Have Backbone, Disagree and Commit
Dissent is an obligation, not a right — and commitment only starts once you've been genuinely heard.
- Difficulty
- Moderate
- Time to result
- ~ongoing to results
- Steps
- 4
- Confidence
- 92%
The most misunderstood of Amazon's leadership principles. It has two halves and people usually skip the first: if you are part of a decision, you are obligated to voice a dissenting point of view, all the way up the chain if the issue is important. The commit half is not 'I'll go along with it' — it triggers only once the decision-maker has demonstrably heard, understood, and weighed your input, and it requires you to understand their reasoning well enough to explain it to your own organization.
Origin
One of Amazon's original ten leadership principles, described by Bill Carr as the least well understood and most nuanced of them. Carr grounds the underlying decision theory in Peter Drucker, whom he considers the best writer on decision-making.
Core principles
- 01The purpose of disagreeing is to supply information or a point of view the decision-maker has not considered — not to register displeasure.
- 02Good decisions come from first understanding all points of view and the pros and cons; great leaders actively solicit opposing views.
- 03The king-and-court model: absent politics, the court exists to give the king opposing views so the king can reach the right decision.
- 04You stop disagreeing at a specific, observable moment: when the leader reflects your reasoning back and shows they weighed it.
- 05Grudging compliance is not commitment.
How to run it
- 1
Voice the disagreement, with new information
If you are part of the unit making an important decision, you are obligated to speak. Bring additional data, a new point of view, or a consideration the group has not weighed — and escalate up the chain if an important issue is genuinely not being heard.
Pro tip Lead with the information, not the objection. 'Here's a factor I don't think we've weighed' beats 'I disagree'.
- 2
Listen for the reflect-back
The leader's job is to say either 'that's new, thank you' or 'I hear your disagreement, we've already considered that factor, and here are the other factors that outweigh it.' The disagreer's job is to check that the leader genuinely understands the reasoning and has taken it into account.
Pro tip If the leader cannot restate your argument, you have not actually been heard yet — keep going.
Watch out This is the moment people miss, which is why they don't know when they are supposed to stop disagreeing.
- 3
Commit on the basis of understanding, not obedience
Once heard and weighed, commit — but commit having actually processed the argument for the chosen direction, so you can reflect that reasoning back to your own organization.
Pro tip The test: can you explain to your team why this is the right call without editorialising?
Watch out Saying 'I'm committed but I still think it's wrong' to your team is not commitment — it is sabotage with good manners.
- 4
When you still don't agree, hunt for the kernel
Carr's method for the residual disagreement: find the kernel or core of why the other person believes this — the thinking vector they are on — and take that kernel and run with it. Expand it, productize it, make it viable within the real business constraints.
Pro tip Carr got this from his manager Steve Kessel: look for what the kernel is, then your job is to run with it.
Watch out Going through the motions — 'stomp stomp stomp through it' — is visible to everyone, and Carr has watched people's careers stall on it.
In the wild
Carr repeatedly found himself post-debate still disagreeing with Bezos, who thought on a creativity level and a time scale nobody else at Amazon did. Rather than executing resentfully, Carr would isolate the kernel of why Bezos wanted it, then work on how to make that kernel viable — how to productize it, how to make it work against the real business constraints.
→ Carr credits this posture — some degree of faith that there's something there, plus genuine effort on the kernel — as the difference between careers that go far at Amazon and careers that don't. It doesn't always work, but going through the motions never does.
As a leader, Carr regularly had people bring him a disagreement and cite reasoning he and the team had already worked through. His response was to say explicitly: I hear your disagreement, we have already considered that factor, and here are the other factors that outweigh it.
→ That explicit reflect-back is what earns the right to expect commitment — and it is the signal to the disagreer that the disagree phase is over.
Common mistakes
Skipping the backbone half
People remember 'disagree and commit' and forget it starts with 'have backbone'. Silence in the room followed by compliance is not the principle — voicing your view when you dissent is an obligation.
Not knowing when to stop disagreeing
Without the reflect-back signal, people either keep relitigating a settled decision or fold too early. The stopping condition is specific: the leader has demonstrated understanding of your reasoning and weighed it.
Fake commitment
'I'm committed to this but I don't really agree and I still think it's wrong' is not commitment. Real commitment means having processed the argument well enough that you can reflect the reasoning back to your own organization.
Is it for you?
Best for
Managers and senior individual contributors in high-debate cultures who need a clear stopping rule for dissent and a real definition of buy-in.
Not ideal for
Low-trust or politically-charged environments where dissent is punished — the mechanism assumes the leader will genuinely engage rather than retaliate.
From the transcript
“it is your obligation to voice your point of view if you disagree”
“as long as the disagreer is hearing back from the leader that they understand their point of view”
“we're committed to this I don't really agree and I still think it's wrong but I'm committed to it that's that's not that's not actually…”
“your job is to then take that kernel and try to run with it”
From the episode
Unpacking Amazon’s unique ways of working
Bill Carr (author of Working Backwards)