Before a restructuring, assess what the new design asks of the organisation: which decisions move, who will own them, which handoffs are created, and what current ways of working will resist. Leave alone what the new strategy does not load. Many redesigns fall short on the connections between the boxes.
Where should the assessment start?
With the plan the restructuring serves. A new operating model is justified by what it lets the business do differently, so the first question is which decisions, handoffs and priorities have to change for that to happen. The structure is the answer to that question, and it can only be judged against it.
Many redesigns begin from the chart instead: fewer layers, clearer lines, a new matrix. The chart is easy to draw and to present, which is part of the problem. McKinsey's research on redesigns names an outsize focus on lines and boxes among the problems companies meet, and it is a predictable one when the design starts there.
Which decisions move in the new design, and who will own them?
List the decisions that change hands, and name an owner for each before the new structure goes live. A redesign that moves decision rights without saying so leaves two leaders who both believe they own the same call, or none who do. Either usually shows up as slower decisions.
Speed is a known weakness. In McKinsey's research on decision making, just 37% of respondents said their organisations' decisions were both high in quality and fast. A restructuring that adds a new interface for a decision can make it slower before it makes it better, and planning for that is cheaper than discovering it.
The useful artefact is short: the ten or so decisions that matter most to the plan, who decides each today, who will decide it after the change, and who has to be consulted. Where the answers disagree, the design is not finished.
Which new handoffs does the redesign create?
A reorganisation cuts some old dependencies and creates new ones. Map the handoffs the plan will rely on after the change, especially those between units that did not previously work together, and give each an owner. New interfaces tend to carry more execution risk, because nothing about them is yet routine.
A regional structure replaced by product lines is a typical case. Pricing, forecasting and key accounts now cross boundaries that used to sit inside one leader's remit. The design document rarely says who reconciles a disagreement between a product head and a country manager. Somebody will, informally, and the organisation will depend on that person without having chosen them.
What should be left alone?
The parts the new plan does not put under more demand. A team that works well and whose work the strategy does not change is a reserve, not a target. Redesigning it spends change capacity the organisation needs elsewhere, and adds strain without adding anything the plan requires.
Change capacity is finite. People absorb a restructuring on top of their current work, and each additional change competes for the same attention. Restricting the redesign to the places the plan loads is a practical way to protect the work that is already going well.
When should individual leaders be assessed?
After the design, the decision rights and the handoffs are settled, and against the roles as they will be. Assessing people first judges them against a job that no longer exists. Assessing them against the new demand shows who fits the new role, and where the role itself needs more support.
The order matters for fairness as much as for accuracy. A leader who struggles in the new design may be carrying a role that changed while its support did not. That is worth knowing before any decision about the person, and the fix is often in the role.
Where this stops
This note is about the organisational conditions a restructuring depends on. It does not assess whether the new operating model is the right strategic choice, what it will cost, or how the change should be communicated and consulted on. Where works councils or co-determination apply, those processes set their own sequence, and this note does not replace them.
How Atlas reads it
Atlas reads a new operating model against the plan it serves. The formal design is one layer. Real ownership, dependencies, operating patterns and what people report are the others, and the redesign is judged by how the five fit what the plan now demands. Areas where demand rises and present support is low are exposed. Areas the plan does not load are held in reserve.
The recommendation follows the intervention ladder, from clarifying a mandate to redesigning a role or forum, and it reaches the question of leadership last, when the demand and the evidence warrant it. The reading is an Execution Forecast before the change goes live and an Execution Reforecast once it has.
Sources
Questions
The decisions that will change hands and who will own them, the new handoffs the design creates, the priorities the organisation will carry during the change, and only then the leaders against the new roles. Starting from the plan the reorganisation serves keeps the assessment focused on what the new design has to deliver.
In McKinsey's 2014 survey, executives reported that three-quarters of redesigns fell short of their objectives and of improving performance. Common reasons include focusing on the chart instead of how work and decisions flow, and underestimating the strain on the organisation. Redesigns that settle decision rights and handoffs before launch give themselves a better starting point.
After. Assessing leaders against the old roles tells you little about the new ones. Once the design, decision rights and handoffs are settled, each leader can be read against what the new role demands, which also shows where a role needs more support rather than a different person.
Look at where the new plan increases demand. Teams whose work the plan does not change, and that are performing well, can usually be left as they are. Redesigning them spends change capacity that the organisation needs for the areas the new plan will load most heavily.
Longer than most plans assume. In McKinsey's research, nearly 30% of fully implemented redesigns took more than twelve months from the decision on objectives to full implementation. Planning for a period of slower decisions, and naming owners early, reduces the cost of that transition.