Matrix structures were created to coordinate increasingly complex organisations. However, some of the difficulties now attributed to them arise when the mechanisms created to coordinate begin to multiply without being reviewed.
—Who can approve this change?
—It depends.
—On what?
—On the areas affected.
—And who decides which areas need to be involved?
The conversation often stops there. Not because the organisation is poorly designed or because no one wants to take responsibility. In many complex organisations, the answer genuinely depends on the type of decision, the context and its impact on other areas.
What becomes interesting happens a few seconds later. The conversation stops being about the change itself and starts being about how to coordinate it. Another validation appears, a new forum, another stakeholder or an exception. Each measure responds to a legitimate need.
This is where a pattern begins that goes unnoticed in many organisations.
Coordination ends up creating new coordination needs because the mechanisms designed to solve specific problems remain in place long after they are no longer necessary.
Complexity rarely appears the day the matrix is created
Matrix structures emerged when reality could no longer fit within a single hierarchical line.
An international company needs to coordinate markets with corporate functions. A bank must integrate business, risk, technology and regulation. An insurer needs to combine a customer perspective with highly specialised capabilities. In all these cases, the matrix responds to a legitimate need: coordinating different perspectives around the same decision.
The question, therefore, is not whether matrix structures work. Complex organisations need cross-functional mechanisms to coordinate people, capabilities and priorities. The difficulty arises when every point of friction leaves behind a permanent mechanism created to solve a temporary problem.
Research into matrix structures has been describing this phenomenon from different perspectives for years. MIT Sloan and Harvard Business Review attribute much of the difficulty to a lack of clarity over who decides what, how disagreements are resolved and when a decision should be escalated.
Deloitte adds a relevant finding: organisations with greater organisational design maturity are up to six times more capable of adapting their structure when the context changes.
That ability depends less on reducing the number of dimensions than on periodically reviewing the design and the elements that support it.
From the outside, two organisations may look very similar. The difference is often found in everything they have gradually built around that structure to keep it working.
When the system starts reproducing itself
The central question is this:
When does coordination start producing more coordination?
The pattern usually develops gradually. An exception requires a specific criterion. That criterion leads to a forum. The forum brings in participants who were not previously involved. When a similar situation arises, that route already exists and eventually becomes part of the standard process.
Over time, the organisation also starts managing the mechanisms it created to coordinate itself.
The literature on matrix management describes product launch, resource allocation and portfolio prioritisation decisions in which between four and seven stakeholders may be involved before a resolution is reached.
The problem is rarely the number of participants itself, but the lack of clarity over who decides, who influences and when a disagreement needs to be resolved.
That is why organisations with very similar formal structures can make decisions at completely different speeds.
In many cases, the design explains only part of the problem. The other part lies in everything the organisation has accumulated around it while responding to new organisational needs.
The cost rarely appears where organisations look for it
When decisions slow down, the explanation is often sought in excessive meetings, bureaucracy, lack of alignment or resistance to change. All of these can be genuine symptoms, but they do not fully explain why similar decisions now require more stakeholders and more validation than they did a few years ago.
The cost becomes significant when the organisation normalises increasingly long decision paths. Part of that cost is distributed across additional meetings, rework, successive validations and alignment cycles that rarely appear as a separate line item, even though they affect the speed at which the organisation executes.
The pattern takes different forms depending on the sector. In telecommunications, it can delay deployments or new services. In energy and industry, it can extend decisions on critical investments. In manufacturing, it can make it harder to respond quickly to changes in the supply chain. And in the public sector, it can lengthen approval processes for cross-functional projects.
Each addition usually responds to a legitimate need. What is less common is reviewing whether the mechanisms created to address that need still provide clarity or simply remain because no one has questioned them again.
Coordination also needs an expiry date
Matrix structures will continue to be part of many organisations. Specialisation, regulation and the need to integrate different capabilities make completely linear models difficult to imagine.
The conversation, therefore, may not be about eliminating these mechanisms.
It is about a different question.
Who reviews coordination?
Perhaps that review should be part of the same cycle already used to review processes, technology or risk.
Recent organisational redesign programmes reveal an interesting pattern. They do not simply focus on simplifying organisational charts. They periodically review the mechanisms the organisation itself has introduced to coordinate: decision rights, validation forums, escalation paths, committees and approval processes. Some remain. Others disappear because they no longer address the problem they were created to solve.
At Cognodata, we often encounter this point when an organisation is already examining why its decisions are slow. The conversation rarely begins by questioning these mechanisms. It usually reaches them after processes, technology or indicators have been reviewed without providing a sufficient explanation for the lack of speed.
Coordination tends to be designed as though it were permanent.
The organisation is not.
Few organisations would accept maintaining a technology system, process or risk policy indefinitely without reviewing it. Coordination mechanisms are more often allowed to persist without the same scrutiny.
Most coordination mechanisms are created to solve a specific need. Yet few organisations establish from the outset when they will revisit whether that mechanism still makes sense.
The question may not be how many dimensions a matrix has or how many committees are involved in a decision.
It may be something else.
Which coordination mechanisms in your organisation still provide clarity, and which remain simply because no one has reviewed them again?
Because complexity does not always increase when a new dimension appears.
The ability to coordinate is also an organisational asset. And, like any other asset, it stops adding value when it stops being reviewed.
Frequently asked questions about matrix structures and organisational coordination
What is a matrix organisational structure?
A matrix organisational structure combines different lines of responsibility within the same organisation. It allows functional, geographical, business and specialist perspectives to be integrated when a single hierarchy is not sufficient to coordinate complex decisions.
Why can a matrix structure slow down decision-making?
The problem does not necessarily lie in the matrix itself. Decisions slow down when decision rights are unclear, when it is not clear when disagreements should be escalated, or when coordination mechanisms remain in place after they have stopped being necessary.
How can an organisation tell if it has too many coordination mechanisms?
One sign is that similar decisions progressively require more validations, stakeholders, forums or escalation steps without those additions providing greater clarity to the decision.
How can organisations reduce the complexity of a matrix structure?
Not necessarily by eliminating reporting lines or committees. The first step is to review coordination mechanisms periodically, identify the problem each one was created to solve and determine whether it still adds value in the current context.
