What to do when your executive team splits into factions

When an executive team splits into factions, examine the shared decisions the groups keep obstructing. Map their incentives, resource dependencies, and conflicting instructions before asking for better teamwork. Establish one decision standard and an accountable owner for the work that crosses the boundary.

Executive faction conflict is a pattern in which competing leadership groups organize cooperation around group allegiance rather than agreed operating responsibilities. This is an operational description. A relevant research distinction comes from Jehn’s 1995 study, which separates disagreement about task content from interpersonal incompatibility. That paper does not supply a diagnostic test for factions or establish that one group has the better strategy.

Start where the split reaches the work

Composite scene: You are driving home after an executive meeting and call a trusted friend. You explain the argument by naming the old team and the new team. When your friend asks what the company decided, you cannot answer. The delivery issue remains unresolved, while you have supplied a convincing account of the camps.

Write down the unresolved decision before drawing a political map. Identify the customer commitment, allocation, or handoff affected. Ask what each executive has instructed their team to do. You need a record of the operating conflict, not an inventory of who likes whom.

Avoid assuming that people who agree regularly form an improper faction. They may share relevant information or hold a defensible view. The concern is whether allegiance prevents evidence from being considered or creates incompatible operating rules.

Distinguish a strategic disagreement from a coalition advantage

A team can disagree about a genuine tradeoff. One function may be protecting delivery reliability while another seeks faster customer commitments. The CEO must resolve the priority using evidence and proper authority.

A coalition advantage is different. A group may gain exceptions, private access, or information unavailable to others without an explained work reason. Verify the actual arrangement before assigning intent. What looks like favoritism may sometimes reflect a role requirement that has never been communicated.

For a hypothetical launch dispute, compare the approved sales promise with operations’ capacity plan. If they conflict, management has an operating problem even if both executives behave politely. If one side can bypass the shared approval whenever the decision is inconvenient, the decision process also needs repair.

The shared-decision faction review

1. Select a recurring boundary failure

Choose one decision that requires both groups to cooperate. Define what a usable outcome would look like. For the launch example, that might be an authorized customer date linked to an acknowledged delivery plan.

Avoid beginning with a workshop about the whole culture. A bounded decision lets you inspect actual behavior and gives participants a specific question to answer. If several disputes appear related, record the connection as a hypothesis rather than declaring a single hidden motive behind them.

2. Map incentives and dependencies

Ask what each group is accountable for, what resources it controls, and what it needs from the other side. Include the informal exceptions you grant as CEO. Use permitted operating records and direct questions, not covert monitoring of private relationships.

Emerson’s 1962 power-dependence theory examines influence through dependence and alternatives. Apply that lens to the handoff: who can delay a resource, who bears the cost, and what feasible alternative exists? The theory helps frame questions. It does not prove that a particular executive is manipulating the relationship.

3. Establish a common evidence standard

Specify what information must support a recommendation. In the launch example, require verified customer requirements and a current delivery assessment. Apply the same rule to proposals from both groups.

A familiar colleague should not win through historical access. A newly hired executive should not win because their language sounds more current. Compare the work assumptions and ask what evidence would change each recommendation. Let contributors correct material inaccuracies before a decision is made.

4. Assign ownership across the boundary

Name the decision owner, the required consultation, and the limits on private exceptions. Clarify who communicates the final instruction to the teams. If the owner lacks authority to resolve the relevant tradeoff, bring it to someone who has that authority.

Do not create joint accountability that leaves nobody able to choose. Several people can contribute evidence while one authorized role makes the decision. Where formal governance or employment authority is disputed, obtain qualified advice rather than settling it through a team exercise.

5. Inspect the next ordinary handoff

After the decision, examine a real transfer of work. Did the customer message match the delivery instruction? Did the required information reach the receiving manager? Did a private override change the plan without the proper review?

Follow the evidence back to the specific failure. A late handoff may expose a capacity issue. An unannounced override may expose an exception problem. Different failures require different repairs, even if the same groups are involved.

Give each group a fair chance to describe the problem

Use separate preparation where it helps gather facts, then bring the relevant evidence into a shared decision process. Ask each executive the same core questions: What are you trying to protect? What do you need from the other group? Which current instruction makes the work incompatible?

Do not ask people to reveal private loyalties or speculate about colleagues’ emotional states. You are evaluating work arrangements. If someone alleges misconduct, preserve the issue and use the appropriate HR, legal, or investigative route.

Edmondson’s 1999 research on psychological safety concerns interpersonal risk within teams. Applied here, it gives you a reason to ask whether challenging a camp’s position carries a cost. The paper does not establish that a structured meeting removes that cost. Inspect what happens to dissenting contributions after the meeting as well.

Use wording that shifts attention to the decision

Try this proposed opening:

We have competing instructions around [shared work]. I am not asking for statements of personal loyalty. I need each proposal assessed against [common evidence]. Explain the work you are protecting, the dependency on the other group, and the assumption that would change your recommendation. [Owner] will resolve the tradeoff through [authorized route], and we will apply the same exception rule afterward.

If someone says, “You always take their side,” ask for the concrete decision or exception they mean. Investigate that example. A general reassurance that you have no favorites will not answer evidence of inconsistent access.

If the complaint is accurate, repair your own practice. Redirect a private request into the agreed process and tell the relevant owners what changed. Do not expect executives to respect a boundary you routinely bypass.

Change the conditions that reward the split

In the hypothetical launch dispute, suppose the groups agree during meetings but each still gains approval through a separate CEO conversation. A team-building event will not resolve that arrangement. Stop providing incompatible private instructions and make changes visible to the accountable owner.

Suppose instead that both executives follow the rule, but their performance measures demand incompatible outcomes. Ask the authorized owners to review the measures and resource tradeoff. Seek employment or compensation advice where changes affect agreed terms.

Do not promise that removing one incentive will remove every conflict. The proposed review is an operating intervention, not a proven cure for organizational politics. Test whether the specific shared decision becomes workable.

Industrial/organizational psychology, the psychology of work and organizations, is useful because it connects group behavior to task arrangements and authority. Keep the application tied to those observable conditions rather than a story about good and bad personalities.

Preserve a sound proposal from an unpopular group

A faction review can create its own bias if management dismisses an idea because of the people advancing it. Separate the proposal from the access problem. A group may have used an inappropriate private route and still have identified a real customer constraint. Correct the route, then assess the constraint under the common standard.

Explain both decisions. State why the evidence changed the operating choice and why future requests must use the agreed channel. That makes it possible to correct the work without rewarding an exception or pretending the proposal lacked merit.

Do not force personal reconciliation

Executives may continue to disagree or prefer different colleagues after the process improves. The work requirement is compatible conduct and fulfillment of agreed responsibilities. A declaration that everyone is now one team adds little if no one can describe the handoff.

Where a past dispute still affects ordinary cooperation, use a future-work agreement with explicit response and escalation expectations. Do not make friendship, forgiveness, or a public apology the default condition for carrying out a role.

If the conduct includes retaliation, harassment, discrimination, threats, or other alleged misconduct, qualified HR and legal support must guide the response. A shared-decision exercise must not become a way to pressure someone to withdraw a complaint.

Review the pattern without creating a new faction list

Keep notes about decisions, resources, and observed conduct. Limit personal information to what the legitimate purpose requires. Do not circulate an unofficial roster of allies and enemies.

At the next review, compare a small set of actual handoffs with the agreed process. Look for equivalent evidence standards, visible exceptions, and a decision owner who can act. Ask whether people can raise a material concern without choosing a camp to obtain a hearing.

If the same failure returns, identify the unresolved condition. It may concern a missing decision right, a resource conflict, or your own private approval practice. Name that condition and assign its repair. The useful result is shared work that proceeds under a common rule even when executives retain different views about the company’s future.