Organisation Design and Operating Model
Before You Redesign
Executive Summary
When performance slips, the first place most senior teams look is the org chart. The boxes are familiar. The boundaries are easy to redraw. A reorganisation produces an announcement, a deck, a town hall, and the comforting feeling that something decisive has been done. What it does not usually produce is a fix. Most performance problems are not structural. They are operational—lodged in how work moves through the organisation, how decisions close, how teams coordinate, what the metrics quietly reward. They are operating model problems, and most of them can be resolved without changing a single reporting line.
None of which makes organisation design irrelevant. Sometimes the structure is exactly what needs to move. Sometimes both the structure and the operating model need to move—when technology is changing the work itself, or when the strategy requires capabilities the current shape of the organisation cannot develop. The point is to diagnose before you commit. Restructuring is expensive, disruptive, and slow to reverse, and the question worth asking is not whether to redesign. The useful question is whether the evidence points to the structure, the operating model, or both—and whether the diagnostic work has actually been done.
What follows is mBolden’s framework for making that call, and for executing on the intervention the diagnosis points to. It draws on advisory work across more than 80 organisations in 14 countries, including one case in which a long-running cross-functional conflict was resolved entirely through operating model changes—without a single reporting line moving.
The Restructuring Default and Why It Persists
Restructuring is the most visible move a senior leader can make. It produces a deck, a town hall, a new chart, and an announcement the board can point to. The arc is familiar, the milestones are legible, the narrative is practised. And in most cases, it is the wrong place to start.
The evidence is uncomfortable. McKinsey research finds that only about 23 percent of reorganisations are judged successful by the companies that run them; the remainder stall partway or fail to lift performance at all.1 The default persists for two reasons, optics and training. Most senior leaders were taught to diagnose organisational problems through a structural lens. When something is not working, the reflex is to ask who owns it, where it should sit, who should report to whom. Those are sometimes the right questions. Asked before the more fundamental one—how does work actually move through this organisation, and where is it breaking down—they end up addressing the symptom rather than the cause.
Restructuring is also more expensive than most balance sheets capture. The direct costs are easy enough to enumerate: severance, recruitment, the productivity dip during transition. The indirect costs run larger and surface later. Months of senior leadership bandwidth absorbed by the redesign. Teams destabilised at the moment the strategy needs them performing at their best. Talent that opts out rather than absorb another reset. Institutional knowledge lost in the handovers. And, six to twelve months on, the recognition that the performance problem is still there—because the structure was never what produced it. None of which makes restructuring wrong. It makes it too costly, in time, talent, and organisational trust, to use as a first response to problems that may not be structural to begin with.
01
The Operating Model: What It Is and Why It Matters
Between strategy and structure sits the operating model. It is the system that determines how the organisation actually delivers on what the strategy commits it to. Not how the chart says work should move, but how it does: which decisions close at which level, where work passes between teams, what is measured, what those measures reward. Most organisations document the structure carefully and leave the operating model largely undocumented. The chart is straightforward to draw and communicate. The operating model evolved organically, varies by function, and has rarely been examined with anything like the rigour brought to strategy or structure.
Its invisibility is precisely what makes it powerful. The operating model shapes behaviour more reliably than almost any other organisational variable, because it determines what is easy and what is hard inside the firm. Which decisions close in days and which drift for weeks. Whether cross-functional work produces coordination or conflict. Whether the metrics on the dashboard are pulling toward the same outcome the strategy is pursuing. The chart describes who reports to whom. The operating model describes how the work actually moves. Most performance problems live in the second.
A well-designed operating model has five interconnected components, each capable of producing performance problems on its own and each available to be redesigned alone or alongside the others. Governance and decision rights: who owns which decisions, and how quickly those decisions close. Work design and process: how work moves between people and teams, where friction accumulates. Metrics and incentives: what is measured, and whether the measures across functions point in the same direction. Capabilities and roles: what skills the work demands, and what is best done by humans, by machines, or by the two together. Rhythms and communication: how often the right people are in the room, and whether the meetings produce decisions or simply circulate updates. Structure—who reports to whom—is the sixth. It matters, but it is one input into a system that has five others. Change the structure without attending to the rest, and the result is the same performance in new boxes.
02
When the Problem Is the Operating Model, Not the Structure
The clearest signal that an operating model problem is being treated as a structural one is repetition. When the same friction, the same decision delays, and the same cross-functional disputes resurface after a restructure, the structure was never what produced them. Other indicators are sharper but more specific. Decisions stall at the boundaries between teams rather than within them, which is almost always a metrics or decision-rights problem rather than a structural one. Two functions stay in continuing conflict despite leadership endorsing the same strategy, usually because their scorecards measure success differently. And cross-functional initiatives underdeliver even when each individual team has performed well, which says something about how the work was coordinated, not about who was assigned to it.
The most diagnostic indicator of all is what the best people do. When the most effective operators in an organisation succeed by routing around the formal system rather than working through it, the formal system is the problem. And the formal system, in that case, is the operating model—the architecture of decisions, metrics, processes, and rhythms that governs daily work. Not the org chart.
When Structure Is the Right Answer
None of this implies that restructuring is wrong. It implies that restructuring should be earned—supported by evidence that the structure, rather than the operating model, is producing the failure. The conditions in which structural change is the right answer are reasonably clear. When spans of control are distorted to the point that leaders are accountable for outcomes they cannot influence, or where oversight has become impossible, structural redesign addresses a constraint that operating model changes cannot reach. When the strategy requires capabilities the current organisation does not have—a new business model, an unfamiliar customer segment, a delivery approach the firm has never built—the question of whether the structure can develop or attract those capabilities becomes structural by definition. Operating model redesign improves how existing capabilities are deployed. It does not create capabilities the organisation does not have.
When technology is altering the nature of work, the situation becomes structural in a different way. The advance of AI and automation is shifting which tasks belong to humans, which belong to machines, and which depend on the two operating together. The result is a real structural question about which roles and teams need to be redesigned around new workflows. The organisations building durable advantage are not simply deploying AI tools. They are redesigning how the work is done in light of what the technology now makes possible—which means redesigning roles, teams, and structures alongside the processes and metrics that govern them. In most cases this is not an either-or decision but a both, with the operating model clarified first so that the structure can be built around how the work is intended to flow. Designing the structure before the decision rights, the metrics, and the work design have been settled is building the house before deciding what it needs to contain. This is a sequencing principle, not an iron rule. Occasionally a structural misalignment is severe enough that the operating model work cannot proceed until the structure has been addressed.
A Case in Point: When Metrics Created the Conflict
The Metrics Were Incompatible
Sales was measured on availability, distribution on flow efficiency, and there was no shared measure that reflected what the business actually wanted: product on shelves without supply chain operations breaking under the load. Each team was rationally optimising for its own scorecard, and those scorecards were in structural conflict with each other. The friction was not evidence that the wrong people held the roles. It was evidence that the incentive system was pointing two capable teams in opposite directions. The fix was a single shared metric, reflecting the actual business objective rather than each function’s historical measure of success. With one number to decide from, every conversation stopped being about defending scorecards and started being about solving the underlying problem. The question shifted from who is right to what the shared number is telling the business to do.
Visibility Was Fragmented
Each team operated from its own data inside its own systems. Sales had availability figures, distribution had flow figures, and neither had a complete picture of the reality the other was managing. The fix was a shared dashboard, giving both teams open, real-time access to the other’s operating reality. After that, there was no longer a sales version of the truth competing with a distribution version. There was one version, visible to everyone, which removed the basis for most of the disagreement and supplied the shared context that joint decision-making depends on.
Collaboration Was Unstructured
The two teams had no regular forum for joint decisions. Cross-functional issues, which surfaced constantly, were either resolved informally through individual relationships or pushed up to a level of leadership further from the work. The fix was a recurring cross-functional sync designed to produce decisions rather than circulate updates—the same people, the shared data, an explicit mandate to settle the matter in the room. The results registered quickly. Friction dropped. Decisions that had taken weeks began to close in days. The daily strain of navigating conflict largely dissipated, and the business benefited from two capable teams now pulling in the same direction. The reporting lines were unchanged. The conflict had been an operating model problem from the start. The pattern recurs across industries. The most damaging cross-functional conflicts are rarely personality conflicts or power struggles. They are incentive conflicts. When metrics push two functions in opposite directions, capable people behave in ways that look like resistance but are entirely rational responses to the system they have been asked to perform inside. Fixing the system is faster and more durable than trying to fix the people.
A Framework for the Decision
The choice between operating model work, structural work, or both is not a matter of preference. It is a diagnostic conclusion, reached through five disciplined steps. The first is to locate where performance is actually breaking down—not where the organisation thinks the problem sits, but where decisions are stalling, conflicts are emerging, and outcomes are diverging from intent. The richest data comes from the people closest to the work. The workarounds they have built and the informal escalation paths they rely on are more diagnostic than any process map. The second step is to test whether the failure is structural or operational. Structural failures concern spans, levels, accountability misalignment, or capability gaps the current structure cannot close. Operational failures concern metrics, decision rights, process design, information flow, and collaboration rhythms. Most organisations have both, in different proportions, and the relative severity determines what gets addressed first.
The third step is to assess capability and capacity for what is ahead: what the organisation needs to be able to do over the next three to five years, whether the current structure and operating model support building that, and—candidly—which roles are becoming less central as technology takes on more of the underlying work. The fourth is to sequence the interventions by what the evidence shows rather than by convention. Where the evidence points to operating model failures, address those first. They are faster to implement, cheaper to reverse if the analysis was wrong, and often eliminate the perceived need for structural change. Where both are required, sequence the operating model work first or in parallel with structural change so the new structure is built around a clear picture of how work will move. The fifth step is to design with the human-machine boundary explicit: which tasks are candidates for automation, which roles will change in scope, and which new capabilities need to be built into an organisation that does not yet have them.
The mBolden Perspective
01
Operating Model Design
The architecture of governance and decision rights, the design of work and process, the alignment of metrics and incentives, the cadence of collaboration, and the boundary between human and machine work.
02
Organisation Design
Design. Spans and layers, role definition, capability and capacity assessment, sizing the structure for the strategy ahead, and sequencing structural change inside broader transformation.
03
Workforce Planning and Capability Build
Aligning roles, skills, and capacity to the demand the organisation will face, including what technology now enables and what it changes about the nature of human contribution.
04
Change Enablement
Removing the structural and behavioural barriers that cause well-designed operating model and structural changes to stall once implementation begins
Conclusion
Whether to redesign is not the useful question. Most organisations navigating digital transformation, talent shortages, regulatory pressure, and competitive disruption will need to change both how they are organised and how they operate, and they know it. The useful question is which change, in which sequence, on what evidence. Most skip the diagnostic step and move from a performance problem to a restructuring announcement, because the restructure is visible, familiar, and reads as decisive. The cost of that shortcut—leadership bandwidth lost for months, talent attrition during the transition, and the slow erosion of organisational trust that follows—is substantial. And the original problem reappears, usually inside twelve months, because the source of it was never addressed.
The operating model is where most performance problems actually live. It is where cross-functional conflict originates, where decisions stall, where metrics create the wrong incentives, and where the gap between what the strategy requires and what daily work produces becomes permanent. Fixing it is less disruptive than restructuring, and the results are more durable, because the intervention reaches the system that governs behaviour rather than the boxes that describe it. Design around how the work needs to move, and let the structure follow the work rather than precede it. Begin with the diagnostic, allow the evidence to point to the right intervention, then sequence the work with a clear view of what the organisation needs to be able to do over the years ahead—not just the quarter ahead. That is the difference between redesign that delivers and redesign that has to be done again.
Frequently Asked Questions
What Is the Difference Between Organisation Structure and Operating Model?
Structure is who reports to whom: the boxes and reporting lines on the chart. The operating model is how work actually moves through the organisation—how decisions are made, how teams coordinate, what is measured, and what the measures push people to do. Structure is one component of the operating model, not a substitute for it, and most performance problems live in the operating model rather than the chart.
Should We Restructure to Fix a Performance Problem?
How Do You Know If the Problem Is the Operating Model, Not the Structure?
When Is Restructuring the Right Answer?
What Does an Operating Model Include?
About the Author
Suzanne Knight, MBA, MA is the founder and CEO of mBolden, a global management consulting firm. She has advised more than 80 organisations across 14 countries on transformation, organisation design, and execution, with engagements spanning Fortune 500 companies, financial institutions, health systems, and federal and provincial governments.
A former Fortune 100 transformation executive, she now keynotes internationally on digital transformation, leadership, and leading through change, and has been a long-standing instructor at the Schulich School of Business Executive Education. The dual vantage point—senior corporate executive and global advisor—means she has lived through the consequences of the decisions she now helps others make.
Her work rests on a single thesis: transformation succeeds or fails on execution, and no quality of strategy can rescue ineffective delivery.
