper fix behind all of Janis's remedies is simple: make the reasoning visible. Groupthink thrives in the ambiguity of the unspoken—the silence read as assent, the objection never recorded, the alternative never named. When a meeting captures not just what was decided but what was considered and set aside, and by whom, the illusion of unanimity loses its cover. A doubt that has been written down cannot be quietly forgotten. A dissent that is on the record cannot be mistaken for agreement later.

This is part of why meetingmortem exists—to give a team a running post-mortem of its own decisions, capturing what was chosen, what was rejected, and the reservations that were voiced along the way, so that the smooth meeting and the honest one can finally be told apart. The next time a room agrees a little too easily, you'll have a record of whether anyone actually doubted—and the standing invitation to ask. If your best meetings have started to feel suspiciously frictionless, it may be worth a look: https://meetingmortem.lumenlabs.works