Macro-structure is where you make the big calls: what your top-level units are, who holds P&L, and how you slice the enterprise into coherent chunks of accountability. Once you set it, everything else—roles, governance, talent flows—tends to follow. That makes macro-structure one of the highest-leverage and highest-risk decisions in any org design. The goal in this chapter is to give you a practical way to select the primary grouping logic, construct a small set of options, evaluate trade-offs, and define the roles of business units and the corporate center, including the balance between global control and local autonomy.
8.1 Choosing the Primary Basis of Grouping (Product, Geography, Segment, Function)
The first macro-structural decision is deceptively simple: “What is the primary logic by which we group the organization?” In practice, this means deciding which dimension—product, customer segment, geography, function, or value stream—will drive the top-level units and P&Ls. Most organizations end up with some hybrid, but you still need to pick a primary logic. If you do not, you drift into a de facto matrix without ever deciding where accountability truly sits.
Start with your strategy and value agenda. Ask: where do you most need focused, end-to-end accountability to win? If distinct product families have very different economics, innovation cycles, and competitive sets, product-based business units may be right. If different customer segments buy in very different ways and you win primarily through relationship and solution tailoring, segment-based or client-industry-based units often make more sense. If regulation, logistics, or local market structure dominate, geography may need to be primary. Where economies of scale in specialist work are paramount and offerings are relatively stable, a strongly functional logic can be effective, especially in early-stage or highly standardized businesses.
You rarely get to optimize all dimensions equally. Picking a primary grouping logic is explicitly choosing which dimension will experience the least friction and which will carry more of the coordination burden. For example, if you choose product as primary, customer segments and regions must live with some fragmentation and rely more heavily on cross-unit processes and governance. If you choose geography as primary, your global product strategy and innovation processes must be designed to work across many local P&Ls.
A practical sequence to choose the primary basis of grouping:
- List your key strategic priorities and value drivers for the next 3–5 years.
- For each candidate dimension (product, segment, geography, function), ask:
- If we made this the primary grouping, which priorities would benefit most?
- Which priorities would be made harder and require more coordination?
- Rank the dimensions by “must be frictionless” vs “can tolerate coordination.”
- Choose the primary basis of grouping accordingly, and be explicit about which dimensions you are not optimizing structurally.
You may still need secondary dimensions reflected in your design—global product councils, segment overlays, or regional clusters—but you should treat these as overlays on a primary structure, not equal pillars in an unspoken matrix. When people know which logic is primary, they know whose P&L they ultimately serve and who will make the final call when interests collide.
8.2 Evaluating Structural Options: Pros, Cons, and Trade-Offs
Once you have a primary grouping logic, you will almost always have multiple viable ways to implement it. For example, within a product-led model you might have three large global business units, or six more focused ones, or a two-tier model with global product lines and regional execution units. The design challenge is not to generate endless variants, but to define a small, coherent set of options and evaluate them transparently.
Start by setting a discipline: generate two to four serious options, plus the current state as a reference. More than that is confusing; fewer than that often means you are not exploring real trade-offs. Each option should be described with enough specificity to be testable: top-level units, P&L locations, how major functions and regions connect, and where key capabilities sit.
Evaluate options using a consistent set of criteria, tied to your design objectives and principles. Typical criteria include:
- Strategic fit: How well does the option align with the strategy and value agenda? Does each top-level unit have a clear strategic mandate?
- Customer impact: Does it improve or degrade clarity of ownership for key customers and segments? How many hand-offs does it introduce across the customer journey?
- Economic impact: What are the implications for structural cost (spans, layers, duplication), and where might it unlock growth or productivity?
- Decision speed and clarity: Does it sharpen or blur decision rights for critical choices (pricing, investment, product priorities, risk decisions)?
- Feasibility and risk: How complex is the transition? What regulatory, labor, or operational risks does it introduce?
- Talent and leadership: Do you have (or can you develop) leaders who fit the key roles? Does it create realistic, attractive leadership jobs?
You do not need a complex scoring model. A simple “high / medium / low” or red-amber-green assessment against each criterion is enough, as long as it is done consistently across options. The important part is the conversation: why, for example, Option A is better for strategic fit but worse for cost than Option B, and which trade-off the enterprise is willing to accept.
One discipline that dramatically improves the quality of evaluation is to separate structure from people in the early stages. Design roles and units as if you were hiring from a blank global market. Only once you have agreed on the best structure for the business should you discuss who might fill which role. If you design around current individuals too early, you will optimize for comfort and continuity, not for strategic fit.
Finally, be explicit about what each option sacrifices. Every design creates losers and winners along certain dimensions. Write this down for each option: “We gain X and Y, we give up some degree of Z.” When the top team chooses, they are choosing that pattern of gains and losses, not just a picture of boxes and lines. This clarity makes later communication more credible: you can explain not only what you chose, but also what you consciously did not prioritize.
8.3 Designing Business Units, Divisions, and Corporate Center Roles
With a preferred macro-structure logic and option selected, you need to define what your business units, divisions, and corporate center actually do. Many organizations get into trouble because they draw BUs and a corporate center, but never truly define their mandates, resulting in constant friction and duplication.
Think of a business unit as a mini-enterprise with a coherent strategy, a set of customers or products, and enough scale and scope to justify a dedicated leadership team. Good candidates for BU status typically:
- Have distinct economics (margins, capital intensity, growth) versus other parts of the business.
- Face different competitors or regulatory regimes.
- Require tailored strategies, capabilities, and talent profiles.
- Are large enough in value (revenue, profit, or strategic importance) to justify dedicated leadership.
Within BUs, you may create divisions or sub-units where there is a second meaningful clustering—by region, product line, or segment—but they usually do not carry full strategic autonomy. Be careful not to proliferate divisions solely because of historical leadership preferences; each division should have a clear, distinct mandate and performance logic.
The corporate center often requires the most deliberate design. Left unchecked, corporate centers tend to grow opportunistically, accumulating activities and roles that no one else wants. A helpful way to define the center is to group its roles into four categories:
- Direction and portfolio: Strategy, capital allocation, portfolio management, enterprise brand, investor relations.
- Control and assurance: Risk, compliance, internal audit, ethics, key policies, and standards.
- Shared services and platforms: Finance operations, HR operations, IT infrastructure, procurement, data and digital platforms.
- Enterprise capabilities and stewardship: Talent management, leadership development, functional excellence, enterprise-wide programs (e.g., sustainability, safety, transformation).
For each category, ask three questions:
- What must be done at the center for reasons of scale, risk, or enterprise coherence?
- What should not be done at the center because it constrains BUs without adding value?
- Where is the center’s role to set direction and guardrails, versus to execute?
The answers should then be translated into clear charters for both the corporate center and the BUs. A BU charter clarifies its P&L scope, key responsibilities, decision authorities, and required interactions with the center. A corporate center charter clarifies what it owns outright, what it co-decides with BUs, and where it is a service provider versus a regulator.
As a rule of thumb, the corporate center should be as small and high-value as possible, consistent with risk and strategy. If an activity is neither about enterprise direction, control, platforms, nor critical capabilities, it is a candidate to push down to BUs or out to shared services. Your macro-structure design is significantly more likely to succeed if everyone understands whether the center is a light-touch “value-adding parent,” a strong operator, or something in between—and why.
8.4 Global vs. Local: Control, Autonomy, and Standardization
For any organization operating across multiple geographies, the question of global vs. local is central to macro-structure. Even in a domestic business, the same logic applies between central functions and distributed units. The core tension is always the same: where to centralize for scale and control, and where to devolve for customer relevance and speed.
Instead of treating this as a philosophical debate, treat it as a design problem by domain. Different activities warrant different answers. For each major area—strategy, brand, pricing, product, operations, people, risk—ask four questions:
- Where does scale or consistency create clear value?
- Where does local knowledge and proximity create clear value?
- What are the risks of fragmentation if we decentralize, and can they be mitigated?
- What are the costs, in responsiveness and ownership, if we centralize?
For example, brand positioning might best sit centrally to ensure a single global promise, while marketing execution is localized to adapt to cultural and channel differences. Product architecture and platforms may be global, but feature prioritization and packaging localized. Risk appetite and policy may be set centrally, while risk decisions within that appetite are taken as close as possible to the front line.
You can formalize this as a set of mandates for global and local roles. A global function or BU might have the mandate to “set standards, design platforms, and approve deviations,” while local units “execute within standards, provide market insight, and innovate within defined boundaries.” The same applies to global product vs regional sales: who owns list prices, who can approve discounts, who decides which products to launch where, and who funds local adaptations?
A helpful discipline is to avoid vague “shared ownership.” For any given topic, define:
- What is global-only (decided centrally, executed everywhere).
- What is local-only (decided and executed locally).
- What is global design / local adapt (central designs the frame; local chooses within or customizes with approval).
- What is local innovate / global scale (local experiments; central scales successful patterns).
Macro-structure choices should then reinforce these mandates. If you want genuinely strong regional autonomy, you may need regional P&Ls with substantial functional resources embedded. If you want tighter global control in a highly regulated business, you may design more direct reporting into global functions and limit local discretion.
It is also important to recognize that global-local balance can differ by BU or function. A commodity manufacturing BU may centralize most decisions; a solutions or consulting BU may operate with more local autonomy. Technology or data platforms may be global even in largely local businesses. The point is not uniformity; it is coherence and clarity.
8.5 Macro-Structure Design Review Checklist
Before you lock in your macro-structure, you should subject the preferred design to a disciplined review. This is where you test whether the structure truly reflects your strategy, design principles, and guardrails—not just what feels familiar or politically acceptable.
Use this checklist as a structured walk-through with your top team:
Strategic alignment and focus
- Can we clearly state the strategic mandate of each top-level unit in one or two sentences?
- Can we map each major strategic priority to a specific accountable unit or combination of units, with one clear “owner of last resort”?
- Does the structure make it easier, not harder, to deliver the 3–5 key value drivers we identified?
Customer and market impact
- For each priority customer segment, can we name a single leader accountable for its performance and experience?
- Does the structure reduce unnecessary hand-offs along critical customer journeys?
- Are there any important customers or segments that fall between units or are split across too many owners?
Economic and operational implications
- Does the design create obvious duplication of similar activities that could be avoided without harming performance?
- Is the number of top-level units and P&Ls manageable, given the leadership capacity and governance load?
- Do spans and layers implied by the macro-structure look realistic, or are we baking in excessive layering or heroic spans at the outset?
Governance, risk, and control
- Does the location of risk, compliance, and key control roles satisfy regulatory and internal policy requirements?
- Are decision rights for critical enterprise decisions (pricing, capital allocation, major investments, risk appetite) clear from the structure?
- Is the corporate center’s role in direction, control, platforms, and capabilities well-defined and bounded?
Global–local balance
- For each major function and BU, is it clear which decisions and activities are global vs local vs shared?
- Does the structure support the level of local responsiveness the strategy requires?
- Are we relying on an unspoken matrix, or have we intentionally chosen where to centralize and where to decentralize?
Talent and feasibility
- Does the new structure create a set of top roles that are realistic, attractive, and not overloaded?
- Do we have, or can we develop, the leadership bench to staff these roles within a reasonable timeframe?
- Is the transition from current state to target state feasible within our financial, legal, and change capacity guardrails?
If you cannot answer “yes” to most of these questions, or if you uncover contradictions (for example, a BU with a broad strategic mandate but limited control over key levers), you are not done with macro-structure. Go back, refine the option, or explore a variant. It is far cheaper to iterate on the whiteboard than to discover misalignments during implementation.
Once the macro-structure passes this review and the top team is aligned, you can move with confidence into micro-structure, roles, and decision rights. The heavy lifting of “who owns what” at the top of the house will be behind you, and the rest of the design will be about making that logic work in practice.
