Disagree and Commit Needs Terms to Hold

The phrase travels well and arrives badly. A leadership team adopts it as a norm, usually after a long stretch of reopened decisions, and expects it to end the reopening. It rarely does. People are told to commit before anyone has established that their objection was understood, so the commitment is a posture and the objection goes underground. It returns in the delivery, as slow follow-through, quiet workarounds and a steady supply of reasons the plan is not working.

What the norm assumes

Disagree and commit is a two-part rule, and the order matters. The disagreement comes first, and it has to be complete: the objection stated, the evidence behind it named, the decider's answer to it given. Only then does the commitment mean anything. A commitment extracted before the objection has been answered is compliance, and compliance is what produces the pattern the norm was meant to cure.

The banks, product teams and sales organisations we work with tend to adopt the second half and skip the first. The decider announces the choice, someone says the team should disagree and commit, and the meeting ends. Nobody wrote down what the dissenting director thought would go wrong, so nobody can later say whether it did. The method below puts the missing half back, in four parts.

An objection that exists only in the room is deniable by everyone, including the person who raised it. Once the meeting is over, the dissenter remembers a stronger case than the one made, the decider remembers a weaker one, and the rest remember that there was some tension. A single written line fixes this.

The one-line dissent

Before the decision closes, whoever disagrees states the objection in a sentence and finishes it with a test: the thing that, if observed, would show the objection was right. A director who thinks a pricing change will drive away mid-size accounts writes that, and adds that renewal rates in that segment falling for two consecutive months would confirm it. The decider reads it back and says what the answer is.

This is not a courtesy to the dissenter. It converts an opinion into a forecast, and a forecast can be checked. It also makes dissent cheaper to offer, since an objection with a stated test is hard to read as obstruction. Teams that adopt this tend to hear fewer objections of the vague kind, because a person asked to name the evidence that would prove them right often finds they have not yet formed a view worth defending.

An objection that exists only in the room is deniable by everyone, including the person who raised it.

Most teams never say what committing involves, which leaves each person to supply a private definition. One reads it as not arguing in public. Another reads it as doing the work but not defending the decision to their own staff. A third reads it as full effort until the first bad number arrives.

Three observable commitments

The definition can be short, and it should be stated when the norm is introduced rather than invoked afterwards as a rebuke. In practice it comes down to three things a team can observe in each other.

• The decision is explained to the people who report in, as the team's decision, without reference to who opposed it.
• Resources agreed in the meeting are released on the date agreed, not held back pending further evidence.
• New information is brought to the decider through the agreed channel, and not into corridors and side threads.

None of these asks anyone to change their mind. A dissenter may go on believing the decision is wrong, and the stated test gives that belief a legitimate home. What the dissenter gives up is the right to act on the belief outside the channel. That distinction is what makes the norm acceptable to capable people, who tend to resist commitment when it is framed as agreement and accept it when it is framed as conduct.

The part of the norm most often left out is its end. A commitment with no term is a permanent suppression of dissent, and a capable team will sense that and resist it. The practical version carries a review date and a named signal.

A review date and a tripwire

The review date is set when the decision closes, typically at the next point where the team will have real data. At that date, the dissent line is read aloud next to what happened. The test either fired or it did not, and the discussion that follows is short because the terms were agreed in advance. The tripwire is the earlier version of the same thing: if the signal named in the dissent appears before the review date, the decider convenes the review early.

This is the link to how a team stops a decision from being reopened, which treats the reopening condition as part of the decision itself. Disagree and commit and the reopening rule are one mechanism seen from two sides. One protects the decision from casual revision, and the other protects the dissenter from permanent silence. A team needs both, because a decision that cannot be revisited breeds resentment and a decision that can be revisited at will is never made.

Where the norm does not apply

Some decisions are not eligible. Where the dissent concerns a regulatory obligation, a safety control or a legal exposure, committing to a plan the dissenter believes is non-compliant is not a team norm at work. It is a risk transferred to whoever stays quiet. Risk and compliance functions in a bank hold a different kind of right from product, and that distinction has to be stated before the norm is introduced, so that nobody is asked to commit to something they are professionally obliged to escalate.

Introducing the norm is a small piece of design with several parts. The meeting format needs a place for the dissent line. The decider needs to be named, since disagree and commit presupposes someone whose call it is, and the notes on decision rights describe how often that has not been settled. The review dates need a home in the operating calendar. The wider decision architecture pillar collects the surrounding pieces.

The test of the norm is whether a dissenter can say, a quarter later, what they objected to and whether they were right. If they cannot, the objection was never recorded and the commitment was never earned.

The first piece of work is retrospective. Take the last five decisions that were contested and note, for each, whether the objection was written anywhere, whether the commitment was ever defined, and whether a review date existed. The gaps usually explain why the team keeps returning to the same questions. That diagnosis, and the redesign of the decision forum that follows it, is the kind of work our consulting practice does with leadership teams, and our training programmes cover the facilitation habits that keep the dissent line and the close in use after the first few meetings.