The standard model places conflict styles on two axes, assertiveness and cooperativeness, and none of the five modes is correct in every situation.
Why it matters when the plan changes
Plans run through interfaces, and interfaces are where interests collide. How the people at a boundary handle that collision decides whether it resolves, escalates or stalls. A team whose members all avoid conflict produces the same delay as a team whose members all compete for it, and neither delay looks like a conflict problem from outside. Cross-unit dependencies are the weakest link in execution; Sull, Homkes and Sull found that only 9% of managers say they can rely on colleagues in other functions and units all the time, which means the style each side brings to that boundary matters more than it appears to from a status report.
The tension is that no style is better in general. Competing settles a decision fast and costs commitment; accommodating preserves the relationship and defers the issue. The five modes sit on two axes, assertiveness and cooperativeness, and the useful question is whether the styles present fit the interfaces the plan runs through, not whether anyone is handling conflict correctly.
In practice
Two functions share a dependency and both leads default to accommodating. Each defers to the other's priority in the moment and neither raises the sequencing conflict that is actually blocking them. Relations are excellent, the milestone slips twice, and every status report describes alignment.
Evidence
The standard instrument identifies five conflict styles positioned on axes of assertiveness and cooperativeness.
Thomas-Kilmann Conflict Mode Instrument, Wikipedia (2026)Cross-unit commitments are the least reliable part of execution, which is where conflict style is most consequential.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)
What it cannot tell you
Conflict style describes a behavioural tendency, not the merits of the position someone is defending. Two people can each use a fitting style and still be wrong about the underlying decision. The model is also silent on power: a style that works between peers can fail entirely across an authority gap.
Questions
The Thomas-Kilmann Conflict Mode Instrument names five styles: competing, assertive and uncooperative; avoiding, unassertive and uncooperative; accommodating, unassertive and cooperative; collaborating, assertive and cooperative; and compromising, intermediate on both axes. Most people have a default mode and a fallback rather than a single fixed one.
No. It is expensive in time and attention and it is the right choice when the issue matters and both parties have to own the outcome. For a low-stakes decision under time pressure it is a way of avoiding a decision, which is why treating one style as correct produces its own failures.
At interfaces. Sull, Homkes and Sull, writing in Harvard Business Review in 2015, found only 9% of managers say they can rely on colleagues in other functions and units all the time. Conflict style at that boundary decides whether the gap resolves, escalates or shows up only as delay.
The default is fairly stable, and the range of styles someone can use widens with awareness and practice. What changes considerably faster is the situation: clarifying who decides removes the conflict altogether, rather than requiring anyone to handle it differently.
A team with only one style has a predictable failure mode, so variety helps. More important than balance is whether the styles present fit the interfaces the plan actually runs through, which is a question about the plan as much as about the people.