What Makes a Good Team Produce a Poor Decision?

Group Think

When agreement becomes more dangerous than ignorance

Framing the Question

Groupthink is what happens when a team’s desire for harmony, confidence, or speed quietly outranks its desire for truth. A good team can make a poor decision not because its members lack intelligence, but because the group structure punishes doubt and rewards agreement. The danger is especially high when the stakes are high, the leader is strong, the team is cohesive, and dissent feels socially expensive. The better question is not “How smart are the people?” but “What is the group making hard to say?”

Why This Question Matters

A good team produces a bad decision when its intelligence gets trapped inside its social dynamics.

That is the short answer.

Groupthink occurs when members of a cohesive group accept what seems to be the group consensus, even when they privately doubt it. Britannica describes groupthink as a mode of thinking where people in small cohesive groups tend to accept a conclusion that represents perceived consensus, whether or not they personally believe it is valid or optimal. Irving Janis developed the theory in his 1972 work on foreign-policy failures.

That is why the question matters. We often assume bad decisions come from bad information, weak talent, or poor effort. But some decisions fail for a stranger reason: the team has enough intelligence to notice the problem, yet not enough psychological permission to say it clearly.

The room knows. The room does not speak.

This is how bright people approve weak strategies, launch risky products, ignore warning signs, or pretend a plan is stronger than it is. The bad decision is not born from stupidity. It is born from suppressed friction.

What the Question Reveals

The question reveals a central tension in teamwork: teams need cohesion to move, but too much cohesion can make them blind.

A team is not just a collection of brains. It is a social system. People watch status, tone, facial expressions, career risk, and who gets rewarded for being “constructive.” Over time, they learn what kind of truth is welcome.

Groupthink thrives when several forces combine:

• Authority bias: people defer too quickly to the highest-status voice.
• Social proof: silence gets mistaken for agreement.
• Confirmation bias: evidence that supports the preferred plan gets more attention than evidence that challenges it.
• Self-censorship: people trim their concerns before anyone else has to.
• Illusion of unanimity: because no one objects, everyone assumes everyone agrees.

The most dangerous sentence in a meeting is not “I disagree.” It is “Sounds good to me,” said by someone who has serious doubts but no appetite for becoming the problem.

A useful analogy is a smoke detector with the batteries removed. The room becomes quieter. Everyone feels less interrupted. But the system has lost the very signal designed to prevent disaster. Groupthink does the same thing to teams. It removes the alarms and calls the silence alignment.

A Real-World Example

The Bay of Pigs invasion is one of the classic examples associated with groupthink. Janis’s theory examined foreign-policy decisions including the Bay of Pigs, and the failed 1961 invasion remains a useful case for understanding how capable leaders and advisers can make poor collective judgments.

The JFK Library explains that Kennedy was briefed before his inauguration on a CIA plan developed during the Eisenhower administration to train Cuban exiles for an invasion of Cuba. The plan assumed the Cuban people and parts of the Cuban military would support the invasion, with the goal of overthrowing Fidel Castro.

The operation went badly. Bombers missed many targets, Castro’s air force remained largely intact, a second air strike was canceled, and the invasion force came under heavy fire after landing at the Bay of Pigs. Nearly 1,200 members of Brigade 2506 surrendered, and more than 100 were killed.

The lesson is not that one person should have magically known everything. The lesson is that high-stakes teams need systems that force doubt into the room before reality does.

Smart people are especially vulnerable here because intelligence can become a tool for rationalization. A talented group can explain away weak evidence more elegantly than an average group. It can build a beautiful bridge to the wrong conclusion.

A Different Perspective

Instead of asking:

“What makes a good team produce a poor decision?”

Ask:

“Where is our team protecting agreement at the expense of evidence, dissent, or reality?”

That sharper question changes the investigation.

The original question points to the failure. The better question points to the mechanism. It asks the team to inspect the social cost of truth. Who speaks first? Who speaks last? What happens to dissent? Are objections explored or quietly labeled as negativity? Are we testing the idea, or protecting the people who proposed it?

This is the QuestionClass move: improve the question until it reveals the hidden structure behind the answer.

What to Do With This

To prevent groupthink, do not simply tell people to “speak up.” That is too vague. Build dissent into the process.

Try these practices:

1. Have the leader speak last. Early leader opinions often become invisible instructions.
2. Use private pre-votes. Ask people to write their recommendation before discussion begins.
3. Assign a red team. Make challenge a role, not a personality flaw.
4. Ask for disconfirming evidence. Do not ask, “Does everyone agree?” Ask, “What would make this plan fail?”
5. Run a premortem. Imagine the decision failed six months from now. What caused it?
6. Reward useful dissent publicly. People repeat what gets protected.

The practical insight is simple: better decisions require designed disagreement. Not endless debate. Not cynicism. Not obstruction. Designed disagreement.

A team does not need more conflict. It needs better channels for reality.

Bringing It Together

A good team produces a poor decision when belonging becomes more important than accuracy.

Groupthink is not a character flaw. It is a design flaw. It appears when teams mistake smoothness for strength, confidence for evidence, and silence for consent. The cure is not to make meetings louder. The cure is to make better questions normal.

Before the next major decision, ask: “What are we not saying because it would make the room uncomfortable?”

That question may be the one that saves the decision.

Follow QuestionClass’s Question-a-Day at questionclass.com to practice asking better questions every day.

 Bookmarked for You

These books help explain why intelligent groups sometimes fail, and how better questioning can protect collective judgment.

Victims of Groupthink by Irving L. Janis - The foundational book on how cohesive groups suppress dissent and make flawed high-stakes decisions.

The Wisdom of Crowds by James Surowiecki - Useful for understanding when groups become smarter than individuals—and when they do not.

The Fearless Organization by Amy C. Edmondson - A practical guide to psychological safety, showing why people need permission to speak candidly before teams can learn.

 QuestionStrings to Practice

QuestionStrings are deliberately ordered sequences of questions in which each answer fuels the next, creating a compounding ladder of insight that drives progressively deeper understanding.

Dissent Before Decision String
For when a team seems aligned too quickly:

“What decision are we about to make?” →
“What evidence supports it?” →
“What evidence makes us uncomfortable?” →
“Who would see this differently, and why?” →
“What would we change if disagreement were easier to voice?”

Use this before approving a strategy, hiring decision, product launch, budget move, or major policy shift. The goal is not to slow everything down; it is to prevent false alignment from becoming expensive action.

The smartest teams are not the ones with no disagreement; they are the ones that know how to turn disagreement into better judgment.