How do I resolve conflict between departments?

Conflict between departments is not a sign that something has gone wrong. It is a sign that both functions care enough about their work to defend it. The problem is not the conflict. It is what most matrix leaders do with it.

I was working with a cross-functional programme lead in a global consumer goods company, managing a product launch that required sales and product development to work together across a twelve-month delivery cycle. The conflict started quietly. Product development felt that sales was constantly moving the goalposts, bringing in late customer requirements that disrupted the development roadmap. Sales felt that product development was too internally focused, building features that missed what customers actually needed. Both were right. Both were also describing the same situation from positions so far apart that they had stopped being able to hear each other.

By the time she called me, the conflict had moved past the work and into the relationship. Emails were being copied to senior leaders as a form of protection. Meetings were producing minutes that each side interpreted differently. A decision that had apparently been made three weeks earlier was being relitigated because one function had never really accepted it. The practical problem, how to manage late-stage customer input without destroying the roadmap, was entirely solvable. But the relationship had deteriorated to the point where neither side could engage with the practical problem without the history of the conflict getting in the way.

The instinct in this situation is to find the compromise. Split the difference, give each side something, and move on. This is the wrong approach, for two reasons. First, a compromise leaves both sides with less than they need, which means neither is genuinely committed to the resolution. Second, it settles the surface dispute without touching the condition that produced it. The same conflict will return, usually within a few months, with different content but the same underlying shape.

Functional conflict is structural, not personal

Consider two programme leads dealing with similar inter-departmental conflicts, one in logistics and procurement, one in technology and operations.

The first treats the conflict as a personality problem. She believes that if she can get the two department heads into a room and facilitate a productive conversation, the tension will ease. She convenes a meeting. The conversation is civil. Two weeks later, the conflict has resumed because nothing about the structural conditions that produced it has changed. The department heads are still being measured on targets that pull in opposite directions. The boundary between their functions is still undefined in the area where both feel ownership. The goodwill from the meeting has a short half-life.

The second programme lead treats the conflict as a structural problem with a relational dimension. She still brings people together, but before she does she spends time understanding what kind of conflict this actually is. Is it a boundary dispute, where nobody clearly agreed who owns what and both functions have expanded into the same space? Is it an incentive misalignment, where the two functions are being measured on targets that are structurally incompatible? Or is it a relationship breakdown, where trust has eroded to the point where cooperation is no longer the default and every interaction carries the weight of previous failures? Each of these requires a different response, and treating them all the same way is why the same conflict keeps resurfacing in different forms.

This is the most important shift in how to think about inter-departmental conflict in a matrix. The conflict almost always presents as a dispute about something specific: a resource allocation, a missed deadline, a decision that was made without adequate consultation. But the specific dispute is usually a symptom. The condition producing it is structural: a boundary that was never defined, an incentive system that rewards one function for doing something that costs another, or a relationship that has accumulated enough unresolved tension that the parties can no longer engage in good faith.

Resolving the symptom without addressing the condition is temporary relief. Addressing the condition is what actually changes the pattern.

This article explores how to do both. It is the second of the practical challenge articles in this series, following the deep dives on influence, networks, convening, accountability, and decision making. The foundation article, How do successful leaders create commitment in matrix organisations?, introduced the capabilities that underpin everything in this series. The other articles cover what a matrix organisation is and why companies adopt one, how to build influence without authority, how to build a powerful internal network, how to convene people who don’t report to you, how to create accountability without authority, how decisions get made in a matrix organisation, and how to manage competing priorities in a matrix.

Why inter-departmental conflict is so hard to resolve in a matrix

Treating every conflict as a personality problem: The easiest explanation for functional conflict is that the people involved are difficult, territorial, or not team players. This explanation is almost always incomplete and often wrong. It focuses attention on the individuals rather than the system producing the behaviour, which means the conditions that generated the conflict remain unchanged even when the individuals are managed, moved, or replaced.

Smoothing rather than resolving: Many matrix leaders, aware that they have no formal authority over either function, default to a conflict management style that prioritises keeping relationships comfortable over addressing the real issue. Tensions are softened, concerns are acknowledged without being acted on, and the conflict goes underground rather than being resolved. Underground conflict is more damaging than visible conflict because it accumulates interest. Every unresolved tension makes the next interaction slightly harder.

Compromising when integration is possible: The instinct to find a compromise, to give each side something and move on, is understandable but often produces a worse outcome than a fuller resolution would. A compromise leaves both sides with less than they need. It also signals that the way to get what you want in this organisation is to hold out for a better split, which teaches both functions the wrong lesson about how conflicts get resolved.

Resolving the dispute without addressing the relationship. It is entirely possible to settle the practical question at the centre of a conflict while leaving the working relationship damaged. When that happens, the next interaction between the two functions starts from a lower baseline of trust than the previous one. Over time, the relationship becomes so loaded with history that even routine collaboration feels adversarial:

Staying in the role of mediator when a structural change is needed: A matrix leader can resolve individual conflict episodes indefinitely and still be managing the same underlying conflict in a different form twelve months later, because the structural condition producing it, an unresolved boundary, a misaligned incentive, a resource scarcity that pits both functions against each other, has never been addressed by anyone with the authority to change it. Resolving conflict in a matrix sometimes means escalating the structural condition rather than personally mediating its symptoms.

Avoiding conflict entirely under the guise of collaboration.:The most insidious failure mode is not escalation but avoidance. Leaders who prize harmony over honesty allow conflicts to fester because naming them feels more disruptive than absorbing them. Research on psychological safety shows that teams with genuinely high safety are not conflict-free. They have more visible conflict, because people feel able to surface disagreement before it becomes entrenched. Avoiding conflict is not the same as having good relationships. Often it is the opposite.

Six practices for resolving conflict between departments

Diagnose the conflict before attempting to resolve it

A boundary dispute, an incentive misalignment, and a relationship breakdown look similar from the outside, and all three often present as an argument about something specific. Applying the wrong resolution to the wrong type of conflict wastes time, generates goodwill costs that are not recovered, and leaves the actual problem intact.

A boundary dispute is characterised by both functions expanding into shared space without anyone having clearly defined who owns what. Both sides feel they are defending legitimate territory. Both are right about their perception and wrong about the totality of the situation. Resolution requires clarity, not mediation: a clear definition of where one function’s responsibility ends and another’s begins, agreed explicitly and, ideally, documented.

An incentive misalignment is characterised by both functions doing exactly what they are being asked to do and producing a conflict as a direct result. Sales optimising for short-term revenue and product development optimising for technical quality will generate tension by design if nobody has resolved the trade-off between those two objectives at the system level. Resolution requires a structural change that a matrix leader can surface and advocate for but usually cannot implement alone.

A relationship breakdown is characterised by a conflict that has accumulated enough history that the parties can no longer engage with the practical question without the relationship getting in the way. Resolution requires addressing the relationship directly before attempting to work on the practical issue, because every attempt to work on the practical issue will be filtered through the accumulated mistrust.

A finance director I worked with described a conflict between her treasury function and the group’s regional finance teams that had been running, in different forms, for three years. Four different matrix leads had attempted to mediate it, each settling the current episode without diagnosing what was actually producing them. When she mapped the conflict properly, the underlying cause was a performance measurement system that rewarded the treasury function for central cash pooling and the regional teams for local liquidity, targets that were structurally incompatible in the way the business operated. No amount of relationship work or boundary clarification was going to resolve a conflict that the incentive system was actively regenerating. The diagnosis changed what resolution actually required.

The programme lead’s conflict between sales and product development was primarily a boundary dispute with a relationship dimension. The boundary between “what customers want” and “what goes on the roadmap” had never been formally defined. Each function had assumed ownership of the decision, and both were right that they had a legitimate interest in it. The relationship had deteriorated as a consequence of repeated boundary collisions, not the other way around. That diagnosis changed her approach entirely: boundary clarity first, relationship repair second.

Questions you can use to diagnose the type of conflict you are facing:

• Is this conflict about who owns what, about competing targets, or about the accumulated history of how the two functions have treated each other?
• If the two people most visibly in conflict were replaced tomorrow, would the conflict continue in roughly the same form with different people?
• Are both functions doing what they have been asked to do, and producing a conflict as a direct result?
• Is the practical issue at the centre of this conflict actually resolvable, or is it a symptom of something neither function controls?
• How long has this conflict been running, and what has each previous resolution attempt actually addressed?

Practical actions:

• Before convening any resolution conversation, spend time separately with each function understanding their view of what the conflict is actually about. Listen for the structural story underneath the interpersonal one.
• Map the conflict against three questions: is there an unclear boundary, an incompatible incentive, or a damaged relationship? Treat any answer of “all three” as a signal to sequence your response rather than attempting to address everything simultaneously.
• For boundary disputes, seek written clarity before seeking relational warmth. A well-defined boundary does more for a damaged relationship than a facilitated conversation that leaves the boundary ambiguous.
• For incentive misalignments, document the structural incompatibility clearly and escalate it to whoever has authority to address it, separately from managing the immediate conflict episode.

Name the conflict as legitimate rather than smoothing it over

The most common first response to visible inter-departmental conflict is to reduce its visibility. Leaders step in to smooth things over, acknowledge both sides, find points of agreement, and restore a surface calm that everyone implicitly agrees to maintain. This is understandable and almost always counterproductive.

Smoothing a conflict over sends two messages simultaneously. The first is that the tension is too dangerous or too uncomfortable to address directly. The second is that the way to get attention paid to a problem is to make it visible and uncomfortable. Both messages produce exactly the behaviour that the smoothing was intended to prevent.

The alternative is to name the conflict as legitimate: to treat the fact that two functions see this differently not as a failure of communication or goodwill but as a natural consequence of having two functions with different responsibilities, different information, and different pressures. This is not permission for the conflict to continue unresolved. It is the stance from which productive resolution becomes possible, because it removes the implicit judgment that one side is right and one is wrong and replaces it with the more accurate framing that both sides are right about something and the organisation has not yet resolved the tension between those two things.

An IT director at a professional services firm described a conflict between her infrastructure team and the client-facing technology team that had reached the point where both teams were avoiding each other at the project planning stage. She had tried three times to broker a resolution through bilateral conversations, each time treating the conflict as a communication problem to be fixed rather than a structural tension to be named. When she changed her approach and convened a conversation that opened with “these two teams have genuinely different and legitimate priorities, and the organisation has not resolved what happens when those priorities conflict,” the tone in the room changed within ten minutes. Both teams visibly relaxed. Nobody had given them permission to have the conflict they were having. Naming it as structural rather than personal made it possible to talk about it honestly for the first time.

The programme lead used the same move with her sales and product development teams. Rather than opening the resolution conversation with an appeal for collaboration and goodwill, she opened it by naming the tension directly: these two functions are being asked to do things that sometimes pull in opposite directions, and neither team is wrong to feel that pressure. The acknowledgement did not resolve anything on its own. It changed the conditions in which resolution was possible.

Questions you can use to name a conflict as legitimate:

• Have you explicitly told both functions that the tension between them is a natural feature of how the organisation is structured, rather than a sign that someone is failing?
• What would it mean for each function to be told that their frustration is understandable given the pressures they are under, without that being used as a reason to avoid the resolution conversation?
• Is the conflict currently being treated as a communication problem when it is actually a structural one, and what would change if you named that distinction out loud?
• What is each side’s most legitimate concern, the one that deserves to be taken seriously regardless of how the conflict has been conducted?
• What would it mean for this conflict to be resolved in a way that neither side had to pretend was satisfactory?

Practical actions:

• Open any resolution conversation by naming the structural tension that produced the conflict, before addressing the specific dispute. This separates the people from the problem and gives both sides permission to engage honestly.
• Resist the impulse to reduce the visibility of the conflict before you have addressed its cause. Underground conflict is not resolved conflict.
• When one function complains about another, acknowledge the legitimacy of the frustration before moving to problem solving. “That makes sense given what you are being asked to deliver” is not capitulation. It is the precondition for a productive conversation.
• Make it explicit that you are not there to decide who is right. You are there to help both functions find a way through a situation that the organisation has created and that neither function can resolve alone.

Separate the functions from the individuals

Inter-departmental conflict in a matrix almost always presents as a conflict between individuals. The sales director and the product director cannot agree. The regional finance team and the treasury function are at odds. The operations lead and the IT lead have stopped communicating directly. But the conflict is rarely, at its core, about those individuals. It is about the roles they occupy, the pressures those roles create, and the system that puts both roles in competition for the same resources, decisions, or territory.

This distinction matters practically because it changes both what you look for and what you do about it. If the conflict is between individuals, the solution is to change the individuals or change their behaviour. If the conflict is between functions, the solution is to change the conditions in which the functions interact. Most matrix leaders who address the individual dimension without the functional dimension find that the conflict reconstitutes itself with different people.

It also matters because taking responsibility for your own contribution to a conflict, before asking the other side to take responsibility for theirs, is both more honest and more effective than waiting for the other party to move first. Each side in a functional conflict almost always has a legitimate grievance and has also done something to make the situation worse. The leaders who accelerate resolution are those who can say, on behalf of their function, “here is what we have done that has made this harder,” before making any claim about what the other side has done.

A supply chain director I worked with described a conflict between his procurement team and the engineering function that had been running for eighteen months. Both teams had legitimate grievances. Procurement felt that engineering kept bypassing the agreed supplier process for technical reasons that procurement could not evaluate. Engineering felt that procurement was applying commercial criteria to technical decisions without the expertise to make those judgements. He had been managing the conflict bilaterally, spending time with each team separately and trying to find workable compromises. When he changed his approach and brought both teams together with the explicit framing that both functions had contributed to the current situation and both needed to identify what they would do differently, the conversation moved faster in one session than it had in six months of bilateral management.

The programme lead had been spending most of her time managing the emotions of the conflict, talking separately to each function and trying to keep both sides calm. When she shifted to separating the functional dynamic from the personal one and named explicitly what each function had done that had made the situation harder, the sales director and the product director each became more willing to acknowledge their own contribution. The separation of function from individual had given them a way to be honest without it feeling like a personal concession.

Questions you can use to separate functions from individuals:

• If you described this conflict in terms of the roles and pressures involved, rather than the people, would it look structurally predictable rather than exceptional?
• What has each function done that has made the conflict harder to resolve, separate from the question of who started it or who is more at fault?
• What would each function need to do differently, going forward, regardless of what the other function does?
• Is the person most visibly in conflict the cause of it, or the most visible expression of a systemic tension?
• What would you say to each function about their own contribution to this situation, and have you actually said it?

Practical actions:

• When describing the conflict to either party, use functional language rather than personal language wherever possible. “The way the two teams have been working together” rather than “what your colleague has been doing.”
• In any joint resolution conversation, ask each function to identify what they have done that has contributed to the current situation before asking them to identify what the other side has done. This sets a norm of self-accountability rather than mutual blame.
• Where a specific individual is genuinely making the conflict harder, address that separately and directly, not in the joint resolution session where it will derail the structural conversation.
• After a resolution has been reached, ask both functions what they will each do differently, not just what they have agreed to do together.

Look for integration, not compromise

When two functions are in conflict, the default resolution is compromise: each side gives something up and gets something back, and the practical matter is settled at a point that satisfies neither side fully. Compromise is fast and it is often better than continued conflict. It is rarely the best available resolution.

Integration is the alternative. An integrated resolution finds a solution that meets both functions’ underlying needs fully rather than splitting the difference between their stated positions. It requires more work and more honesty than compromise, because it demands that both sides articulate what they actually need rather than simply defending what they asked for. But it produces a resolution that both sides are genuinely committed to, because neither has had to sacrifice something important to get there.

The distinction between compromise and integration maps directly onto the distinction between positions and interests from the competing priorities article. A compromise settles the positions: each side gets less than it wanted. An integration addresses the interests: each side gets what it actually needed, through a solution that neither side could have found alone. The integration is only possible when both sides have been willing to say what they actually need rather than simply defending what they originally asked for.

A marketing director described a conflict between the brand team and the regional sales teams over campaign localisation that had been generating friction across three regions for two years. The brand team’s position was that campaign consistency was non-negotiable. The regional teams’ position was that local adaptation was essential for market relevance. Every previous resolution had been a compromise: a defined percentage of campaign content that could be adapted locally, set at a level that satisfied nobody. When a new matrix lead asked each side to articulate not what they wanted but what they actually needed, it emerged that the brand team’s real concern was a small set of brand identity elements that were genuinely non-negotiable, and that the regional teams’ real need was creative latitude in execution rather than in the core brand identity. The integration that followed gave the brand team full protection of the elements they actually cared about and gave the regional teams far more latitude than any previous compromise had allowed. Both sides described it as a better outcome than they had expected.

The programme lead’s conflict between sales and product development had been managed through compromises for months: a defined number of late requirements that would be accepted per sprint, a scoring system for prioritisation, a set of escalation criteria. Every compromise had held for a few weeks and then broken down. When she ran an integration conversation, asking each side what they actually needed rather than what they were asking for, the sales team’s real need was confidence that customer-critical feedback would reach the roadmap before it was too late to act on. The product team’s real need was enough stability in the requirements to build to a quality they could defend. The integration was a structured early-warning process that gave sales a clear and fast route to raise genuinely critical customer input, separate from the regular requirements process. Neither side had asked for it. Both sides immediately saw that it served their actual need better than any previous compromise had.

Questions you can use to find an integrated rather than a compromise resolution:

• What does each function actually need from this resolution, separately from what they have been asking for?
• Is there a solution that meets both sides’ underlying needs fully, rather than splitting the difference between their stated positions?
• What would each side be willing to give up if they were confident that what they actually need, rather than what they have asked for, would be protected?
• Have both sides articulated their real concerns, or are they still defending positions?
• What would a resolution look like that neither side would have suggested on their own but both could genuinely commit to?

Practical actions:

• Before proposing any resolution, ask both sides to articulate their underlying need separately from their stated position. Do this in separate conversations before any joint session, so each side can be honest without managing the other’s reaction.
• When a compromise has been tried and has broken down, treat the breakdown as evidence that the underlying need was not addressed, not that the other side is unreasonable.
• In joint resolution sessions, use the question “what would you need to be true for you to be able to commit to this fully?” rather than “can you accept this?” The first question opens up the possibility of integration. The second closes it into a binary.
• After a resolution has been reached, test it against both sides’ underlying needs, not just their stated positions. If either side is quietly accepting something that does not serve their real need, the resolution will not hold.

Rebuild the relationship alongside resolving the issue

A practical resolution and a repaired relationship are not the same thing. It is entirely possible to settle the functional dispute at the centre of a conflict while leaving the working relationship more damaged than it was before the resolution conversation, because the conversation itself was conducted in a way that generated new grievances rather than clearing old ones.

In a matrix, this matters more than it does in a hierarchy, because the relationship is the primary mechanism through which work gets done. A resolution that settles the immediate issue but leaves both functions less willing to engage with each other in future has traded a short-term win for a long-term cost.

Rebuilding a relationship in the context of a conflict resolution requires two things that most matrix leaders underinvest in. The first is acknowledgement: genuine recognition of the impact the conflict has had on each side, including what each function has done that made it harder. This is not the same as apology, and it does not require admitting fault. It requires naming what happened honestly and directly enough that both sides feel the reality of the situation has been acknowledged rather than managed. The second is a specific commitment to a different way of working going forward, not a general aspiration to collaborate better, but a named change in how the two functions will interact that both sides have agreed to and can be held to.

An operations director described a resolution conversation between his logistics team and the customer experience function that had gone well on the practical dimension and badly on the relational one. The specific issue had been resolved in about forty minutes. The remaining two hours had been spent relitigating a series of incidents from the previous six months in a way that left both sides feeling their history had not been properly acknowledged. He had moved too quickly to the practical resolution without giving the relationship dimension enough space. The logistics and customer experience leads left the room with an agreement they could both point to and a working relationship that was, if anything, colder than before they had met.

The programme lead made the same mistake in her first resolution attempt with sales and product development. She had focused entirely on the practical problem: the late requirements process, the roadmap stability issue, the escalation criteria. By the end of the session the practical matter was substantially resolved. Two weeks later, the two teams were barely speaking, because the conversation had never addressed the impact the previous months of conflict had had on the people involved. She convened a second conversation, this time structured around acknowledgement before problem solving, and the change in tone was immediate.

Questions you can use to rebuild a relationship alongside resolving a practical conflict:

• Has each side had a genuine opportunity to name the impact the conflict has had on them, separately from the practical question of how it gets resolved?
• Has each function acknowledged what it has done that made the situation harder, without that acknowledgement being used as a lever in the practical negotiation?
• What specific change in how the two functions interact has been agreed, as distinct from a general commitment to work better together?
• Is there anything from the history of this conflict that has not been named and that both sides know is in the room without being said?
• What would need to happen for both functions to describe this relationship, six months from now, as genuinely different from what it has been?

Practical actions:

• Structure resolution conversations in two parts: acknowledgement first, practical resolution second. Resist the temptation to move to the practical problem before the relational dimension has been given enough space.
• Ask each side, in the joint session, to name one thing that the other function has done well during a difficult period before naming what has made it hard. This is not forced positivity. It sets a norm of fair witness that makes the harder conversation more honest.
• End every resolution conversation with a specific, named commitment from each function about how they will work differently, not a general aspiration to collaborate. Write it down and treat it as a working agreement in the same way the contracting practice from the previous article describes.
• Schedule a follow-up conversation four to six weeks after any significant resolution, not to check whether the practical agreement is holding, but to ask how the relationship is feeling. The relationship dimension recovers more slowly than the practical one and needs its own attention.

Change the structural condition, not just the episode

The most important practice in this article, and the one most often skipped, is to identify and address the structural condition that produced the conflict rather than simply managing its current manifestation. This is the practice that determines whether a resolution lasts or whether the same conflict resurfaces in a different form six months later.

Most inter-departmental conflicts in a matrix are generated by one of three structural conditions: a boundary that was never clearly defined, an incentive system that puts two functions in competition, or a resource scarcity that means genuine cooperation requires one function to accept a worse outcome for itself. Each of these conditions sits above the level at which a matrix leader can typically act alone. Resolving them requires escalating the structural issue, clearly and specifically, to whoever has the authority to address it, which is a different action from escalating the conflict itself.

This is a distinction worth making explicit. Escalating a conflict to a senior leader is usually a sign that the matrix leader has run out of resolution options and is handing the problem upward. Escalating a structural condition to a senior leader is a different kind of action: it is a specific, evidence-based recommendation that something in the design of the organisation is generating a predictable conflict that will not be resolved by any amount of relationship management. The first looks like failure. The second looks like strategic thinking.

A group HR director I worked with described a conflict between the talent acquisition function and the business unit HR teams that had been managed episodically for four years by a series of matrix leads, each resolving the current episode without diagnosing the condition producing it. When she mapped the pattern across four years of incidents, the structural condition was clear: the talent acquisition function was resourced and incentivised to fill roles quickly, and the business unit HR teams were accountable for the quality of hires over a twelve-month period. Both sets of incentives were rational. Together, they were incompatible. She escalated a specific recommendation to the CHRO to restructure the performance targets for both functions to align on a shared metric. The recommendation was implemented. The conflict pattern, which had run for four years across multiple matrix leads, stopped.

The programme lead eventually reached the same conclusion about her sales and product development conflict. The integration she had found addressed the immediate symptoms well. But the underlying condition, a product development roadmap process that had no formal mechanism for incorporating market intelligence in real time, was still generating pressure on the boundary. She escalated a specific structural recommendation to the programme steering group: a quarterly market signal review built into the roadmap process, owned jointly by both functions, with agreed criteria for what qualified as a roadmap-level input. This was not something she could implement herself. It was something she could diagnose, evidence, and advocate for with enough precision that the people who could implement it understood exactly what needed to change and why.

Questions you can use to identify and escalate structural conditions:

• If this conflict were to resurface in six months with different people, what structural condition would be producing it?
• Is the conflict being generated by something that the two functions involved can resolve between themselves, or by something in the design of the organisation that sits above them?
• What specific change in how the organisation is designed, measured, or resourced would reduce the likelihood of this conflict recurring?
• Who has the authority to make that change, and what would they need to see to be persuaded it was necessary?
• Have you escalated the structural condition as a recommendation, with evidence, or have you only escalated the conflict itself as a problem to be managed?

Practical actions:

• After any significant conflict resolution, ask one further question: what would have to change in how this organisation is designed for this conflict not to recur? Write the answer down and treat it as a separate piece of work from the resolution itself.
• Build the structural diagnosis into your escalation. “This conflict is a symptom of an unclear boundary between these two functions, and here is what I think needs to change” is a more useful escalation than “these two functions cannot work together and need senior intervention.”
• Where a structural change requires a decision above your authority, make the recommendation as specifically as possible: what needs to change, why, what it would cost not to change it, and what a reasonable timeline for the change looks like. Vague escalations produce vague responses.
• Track conflict patterns over time, not just conflict episodes. A conflict that recurs in different forms across different people is almost always structural. The pattern is the evidence you need to make the structural case.

Wrapping up

Inter-departmental conflict in a matrix is not a sign of dysfunction. It is a predictable feature of an organisational design that gives multiple functions legitimate but sometimes competing claims on the same resources, decisions, and territory. The leaders who manage it well are not those who eliminate the conflict. They are those who can diagnose what kind of conflict they are facing, name it as legitimate rather than smoothing it over, separate the functional dynamic from the personal one, find integrations rather than settling for compromises, rebuild the relationship alongside resolving the practical issue, and change the structural condition rather than simply managing its current episode.

None of these practices requires formal authority over the functions involved. All of them require the willingness to engage with the conflict honestly rather than managing it at a safe distance. The most damaging thing a matrix leader can do with inter-departmental conflict is to treat it as something to be contained rather than resolved. Contained conflict is not managed conflict. It is deferred conflict, accumulating interest, waiting for the moment when containment fails.

Three questions for reflection

1. Think of a recurring conflict between two functions in your current organisation. Have you addressed the structural condition producing it, or have you only managed its most recent episode?
2. When you are in the middle of an inter-departmental conflict, are you looking for an integration that meets both sides’ underlying needs, or are you looking for a compromise that both sides can live with?
3. Is there a conflict in your current work that you have been smoothing over rather than naming? What is it costing you, and both functions, to leave it unaddressed?

Inspiration

Block, P. (1993) Stewardship: Choosing service over self-interest. San Francisco: Berrett-Koehler Publishers.

Davey, L. (2019) The good fight: Use productive conflict to get your team and organization back on track. Vancouver: Page Two.

Edmondson, A.C. (2018) The fearless organisation: Creating psychological safety in the workplace for learning, innovation, and growth. Hoboken, NJ: Wiley.

Fisher, R. and Ury, W. (1981) Getting to yes: Negotiating agreement without giving in. London: Penguin.

Follett, M.P. (1925) ‘Constructive conflict’, in Metcalf, H.C. and Urwick, L. (eds.) Dynamic administration: The collected papers of Mary Parker Follett. New York: Harper and Brothers, pp. 30–49.

Kahane, A. (2017) Collaborating with the enemy: How to work with people you don’t agree with or like or trust. Oakland, CA: Berrett-Koehler Publishers.

Tamm, J.W. and Luyet, R.J. (2004) Radical collaboration: Five essential skills to overcome defensiveness and build successful relationships. New York: HarperCollins.