The Three Words That Stop Teams Going Round in Circles

By Akash Desai

Three words that saved Amazon countless hours, and could rescue your team from death by meeting. In April 1961, a room full of some of the sharpest minds in Washington signed off on the Bay of Pigs invasion. Several of them had serious doubts. Almost none of them said so out loud. The operation collapsed within three days, and the psychologist Irving Janis later made it his star example of a new idea he called groupthink: clever people agreeing their way into disaster.

Now picture the opposite problem. A team where everyone does speak up, and nobody ever stops. Every decision reopened. Every meeting ending with “let’s revisit next week”. Different symptom, same result: nothing worthwhile gets done.

The sweet spot sits between those two failures, and it has a name. Jeff Bezos made it famous, and you can start using it on Monday morning. It’s called disagree and commit.

The idea in one sentence

Argue as hard as you like before the decision is made, then back it fully once it is, even if you lost the argument. That’s it. Silence before the decision is cowardice. Resistance after it is sabotage. Disagree and commit asks for the reverse of both.

Where it actually came from

Bezos didn’t invent the phrase. It’s usually traced to Intel under Andy Grove, a company famous for blunt internal debate, and to Scott McNealy at Sun Microsystems. Amazon wrote “Have Backbone; Disagree and Commit” into its leadership principles around 2010. But it was Bezos’s 2016 letter to shareholders that turned it into a phrase leaders everywhere now use.

The story that makes it stick

In that letter, Bezos describes a show the Amazon Studios team wanted to greenlight. He wasn’t convinced. He thought it might not be interesting enough, would be complicated to make, and came with poor business terms. The team saw it completely differently.

So he wrote back straight away: “I disagree and commit”, adding that he hoped it would become the most-watched thing they’d ever made.

The lesson isn’t the show. It’s who said it. The most powerful person in the building had an opinion, voiced it, and then chose not to use it as a veto. As Bezos pointed out, the decision would have dragged on far longer if the team had needed to win him over rather than simply get his commitment.

Three ideas worth stealing

1. Decide at around 70%. In the same letter, Bezos argued that most decisions should be made with roughly 70% of the information you wish you had. Wait for 90% and you’re probably moving too slowly. Being wrong is often cheaper than being slow, provided you’re good at spotting mistakes and correcting course.

2. Know your doors. Bezos split decisions into one-way doors (irreversible, so take your time) and two-way doors (reversible, so walk through and walk back if needed). Most decisions are two-way doors. The classic mistake is running a slow, cautious process on choices you could easily undo.

3. Commitment gives you a clean test. If a team backs a decision wholeheartedly and it still fails, you’ve learnt something real about the idea. If they only half-tried, you’ve learnt nothing at all. You’ll never know whether the idea was weak or the effort was.

What it isn’t

Not “do as you’re told”. If people never get a genuine hearing, it’s not disagree and commit. It’s a dictatorship with better branding.
Not fake nodding. Polite agreement in the room and grumbling in the corridor is the exact disease this cures.
Not quiet sabotage. Doing the bare minimum so you can say “I told you so” later breaks the deal.
Not for crossing lines. If something is unethical, unsafe or illegal, you don’t commit. You escalate.

Try it with your team this week

Before any big call, ask: “Who sees this differently?” Then thank whoever answers, especially if they’re junior.
Make it obvious when the debate ends: “We’ve heard everything. Here’s the decision.”
Say the words out loud: “I disagree, but I’m in. Fully.” It turns a private doubt into a public promise.
Agree a review point in advance, so doubters have a fair moment to raise concerns with evidence instead of reopening it every Monday.
If the dissenters turn out to be right, credit them openly. Otherwise, next time they’ll stay quiet.

The real test of a leader

Anyone can commit to a decision they agree with. That’s not leadership, that’s just enthusiasm. The real test comes when you’ve made your case, lost, and still have to walk out and back the decision as if it were your own. Not grudgingly. Not with a raised eyebrow. Properly. So next time you’re in the room, say what you really think. Push hard. Then, when the call is made, commit like you wrote it.

Sources: