The dynamics that matter are the ones that appear when the group is under load, not the ones it displays in a workshop.
Why it matters when the plan changes
Most of what determines whether a leadership team can carry a plan happens in the interactions rather than in the individuals. A group that cannot disagree productively will approve plans it privately doubts, and a group where one voice dominates will make that person's blind spots the organisation's. Neither shows up in any individual assessment. Upward and cross-unit information weakens as it travels, so a team's account of its own functioning is systematically incomplete even when everyone answering is honest.
The tension is that dynamics are easiest to read exactly when they matter least. In a calm facilitated session a team displays its best behaviour honestly. The patterns that decide execution appear when three demands collide and someone has to lose an argument, which is not a setting anyone schedules. Conflict style research maps behaviour on axes of assertiveness and cooperativeness, from competing through collaborating to avoiding, and which mix shows up under load says more about a team than any of those styles named in isolation.
In practice
A leadership team rates its own collaboration highly and the rating is sincere. Under the pressure of an integration, decisions start being made in pairs before the meeting, and the meeting becomes ratification. Nobody decided to work that way. The dynamic under load is different from the dynamic the team knows about itself.
Evidence
Upward and cross-unit information weakens as it travels, so what a team reports about itself is systematically incomplete.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)Conflict styles sit on axes of assertiveness and cooperativeness, and the mix present shapes how a group handles disagreement.
Thomas-Kilmann Conflict Mode Instrument, Wikipedia (2026)
What it cannot tell you
Team dynamics describe how a group behaves together at a point in time; they do not explain why, and they say nothing about the skill or judgement of any individual member. A group can have healthy dynamics and still be wrong, or strained dynamics and still deliver. Dynamics observed in one setting do not reliably predict dynamics under a different kind of pressure.
Questions
Through how decisions actually get made rather than how the team describes its meetings: where a decision is really settled, what happens to a dissenting view, and how escalation is used. Harvard Business Review reported in 2015 that 84% of managers believe they can rely on colleagues, yet execution still stalls.
Because everyone sees the part they are in and nobody sees the whole, and because the behaviour people display in a reflective session is genuinely their best. Both are honest. The gap is structural rather than a matter of self-awareness.
How disagreement is handled, whether decisions stay decided, and whether bad news travels. The Thomas-Kilmann Conflict Mode Instrument frames disagreement across five styles, from competing to avoiding, and a team's mix under pressure shapes whether it approves plans it privately doubts or genuinely tests them.
Yes, and usually faster by changing structure than by changing behaviour. Naming who decides, changing what a forum is for, or moving a decision to a different level all change the dynamic without asking anyone to be different, which is why they are tried first.
Composition is what the group contains; dynamics are what it does with it. A well-composed team with poor dynamics under-delivers on the range it has, and a narrowly composed team with strong dynamics uses what it has fully. Both are worth reading, and neither substitutes for the other.