Leadership · The leadership decisions behind transformation

Everyone Agreed. So Why Hasn’t Anything Changed?

By Isabella Brusati · 12 November 2026

Agreement becomes useful when leaders can explain what they will change in their own function to honour it.

Imagine a post-acquisition integration meeting. Sales, operations and finance endorse a common customer process. The sponsor records full alignment. Three weeks later, sales still permits local exceptions, operations expects standard service rules and finance prepares a single reporting structure. Each leader believes they are implementing the agreement. Managers inherit the contradictions.

In this hypothetical situation, nobody needs to withdraw support for progress to stall. The group agreed to words that allowed several different interpretations of the work.

My argument is that alignment should be tested through commitments before it is reported as achieved. Ask what each leader will resource, stop, decide and accept. Those answers expose differences while the people with authority are still together, rather than leaving managers to discover them during implementation.

Why a positive meeting can produce little movement

Broad language makes agreement easy because it postpones the distribution of costs and authority. A leader can endorse standardisation while assuming their own exceptions remain. Another can approve a timetable while expecting someone else to provide the specialists. The apparent consensus leaves the consequential choices open.

Silence is equally ambiguous. It can reflect agreement, uncertainty or a belief that speaking will achieve little. Amy Edmondson’s study of 51 manufacturing teams found an association between psychological safety and learning behaviour. It supports attention to the conditions for speaking up; it does not establish why a particular meeting is quiet.

McKinsey’s 2025 analysis found that only 22% of the transformations examined prioritised decision-making behaviours. My practical recommendation is to make the content of agreement observable.

1. Ask leaders to translate the decision independently

Before the next cross-functional review, ask each accountable leader for a short written answer to four questions: What outcome have we agreed? What will my function do differently? What must we stop or defer? Which consequence or uncertainty are we accepting?

Collect the responses before discussion. This gives each person room to articulate their understanding without following the first confident answer in the room.

Compare the differences openly. In the integration example, decide where standardisation ends and local discretion begins. If several functions expect the same specialist time, allocate it and identify what other work must move. The output should be a specific unresolved choice, not an average alignment score.

2. Turn disagreement into an implementable commitment

Ask someone close to the work where the proposal first collides with an existing obligation. Separate disagreement with the outcome from evidence that the proposed route is unworkable.

A manager who cannot release a testing team may support the transformation and still be right about service exposure. Require evidence and available options, but allow a material concern to be raised before the entire solution is known.

Use the established decision authority when consensus is unavailable. Record the owner, resources, accepted consequence and review trigger. Ask each affected executive to confirm the action they will take in their own function. People can implement a decision they did not prefer when its meaning and authority are clear.

3. Inspect what changed below the meeting

Within two weeks, ask affected managers to show the practical consequence. Have named people been released? Has an old report stopped? Has an approval route changed? Do performance discussions reflect the agreed priority?

If the executive commitment is clear but local instructions conflict, bring the discrepancy to the accountable leaders. Avoid sending another general message that leaves managers to reconcile it themselves.

Keep justified exceptions visible. Record who authorised one, why it is needed and when it will be reviewed. Otherwise, separate accommodations can gradually replace the agreement without an explicit decision.

Measure follow-through without rewarding silence

Track the actions due, resources actually released and decisions reopened because their meaning was unclear. Examine reasons as well as counts. New evidence can justify reconsideration; ambiguous commitments call for a different remedy.

More reported conflicts may initially mean that people are describing the work more honestly. Ask whether those conflicts reach the right owner earlier and whether the resulting decision holds. A target of no disagreements can hide the information the group needs.

The test of alignment comes after the meeting, when a leader’s own priorities must change.

Research context: Edmondson, “Psychological Safety and Learning Behavior in Work Teams” (1999); McKinsey, “How to capture the elusive performance edge in true transformations” (2025). The integration example is hypothetical. The commitment test is my recommendation.

Related reading: Change management cannot resolve a decision leadership has not made; When Strong Leadership Becomes a Barrier to Transformation.

← Back to Insights