How you distribute work across locations, what you centralize or decentralize, and which activities you keep in-house versus source externally are not just operational questions. They are core elements of organizational design. They shape cost, risk, talent access, collaboration patterns, and the lived experience of teams.
This chapter focuses on making those footprint choices deliberately and integrating them into your org design rather than treating them as separate “real estate” or “outsourcing” projects.
14.1 Centralization vs. Decentralization of Activities
Centralization and decentralization are often debated as matters of philosophy (“we’re a decentralized company”) when they should be treated as design choices made activity by activity.
Centralization means consolidating similar activities into a single unit or a small number of hubs. Decentralization means placing activities in multiple business units, regions, or local teams. Neither is inherently better; the question is where each approach creates the most value.
A practical way to think about it is to assess each major activity along four dimensions:
- Scale and efficiency
Activities with high standardization potential and strong economies of scale tend to benefit from centralization. Examples:- Payroll and core HR administration
- Transactional finance (AP/AR, general accounting)
- Commodity IT infrastructure and basic helpdesk
Centralization here can reduce cost, improve consistency, and enable automation.
- Business intimacy and responsiveness
Activities that require deep local or segment insight, rapid context-specific decisions, or close customer interaction are candidates for decentralization. Examples:- Front-line sales and account management
- Local marketing activation
- Country-level regulatory engagement in markets with highly specific rules
- Risk and control
Some activities are high-risk if fragmented: policy-setting, risk appetite definition, key compliance standards. These often require central ownership even if execution is distributed. Examples:- Group risk and compliance frameworks
- Information security standards
- Brand and communications guidelines
- Innovation and learning
Innovation can sit centrally (platforms, core tech, global product strategy) or locally (experiments, pilots, local adaptations). Centralization supports reuse and scaling of innovations; decentralization supports variety and market fit. Often you will:- Centralize platform and architecture decisions
- Decentralize experimentation within agreed guardrails
Instead of a blanket policy, create a centralize–decentralize matrix: list key activities and decide for each:
- Central-only (design and execution at center)
- Central design / local execution
- Local with central standards
- Local-only
The key is to be explicit. Many organizations drift into hybrid arrangements by accident: activities are nominally centralized but heavily duplicated locally “just to be safe,” or they are decentralized but still constrained by heavy central approvals. The result is cost without speed, or speed without control.
A short test for your centralization choices:
- Can local leaders clearly explain which levers they own and which the center owns?
- Are there activities that are effectively done twice (central and local) with little incremental value?
- Where decisions must be made fast, do they sit close enough to the front line?
- Where mistakes would be very costly, is oversight strong enough?
If you cannot answer these questions confidently, you likely need to revisit how you have allocated activities across the footprint.
14.2 Shared Services, Centers of Excellence, and Outsourcing
Once you have decided what to centralize, you need to choose how to centralize: in shared services, centers of excellence (CoEs), or through outsourcing and partners. Each has a distinct role in the organizational design.
Shared services
Shared services centralize high-volume, standardized, transactional (or semi-transactional) work into one or a few delivery centers. They serve multiple business units and functions under service-level agreements.
Typical domains:
- Finance operations (AP/AR, record-to-report)
- HR operations (hire-to-retire processes)
- Procurement operations
- Customer service for standard requests
Design considerations:
- Define a clear service catalog: what the shared service does and does not do.
- Set SLAs and performance metrics agreed with business “customers.”
- Decide the governance model: who sets priorities, who funds investments, how disputes are resolved.
- Choose locations based on cost, language, and talent availability; often a mix of onshore, nearshore, and offshore.
Done well, shared services free business units from running small, inefficient back offices and allow them to focus on customers and strategy.
Centers of excellence (CoEs)
CoEs host deep expertise and set standards for specialized areas. They are less about transaction volume and more about quality, innovation, and consistency.
Typical domains:
- Data and analytics
- Cybersecurity
- Advanced manufacturing engineering
- Pricing and revenue management
- ESG, safety, or other specialist topics
Design considerations:
- CoEs should have a clear mandate: set standards, develop methods, provide expert support, and build capability.
- They often operate in a hybrid model: some experts centralized; others embedded in business units but belonging to the CoE community.
- Governance must balance enterprise priorities (platforms, methods) with local needs (customization, proximity).
CoEs can easily become ivory towers if they only issue guidelines without engaging in real work. Anchor them in meaningful delivery projects so they stay connected to reality.
Outsourcing and managed services
Outsourcing shifts activities to external providers under contracts. This is attractive for:
- Highly standardized, non-differentiating tasks (e.g., payroll processing, some IT operations, certain customer-service volumes).
- Capabilities where external providers can invest more in tools and talent than you can justify.
- Flexible capacity requirements (seasonal peaks, 24/7 coverage).
Design implications:
- Outsourcing does not remove the need for internal ownership. Someone must define requirements, manage the vendor, monitor performance and risk, and own the relationship.
- You must decide how you will retain critical knowledge and how easy it would be to switch providers or bring work back in-house if needed.
- Regulatory and data-privacy constraints may limit what can be outsourced or where data can be processed.
Some organizations adopt a layered model:
- Shared services or CoEs in-house for differentiating or sensitive capabilities.
- Outsourcing for commodity or overflow work.
- Strategic partnerships or joint ventures for specialized, high-value domains.
The key is coherence: you should be able to explain which work is done where and why, and how the service ecosystem fits your organizational design and risk appetite.
14.3 Remote, Hybrid, and On-Site Work Patterns
Org design decisions about teams and roles now inherently include decisions about where work is done: remote, hybrid, or on-site. These are not just HR policies; they affect collaboration patterns, leadership requirements, and location strategy.
Consider work patterns by activity and role type, not only by function:
- Work that can be done fully remotely: individual, knowledge-based tasks with limited dependence on physical assets (e.g., many software, analytics, documentation, some design and planning roles).
- Work that requires regular in-person interaction: complex collaboration, physical prototyping, sensitive discussions, early-stage ideation where face-to-face contributes meaningfully.
- Work that is inherently on-site: manufacturing, logistics, in-person service delivery, physical inspections.
For each category, define default expectations:
- Fully remote possible (with occasional travel).
- Hybrid with specified in-office days or events.
- On-site as the norm, with limited flexibility.
From an organizational-design perspective, remote and hybrid choices affect:
- Team design and leadership
- Distributed teams require stronger clarity on roles, goals, and norms.
- Managers must be able to lead without relying on visual supervision; coaching and performance discussions need deliberate cadence.
- Collaboration and interfaces
- Cross-location collaboration depends more on tool choices and disciplined meeting practices.
- You may cluster certain activities in the same time zones to reduce coordination friction.
- Talent and location strategy
- Permitting remote work expands talent pools beyond traditional hubs.
- However, fully atomized remote hiring can create micro-sites with no critical mass, complicating management and culture.
To integrate work-pattern decisions into org design:
- Design teams with clear expectations: is this a co-located team, a hybrid team with anchor days, or a fully distributed team?
- Ensure line of sight: every remote employee should know which site or hub they are anchored to (for leadership, development, and community), even if they rarely go there physically.
- Codify ways of working: default tools, meeting rhythms, core hours for cross-time-zone work, and explicit rules about responsiveness and availability.
Remote/hybrid choices should also be consistent across similar roles to avoid perceived inequities. If two teams doing similar work have radically different flexibility, morale and retention will suffer unless there is a clear, articulated reason.
14.4 Location Strategy: Cost, Talent Pools, Risk, and Regulation
Beyond centralization choices and work patterns, you need a deliberate location strategy: where key units and activities are physically based.
A structured approach considers four main dimensions:
- Cost
- Labor cost differentials across countries and cities.
- Real-estate and facility costs.
- Tax and incentive regimes.
Cost alone should not drive decisions, but it is a material factor for large-scale operations.
- Talent pools
- Availability of relevant skills (e.g., engineering, analytics, customer service in specific languages).
- Competition for those skills and local attrition norms.
- Attractiveness of the location to internal and external talent (quality of life, mobility, visas).
- Risk and resilience
- Political stability, security, and infrastructure reliability.
- Natural-disaster risk, climate risk.
- Concentration risk: how much of a critical capability is you willing to place in a single city or country?
- Regulation and proximity
- Requirements for local presence (e.g., regulators, government clients) and local data storage.
- Proximity to customers, partners, and ecosystems (suppliers, universities, innovation hubs).
Based on these dimensions, organizations often adopt a hub-and-spoke or multi-hub model:
- Global hubs for strategic functions and CoEs, chosen for connectivity, talent, and brand.
- Regional hubs for shared services, regional management, and certain sales and operations activities.
- Local nodes where on-the-ground presence is required (plants, warehouses, sales offices, key-account presence).
When designing your footprint:
- Start from where critical roles and capabilities need to be to best support the strategy.
- Decide how many hubs you need in each category (global, regional, local) and what each hub’s mandate is.
- Avoid uncontrolled proliferation of small sites; they create complexity and dilute leadership attention.
Location strategy must also be dynamic. As markets, costs, and politics evolve, you should periodically review:
- Whether certain hubs have outlived their purpose.
- Whether new talent pools or regulatory changes create opportunities or risks.
- Whether you are over-exposed to a single country or region in a critical function.
From an org-design lens, these changes must be reflected in your structure: reporting lines, unit definitions, leadership roles, and governance.
14.5 Integrating Footprint Decisions into the Org Design
Footprint and sourcing choices should be woven into the org design, not made in parallel. Integration happens at three levels: structure, roles, and transitions.
1. Structure
Ensure that your structural choices reflect footprint decisions:
- If you create shared services or CoEs, they should appear as clearly defined units in the org structure with explicit mandates and reporting lines.
- If you designate global or regional hubs, make sure there are corresponding leadership roles (e.g., site leads, hub heads) with clear responsibilities for local operations, culture, and risk.
- If you shift from local to global or regional ownership of certain activities, update P&L and cost-center structures accordingly.
Ask:
- Can we see, on the org chart, where each major site or hub sits and who is accountable for it?
- Do we have any “orphan” locations with no clear leadership or accountability?
- Does the structure reflect how work actually flows between locations (e.g., from nearshore center to onshore teams)?
2. Roles and interfaces
Location decisions have direct implications for roles:
- Site or hub leaders: define their mandate (operations, people, external representation, risk, continuity). They are often critical in maintaining culture and engagement in multi-location organizations.
- Global process owners and CoE leads: clarify how they work with local teams, which decisions they own, and how they support or challenge local practices.
- Vendor-management roles: if outsourcing or managed services are significant, design roles that own those relationships and outcomes, not just procurement.
Interfaces must be redesigned with the footprint in mind:
- Document core handoffs between onshore and offshore or nearshore teams.
- Align time zones and working hours where possible to minimize friction for high-collaboration tasks.
- Clarify escalation paths across locations and vendors, not just within a single site.
3. Transitions
Changing footprint is one of the most disruptive elements of org design. Moving work from one site to another, creating or closing locations, or shifting from in-house to outsourced models affects people, processes, and customers.
A practical transition approach:
- Stage the moves: pilot in one or two areas before scaling; avoid moving everything at once.
- Define transition states: interim structures and roles while work is being transferred; avoid long periods where no one knows who is in charge.
- Protect customers and operations: identify business-critical periods and avoid major transitions then; run parallel operations and extended testing where risk is high.
- Integrate with people and HR plans: redeployment, reskilling, and exit programs must be synchronized with footprint steps (see Chapter 13).
Finally, build feedback loops:
- Track performance and risk measures for new hubs, shared services, or outsourced relationships.
- Listen to employees and customers in affected locations about service quality and collaboration.
- Be prepared to adjust quickly—adding local capacity, changing governance, or even reversing decisions if assumptions prove wrong.
When footprint decisions are made explicitly, grounded in strategy and integrated into structure, roles, and processes, they become a powerful lever of organizational performance rather than a source of ongoing tension. Your aim is an organization where what is done where and by whom is clear, coherent, and aligned with how you create value and manage risk.
