3.2. Manage Conflicts
💡 First Principle: Conflict itself isn't the failure state PMI is testing for — an unmanaged or wrongly-resolved conflict is. The exam expects you to diagnose the conflict's source before selecting a resolution strategy, because the "right" strategy depends entirely on what's actually driving the disagreement.
Enablers: identify conflict sources, analyze the context surrounding the conflict, implement an agreed-on resolution strategy, communicate conflict-management principles to the team and external stakeholders, establish an environment that fosters adherence to ground rules, and manage/rectify ground-rule violations.
Common conflict sources include competing priorities, scarce resources, personality clashes, and ambiguous authority — each calls for a different resolution approach (collaborating, compromising, smoothing, forcing, withdrawing). PMI's scenario questions almost always describe the conflict's source in the stem, then ask which resolution technique fits — matching technique to source, not defaulting to one favorite style, is the tested skill.
Conflict is not the enemy
Constructive conflict drives better decisions, surfaces hidden risks, and prevents groupthink. A servant leader deliberately builds an environment where people can disagree safely, because a team that never disagrees is usually a team that has stopped telling you things. Conflict is also expected at certain moments: in Tuckman's model the storming stage is where personality clashes and disputes over roles and decisions appear. The right response is to recognize the stage and help the team clarify roles and agree ground rules — not to treat it as a failure. A team whose members have worked together before, using familiar methods, often passes through forming and storming quickly for exactly that reason.
The resolution sequence
Identification often means looking past the presenting complaint. Two leads who have stopped talking and a 20% velocity drop call for a facilitated conversation to find the root cause; a team resisting agile because executives once weaponized velocity is best served by addressing that cause — for example, a flow-based Kanban approach that does not rely on velocity — rather than by overruling the objection.
The five modes
The Thomas-Kilmann model maps five modes across two dimensions — assertiveness (pursuing your own concerns) and cooperativeness (pursuing the other party's). Its lesson is that no mode is universally right.
| Mode | Outcome | Use it when |
|---|---|---|
| Collaborating / problem solving | Win-win, integrating viewpoints | The issue matters, there is time, and parties can explore underlying interests |
| Compromising / reconciling | Partial win for each side | Time is limited and a divisible middle ground is acceptable |
| Smoothing / accommodating | Yield to the other party | The issue matters more to them, the relationship is the priority, or you are wrong |
| Forcing / directing | Win-lose, using authority | A decision cannot wait and the right answer is clear — safety, legal, ethical, emergency, hard deadline |
| Withdrawing / avoiding | Issue deferred | The issue is trivial, emotions are too high, or you need more information first |
Collaboration works on interests, not positions: when two stakeholders hold incompatible requirements, bring them together, surface the needs behind the stated demands, and guide them to a solution or a clear prioritization. This is also PMI's preferred negotiating stance — win-win, built on integrity, mutual trust, and seeing the situation from the other party's point of view.
⚠️ Exam Trap: Compromising a technical decision by splitting it in half. Compromise suits divisible things — budget, time, resources. Half of one architecture plus half of another produces an incoherent design; when time is short and a clear decision is needed, forcing is the better fit.
Read the clock in the stem. Two seniors disagreeing with three weeks left still allows a facilitated, collaborative session against project requirements; the same dispute stalling progress with the deadline at hand calls for a prompt decision. Likewise, when a sponsor who controls funding has heard the team's concern and directs you to proceed, accommodating that decision — while recording the concern — is appropriate; forcing your view on the person with the authority is not.
Where, and by whom
- Private first for individuals. Dismissive comments, behind-the-back criticism, or a colleague quietly covering for an underperformer are handled in private conversations — listen to and validate the affected person, then address the behavior and its impact on trust with the person responsible.
- Reframe in the room. When a developer attacks the product owner in a retrospective, acknowledge the frustration and reframe it as a process issue — how priorities are managed — rather than a personal one.
- Agile teams self-manage most conflict through working agreements and retrospectives; the Scrum Master facilitates rather than decides.
- Escalate when the conflict involves organizational policy, legal or ethical issues, decisions beyond your authority, or when team-level attempts have genuinely failed. Three projects competing for one specialist is not a fight for the loudest PM to win — the PMO allocates by portfolio priority.
Ground rules and the team charter prevent friction by setting expectations for behavior, communication, and decision-making. When a rule is broken, address it promptly and consistently. Cultural awareness matters here too: norms for disagreement, directness, and authority vary, and ignoring that produces misreadings of intent.
⚠️ Exam Trap: Defaulting to "collaborate/problem-solve" as the universal right answer. It's usually correct for value-based disagreements with time to resolve them, but forcing/directing can be the right (and PMI-expected) call when there's a hard deadline and a clear right answer.
Reflection Question: Recall a conflict you resolved by compromising when collaboration would have produced a better outcome (or vice versa). What made you choose the technique you did?
