Clean Sheet Organization Design: The Process

Why the current chart is the wrong starting point

Every organization chart is a record of decisions that were correct at the time they were made. Roles were created for a customer that mattered, a plant that existed, a founder who carried three functions at once. None of that is a criticism. It becomes a problem only when the strategy moves and the structure does not, at which point the chart quietly governs the enterprise in place of the strategy.

When a leadership team redesigns from the existing chart, the conversation is about people, titles and reporting lines within thirty minutes. That conversation is unavoidable, but it belongs at the end of the process, not the beginning. Held first, it produces a settlement between incumbents rather than a design, and the same constraints reappear a year later with new names on them.

A clean sheet inverts the order. The team designs the organization the strategy requires, on an empty page, with no names in the boxes. Only then does it place talent, count cost and sequence the transition. The discipline is not in the drawing; it is in refusing to look at the current structure until the design decisions are made.

An organization chart is an answer. A clean sheet forces the leadership team to restate the question first.

The pre-condition: a stated '80'

A clean sheet is not the first move. It requires a stated view of where economic value is concentrated — the segments, customers, products and channels that carry the profit — and an equally explicit statement of what will be minimized. Compass calls this the vital few: the small share of the business that generates most of the economic value, set against the long tail that consumes disproportionate complexity.

Without that statement, a clean sheet becomes an opinion contest. With it, the design has a test: does this structure put the strongest talent, the shortest decision path and the most capacity against the vital few, and does it deliberately starve the complexity that does not pay?

This is also why the work stays a small-team exercise at the start. A core team of the chief executive, the finance leader and the people leader can hold a design honestly. A committee of twenty cannot, because every function present will defend its own boxes before the design exists.

The five decisions that constitute the design

First, design principles. Before any structure is drawn, the team writes down what it is optimizing for — speed to the customer, cost position, technical depth, geographic reach — and what it will accept as a consequence. A design with no stated trade-off has not been designed.

Second, current-state assessment. Structure, roles and capabilities are mapped against the strategy, not against each other. The output is a short list of gaps and strengths, including the capabilities the strategy requires that the enterprise does not presently have.

Third, future-state options. Serious clean sheets produce two or three genuinely different structures — for example a central functional model, a segment-led model and a hybrid — each costed and each assessed for speed, capability and risk. A single option is a proposal, not a design process.

Fourth, roles, decision rights and capability definition. Each load-bearing role receives a small number of accountabilities, one clear decision it owns, and the capability profile required to hold it. Where two roles claim the same decision, the design is not finished.

Fifth, talent evaluation and workforce planning. Only now are names placed against roles, with an honest read of strong fit, close fit and gap, and a build-or-buy decision for every gap. Held to the end, this conversation is a placement exercise. Held first, it is the design.

Design test

The white space question

After the boxes are drawn, list the outcomes the enterprise is managed on and ask which box owns each one. Anything that lands between boxes is white space, and white space is where strategies fail quietly. Assign every item to a single name before the design is approved.

The absent-executive test

Walk the design as though the chief executive were unavailable for a month. If cross-functional trade-offs cannot be settled without that person, the design has left the integration mechanism inside a human being and will not scale.

Transition is part of the design, not a phase after it

A clean sheet that is approved and then handed to human resources to implement usually degrades within a quarter. Sequencing, communication and the order in which roles are filled are design decisions with consequences: fill the integrating roles first and the rest of the structure holds; fill them last and the organization runs on its old wiring while the new chart hangs on the wall.

Transition therefore carries its own artifacts — a phasing plan with dates, a named owner per move, a communication plan written for four audiences (board, executive, manager, frontline), and a risk register covering both talent and customer exposure during the change.

Then the design is governed rather than filed. Compass runs this through ARC — Assess, Refine, Commit — on a quarterly organization and talent review, with an annual strategic reset. Effectiveness is measured, structure and talent are adjusted, and the improvements are institutionalized. A design that is never re-examined becomes the next chart nobody wants to question.

The design is not finished when it is approved. It is finished when it is installed, measured and owned.

Who owns what

Clean sheet work fails on ownership more often than on analysis. The executive sponsor is the chief executive, and that cannot be delegated: the design decides where authority sits, which only the top of the house can settle. The process owner is the head of people or organization development, who runs the method, holds the artifacts and keeps the sequence honest.

Operating owners are the profit-and-loss leaders, people and strategy leaders who supply the content and carry the consequences. Governance is a standing quarterly organization and talent review plus an annual strategic reset, so the structure is revisited on a cadence rather than in a crisis.

Compass partners with the leadership team through this work and then transfers it. The team that designed the structure runs the review that maintains it — which is the point of the engagement.

Exhibit — SIPOC: Organization Transition & Design Process

The clean sheet as an operable process rather than a workshop. Read across: who supplies the work, what it consumes, the nine steps, what it produces, and who receives it. Each row is a coherent slice of the same process.

Ownership: executive sponsor — CEO. Process owner — CHRO or head of organization development. Operating owners — P&L leaders, HR / OD, strategy. Governance — quarterly organization and talent review plus an annual strategic reset.

How this shows up in the business

• The reorganization conversation starts with names and titles rather than with the strategy.
• Only one future-state option is ever presented, so no trade-off is visible.
• Outcomes the enterprise is measured on have no single owning role.
• Capabilities the strategy requires are assumed to exist because the roles exist.
• The strongest talent is concentrated on the long tail rather than the vital few.
• Cross-functional trade-offs still require the chief executive to be present.
• The new chart was published, but no quarterly review examines whether it works.

A clean sheet is not an act of demolition. It is the discipline of designing the organization the strategy requires, naming who decides what, and installing it with the same seriousness the strategy received — so the structure stops being the reason the strategy underperforms.

A designed structure only produces results once adoption — not announcement — is treated as the outcome.