The deliberate process of configuring an organization's structure, roles, processes, and governance systems to align with its strategy and enable effective execution.
Key Takeaways
Organizational design is the deliberate process of configuring how an organization structures itself to execute strategy. It determines who reports to whom, how decisions get made, where authority sits, and how information flows between teams. Most people think organizational design means drawing boxes and lines on an org chart. That's the visible part, but it's roughly 20% of the actual work. The other 80% involves designing decision rights (who can approve what, and at what level), governance mechanisms (how do teams coordinate when work crosses boundaries), incentive alignment (do rewards encourage the behaviors the strategy needs), and information architecture (does the right data reach the right people at the right time). Jay Galbraith's Star Model, one of the most cited design frameworks, identifies five elements that must align: strategy, structure, processes, rewards, and people practices. Change one without adjusting the others and the design will fail. This is exactly why 70% of reorganizations don't meet their objectives. Companies move boxes around on the org chart but don't touch the underlying processes, decision rights, or incentive structures that actually drive behavior.
The most common triggers are strategic shifts (entering new markets, launching new product lines), mergers and acquisitions, growth that's outpaced the current structure, persistent performance problems, or a new CEO who wants to put their stamp on the organization. The worst reason to redesign, and the most common one, is to fix a people problem. If two VPs can't work together, redesigning the org chart to separate them treats the symptom while leaving the dysfunction intact. Bain & Company estimates that most large organizations go through some form of reorganization every 2-3 years. That frequency should raise questions about whether the redesigns are being done well.
Several frameworks guide design decisions. Each emphasizes different elements, but they share a common insight: structure alone isn't enough.
| Framework | Creator | Core Elements | Best For |
|---|---|---|---|
| Star Model | Jay Galbraith | Strategy, Structure, Processes, Rewards, People | Large-scale redesigns requiring multi-system alignment |
| McKinsey 7-S | Tom Peters & Robert Waterman | Strategy, Structure, Systems, Shared Values, Skills, Staff, Style | Diagnosing misalignment across organizational dimensions |
| Weisbord's Six-Box Model | Marvin Weisbord | Purpose, Structure, Relationships, Rewards, Helpful Mechanisms, Leadership | Quick organizational assessments and diagnostics |
| Nadler-Tushman Congruence Model | David Nadler & Michael Tushman | Input, Strategy, Transformation (work, people, structure, culture), Output | Analyzing how well organizational components fit together |
| Operating Model Canvas | Andrew Campbell | Value chain, processes, organization, locations, suppliers, information, management systems | Translating strategy into operational design decisions |
A disciplined design process reduces the risk of a failed reorganization. Rushing through these steps is the single biggest cause of design failures.
Before touching the org chart, answer one question: what does this organization need to be great at to win? If the strategy requires speed to market, the design must minimize handoffs and approval layers. If the strategy requires deep technical expertise, the design must create centers of excellence. If the strategy requires local responsiveness, decision authority must sit close to the customer. Amazon's two-pizza team structure exists because Jeff Bezos decided speed and autonomy mattered more than coordination efficiency. That's a strategic choice driving a design choice.
Choose the structural model that best serves the strategy: functional (grouped by expertise), divisional (grouped by product, geography, or customer), matrix (dual reporting), network (loosely coupled teams), or hybrid. There's no universally best structure. Each involves trade-offs. Functional structures build deep expertise but create silos. Divisional structures enable local responsiveness but duplicate resources. Matrix structures balance multiple priorities but create confusion about authority. The structure should make the most important work flow easily and accept friction in less critical areas.
This step matters more than the structure itself, and it's the one most companies skip. For every major decision type (pricing, hiring, product launches, budget allocation), document who decides, who provides input, who needs to be informed, and who can veto. The RACI matrix is a common tool, but RAPID (Recommend, Agree, Perform, Input, Decide) from Bain is more nuanced for complex organizations. Without clear decision rights, the new structure will recreate the same gridlock as the old one. Decisions will escalate to senior leaders by default, slowing everything down.
Redesign the processes that cross structural boundaries. Create lateral coordination mechanisms: shared metrics, cross-functional teams, integrator roles, or technology platforms that enable collaboration. Adjust the reward system to incentivize the behaviors the new design requires. If you want cross-selling between divisions but compensation is based on individual division P&L, don't expect collaboration. Finally, assess whether you have the right people in the right roles. New designs often require new capabilities.
The design is only as good as its implementation. Communicate the rationale clearly: why are we changing, what stays the same, what's different, and how will this affect me. Provide a transition plan with milestones. Expect a productivity dip of 6-18 months (Bain's research). Plan for it. Don't declare victory too early. Monitor leading indicators (decision speed, employee sentiment, collaboration metrics) throughout the transition and adjust the design where it's not working. The best designs evolve after launch based on real-world feedback.
Understanding failure patterns helps avoid them. These mistakes account for the majority of the 70% failure rate.
The most common mistake. Companies draw a new org chart, announce the changes, and expect different results. But if the meeting cadences, approval processes, performance metrics, and incentive structures stay the same, people will behave the same way in the new structure. Structure is necessary but not sufficient. Every structural change requires corresponding changes in processes, decision rights, and rewards.
When a company creates a role because they have a talented person they don't want to lose, or merges two departments because the leaders get along, they're designing around people. This approach produces structures that make sense for the current cast of characters but fall apart when anyone leaves. Design should start with strategy. Then find the right people for the roles the strategy requires.
Frequent reorganizations destroy organizational trust and create change fatigue. When employees have lived through three reorgs in four years, they stop investing in making the current structure work because they know another change is coming. A well-designed organization should last 3-5 years with minor adjustments. If you're reorganizing more often than that, the problem probably isn't the structure.
These cases show how design choices directly affect performance, for better and worse.
Spotify's widely copied model organizes around small, autonomous squads (cross-functional teams of 6-12 people), grouped into tribes (collections of squads working on related areas). Chapters (people with the same skill set across squads) and guilds (communities of interest) provide horizontal connections. The design optimized for speed and autonomy. Each squad could ship independently without waiting for other teams. But Spotify itself has acknowledged the model's downsides: coordination between squads is hard, and the lack of traditional management layers makes career progression unclear. Many companies tried to copy the model without understanding the cultural context that made it work at Spotify.
In 2019, P&G restructured from four industry-based sectors into six Sector Business Units (SBUs) with more direct P&L accountability. Each SBU controls its own brand portfolio, innovation pipeline, and go-to-market strategy. This reversed decades of matrix complexity where brand managers, regional presidents, and functional leaders all shared authority. The result: P&G's organic sales growth improved from 1% to 6% within two years. By giving SBU leaders clear accountability and the authority to match, P&G eliminated the decision gridlock that had slowed innovation.
Each model optimizes for different things. The right choice depends on your strategy, not on what's trendy.
| Model | Optimizes For | Biggest Trade-off | Works Best When |
|---|---|---|---|
| Functional | Deep expertise, efficiency, economies of scale | Cross-functional coordination is slow | Single product/market, technical excellence matters most |
| Divisional | Market responsiveness, accountability | Duplicated resources, less knowledge sharing | Multiple products, geographies, or customer segments |
| Matrix | Balancing multiple priorities (product + geography) | Complexity, dual-reporting confusion, slow decisions | Global companies needing both local adaptation and global scale |
| Network / Platform | Speed, innovation, flexibility | Less control, requires high trust | Fast-changing markets, knowledge-intensive work |
| Flat / Team-based | Speed, employee autonomy, reduced bureaucracy | Coordination overhead, unclear career paths | Smaller organizations or units requiring rapid innovation |
Data on how organizations approach design and the results they achieve.