Organizational Design goes beyond physical workspaces. How a company manages its people, assets, and resources also has a significant impact on how a company creates value. As we illustrated in Chapter 2, whether an organization optimizes for resources or flow (whether consciously or not) will have wide-ranging effects throughout the company.
That’s not to say that there is any one “right” way to structure a company. Depending on the business context in which the company operates, one given structure may be more advantageous than another. As we’ve learned, knowledge work requires different structures than what we may need in environments where we’re looking to decrease variance, lower per-unit costs, and utilize resources to the fullest. Some specific organizational patterns are more appropriate than others when it comes to embracing uncertainty, executing with purpose, and creating an environment characterized by robust employee communication and collaboration.
In this section, we’ll first describe some of the most common organizational structures companies deploy and point out some of the benefits and challenges associated with each of them. We’ll then highlight a few nascent structures that have gained traction over the past few years.
We’ll conclude the chapter by presenting an example of an agile Organizational Design that is currently gaining a great deal of momentum, and for good reason: it’s currently working for HERE, Spotify, and ING.
The functional structure is the most common organizational structure in use today, and it has its roots in the management theories we discussed in Chapter 2. As the name implies, a functional organizational structure is optimized for each function within the company; for example, below the top tiers of administration you might have a Production department, a Marketing department, an IT department, and so on, as pictured in Figure 4.4.
Figure 4.4 Example of a Functional Organizational Structure
Each function in this type of organization is run by a functional leader who is responsible for managing her respective staff and ensuring her function runs optimally. Given that the focus of each function is to specialize in a given area of expertise, the people within each function tend to share similar skills, knowledge and abilities.
Because of this structure, there is a rather large difference between the various functions in the organization: having people change roles from marketing to information technology happens rarely, for example, because each function has its own culture. In extreme cases, it may even be viewed as an organization inside an organization.
Functional structures optimize for resources and reduce per-unit costs. They increase efficiencies and facilitate a high degree of control.
Also, staff within each functional area has a clear career path. As employees gain experience in their field, they advance from junior to more senior roles within their specialty. This means the level of knowledge within the particular function is typically quite deep: the more senior employees can have 30+ years of experience.
Although the functional structure facilitates deep areas of knowledge within the company, it also creates distinct silos and makes collaboration between specialty areas difficult. Having this clear delineation and separation has a tendency to slow down decision making and makes adapting to changing market conditions challenging.
People working in companies with functional structures also typically struggle to see the “big picture” of how an organization creates value for its customers. Because their work rarely involves people outside of their functions, employees tend to be limited to a customer view dominated by a functional perspective. For instance, if you happen to work in the Quality department and focus purely on this perspective, you may only notice the number of defects and “bugs” reported by customers, while the more substantial issues preventing higher uptake may be related to poor user design or responsiveness.
A functional organizational structure, while popular, is becoming increasingly undesirable in an environment characterized by Volatility, Uncertainty, Complexity, and Ambiguity (VUCA). Due to the inherent limits of this structure as it relates to adaptability, speed of execution, and customer focus, companies that benefit from this structure are those that operate in stable, highly predictable environments. Government institutions, manufacturing organizations, and religious institutions are examples of organizations that may leverage this structure with some success, although these areas are no longer as stable as before.
A divisional structure is used more often in larger companies that span a wide geographic area or support multiple lines of business and products under a common corporate umbrella, as illustrated in Figure 4.5. In a divisional organizational structure, each division operates semi-autonomously to produce a service or product. Each is headed by its own executive responsible for running the division as a separate business. For example, a bank leveraging an organizational structure may have separate divisions for retail, wealth management, investment banking, and so on.3
Figure 4.5 Example of a Divisional Organizational Structure
The head of the division, typically a VP or general manager, is responsible for hiring, overall budgeting, and sales/marketing for the respective division. Having a clear line of ownership is advantageous because all resources necessary to produce value for this particular product or service are focused on a common goal.
Contrary to a functional organization, where there is constant competition for access to central resources, the divisional structure have its own functions and therefore supports a more product-centric way of running a business.
In my experience, companies with a divisional structure typically move faster than functional organizations. They can change strategic direction more rapidly and with a higher degree of customer focus.
Divisional structures can be powerful, but they have their distinct disadvantages. Communication between the divisions often suffers as each unit is run more or less as its own organization within the company.
This structure can also be expensive because there is duplication of functions inside each division. For example, just as the Retail division will have a marketing function, so will the Investment Banking division. This duplication of effort can add a lot of overhead.
Perhaps more crucially, the divisional structure tends to create unhealthy competition and negative politics between the various divisions. Because each division competes for the same resources from the corporate level, there is little incentive to collaborate between the divisional heads, and this tends to drive a short-term, narrow view of the business as a whole.
For larger companies with operations in multiple locations across the world, a divisional structure may make sense. Having a division in Asia run semi-autonomously from a sister division in Europe can serve the company well because the needs of the customer base in each geography may be different and require custom ways of operating.
Although more suited to deal with today’s complex business environment than the functional structure, one potential drawback of this way of organizing is that without strong leadership and alignment at the top, the semiautonomous nature of this structure can lead to divisional factions and misalignment on a global level. For instance, a customer’s experience of one brand in the U.S. might be very different with the same brand in Europe, if the same brand is operating independently. This can create negative brand perception and customer confusion.
A matrix structure can be viewed as a combination of the functional and divisional structures. Whereas the functional structure optimizes for a given function and a divisional structure optimizes for a particular product, geography, or line of service, a matrix structure aims to marshal a company’s resources and assets toward a common goal, as illustrated in Figure 4.6.
Figure 4.6 Example of a Matrix Organizational Structure
One of the key benefits of a matrix organization is that it drives an increased level of collaboration and communication across the organization. People are not tied to a given function or a product line; instead, people work across boundaries to accomplish the same goal, regardless of where they may reside in the org chart.
Also, because employees are working in an environment where managers are not defining the way things are done, there tends to be a relatively high level of autonomy and employee engagement in matrix structures. Consulting firms are often set up using a matrix structure where employees may have a formal “people manager” who is separate and distinct from the “client engagement manager” who may direct their day-to-day work activities.
The main drawback of the matrix organizational structure is that employees are supervised by multiple managers at once. Because employees are managed by both a “people manager” (who may be responsible for their career path and growth) and a “product manager” (who defines the goals they are working on), employees in matrix structures often report feeling conflicted and confused regarding which priorities to follow. “Do I work on improving a certain process, or do I spend my efforts working on a product deliverable?” Both are important—the former concerns improving how we work while the latter addresses customer needs—but one of these needs to be prioritized higher than the other. Both can’t be equally important.
Having a single manager would help in terms of gaining clarity. With two managers—each coming from different perspectives—you may be faced with delivering two work items, both being number one on the priority list.
Also, because employees are typically working on multiple goals at once (each with its own set of managers and other stakeholders), the matrix structure brings with it significant complexity and overhead, requiring excessive coordination and dependency management. I worked in a matrix organization earlier in my career and experienced some of the challenges firsthand. For instance, my “people manager,” who was ultimately responsible for defining my financial rewards and career recognition, would sometimes disagree with my “product manager,” whose goals I was assigned to help achieve. As an employee, these situations could be uncomfortable; do I follow my people manager, who wants me to help redesign our existing workflow so we can be more effective in the future, or do I follow my product manager, who wants me to focus more on immediate feature deliverables? When management was well aligned, this type of structure worked well. When it was not, it tended to create conflicts and office politics.
The matrix structure may at first appear as if it combines the best of the functional and divisional structures. However, when used in larger organizations, many of its inherent benefits disappear amidst complexity and coordination between conflicting managers and goals.
For smaller organizations (< 150 people), this structure may work quite well, but as the organization moves beyond a few hundred people, the matrix structure tends to be less effective—both for executing in an efficient manner and adapting quickly to changing business environments.
Emergent Organizational Structures: Sociocracy and Holacracy
The organizational structures discussed in the preceding sections are well established and have been in place in some form or another since the concept of the modern corporation was introduced in the 19th century. The business environment has changed dramatically since then, however, and some organizational structures have emerged that aim to better support a more human-centric, adaptable organization. Two of the most popular emergent structures are Sociocracy and Holacracy. In the following section, we’ll describe these structures and highlight some of their differences. We’ll also highlight some of the challenges associated with each.
The term Sociocracy comes from the Latin and Greek words socius (companion) and kratein (to govern). Sociocracy aims to put people first and considers all employees to be equals. In contrast to the more traditional organizational structures we discussed earlier where there is a single decider at the top, decision-making in Sociocracy is based on consensus—the notion that a group of individuals can get behind a decision together without objections.
Gerard Endenburg, an electrical engineer, is credited with defining the modern version of Sociocracy we know today. Combining ideas from the Quakers’ consensus principles with his understanding of engineering and systems thinking, Endenburg defined an “operating system” of decision making and collaboration through a set of hierarchical circles—each corresponding to a department of an organization.
In this structure, policy decisions require consent of all members of the circle. Regular decision making is made by the operations leader within the policies established in circle meetings. If other circles are affected by a decision, then a circle of representatives from each affected circle is empowered to make the decision. By linking circles and making decisions by consent, efficiencies are preserved while maintaining equality for the circles and the people in them (see Figure 4.7).4
Figure 4.7 An Example of a Sociocracy Organizational Structure. Representatives from Various Circles are Empowered to Influence Decisions that Affect Them. Decisions are Made by Consensus.
Although influenced by Sociocracy, Holacracy is a more recent innovation and is based on Brian Robertsen’s organizational experiments at Ternary Software, where he wanted to explore more democratic forms of governance. Robertsen first defined Holacracy as a concept in 2007 and later codified it with definitions of practices and principles in his 2010 manifesto, “Holacracy Constitution.”5
A central idea in Holacracy is the focus on roles. A role is not a job description in the traditional sense; it is a definition of the purpose, span of control, and accountabilities involved in getting work done. This means a person in an organization can take on multiple roles; he is not solely dedicated to one purpose, as in a Sociocracy.
Holacracy also leverages the circle-based decision model, but it is more self-organizing and less self-directed than in Sociocracy. In Holacracy, decision making is more hierarchical; circles are connected by two roles known as “lead link” and “rep link,” which are accountable for alignment with the broader organization’s mission and strategy. If there is a potential conflict between circles, the “lead link” gathers input and feeds this into existing roles so forward momentum can be made. The roles are empowered to do everything possible (within legal and ethical limits) to achieve the goals defined for them. Advocates of Holocracy therefore claim a bias toward action and autonomy—more so than with Sociocracy. A person whose role is “Community Liaison” is empowered to make decisions within her area of responsibility without conferring with others in her circle, for instance. Sociocracy would encourage a more consensus-based approach.6
One of the key benefits of Sociocracy is that everyone in the organization feels included, and their voices are represented in decision making. This structure also helps create trust among the employees and a strong sense of working toward a common purpose.
It is perhaps not a coincidence that Sociocracy has been adopted quite successfully in nonprofit, religious, and cohousing communities—organizations that thrive on empathy and common understanding among constituents.
Holacracy shares many of the same benefits and places a high degree of emphasis on personal accountability and actions. Compared to Sociocracy, in Holacracy, there is increased focus on defined roles and their ability to make decisions, and there is less focus on group consent. Yet, by emphasizing self-organization, an organization leveraging Holacracy should be more adaptable to change and more apt for a world where environments are fluctuating widely.
With its focus on equality and transparency, Sociocracy will appear quite radical to businesses accustomed to operating with top-down decision making and need-to-know information-sharing policies. The corporate culture required to make Sociocracy work can also be alien to multinational companies accustomed to driving strategic direction from headquarters rather than having decision making happen organically within a group of employees viewed as equals. Most implementations of Sociocracy known today are of small-to-medium organizations; larger enterprises have yet to adapt this model.
Holacracy has had a few high-profile adaptations, notably Zappos and Medium; however, both implementations have been challenging, to say the least. Medium ended up abandoning Holacracy after a few years because of challenges related to cumbersome coordination for larger product development efforts. In addition, the rigorous and detailed nature by which the roles and policies were defined ended up hindering personal ownership and initiative.
Zappos’ Holacracy efforts are still ongoing, but the company has generated some high-profile negative attention as a double-digit percentage of people left the company when CEO Tony Hsieh posed an ultimatum asking employees to commit to Holacracy or leave. It is too early to tell whether or not Zappos’ Holocracy effort can be called a success or a failure, but it is clear that it has been an extremely challenging transition.7
Sociocracy and Holacracy are genuine innovations in Organizational Design. Although it’s difficult to implement, there are examples of smaller organizations that have been able to adapt this way of operating with success. The Morning Star Company, a maker of tomato products, has been leveraging Holacracy for some time, as has Valve, a developer of software gaming platforms. These are relatively small companies, however.
For larger corporations, these governance structures still appear too immature and risky to constitute a pragmatic alternative to one of the traditional Organizational Designs discussed earlier. Although there is a lot to like—self-organization, consensus-building, transparency—the cohesive cultural environment, leadership maturity, and communication structures necessary to make these designs work in large enterprises may not be there yet.8
An Agile Organizational Structure?
After exploring these organizational structures at a high level, where does this leave us? The formal structures we discussed earlier are rooted in traditional, plan-driven ways of running a business; they optimize for resource control and compliance rather than speed and agility. And the modern structures we looked at optimize for self-organization and personal fulfillment but do not confidently address customer needs and a changing business environment.
What, then, is an Organizational Design that supports business agility?
The short answer is that there is no one “correct” Organizational Design. Rather, an agile Organizational Design is dynamic, flexible, and ultimately optimizes for customer value (present and future) rather than resources or leadership control. In the next section, I outline the Organizational Design I was involved in creating at HERE and cover the highlights of an organizational setup at ING (greatly inspired by Spotify), a Dutch financial institution that has been making some innovative organizational changes.
Optimizing for Value: Organizational Design at HERE and ING
When HERE (formerly NAVTEQ) started its agile transformation in 2009, the initial focus was on implementing Scrum (and in some cases Kanban) to boost teams’ ability to deliver software faster, with higher quality, and with more transparency. Over the course of several months, the Agile Working Group (AWG) and the coaching staff trained teams, empowered Scrum Masters, and gave Product Owners the authority to own their respective Backlogs to prioritize work. By all measures, the initial efforts were a significant success: quality improved significantly, teams delivered value faster, and management had more visibility into what was delivered over time.
But everything was not all roses. Part of the reason the teams were doing so well was that they were doing quite poorly before due to a bogged-down, highly bureaucratic process that had been imposed on the teams a few years earlier. The relative improvement was significant, but still not enough to be competitive with more nimble competitors entering the marketplace.
Perhaps most troubling: The teams were starting to operate more as silos, making it difficult not only to integrate code as teams prepared to go to production, but also to appear to the customer as a cohesive end-to-end product. For instance, one team would be working on a radically different way to process map data throughout the vast database of Points Of Interest (POI), yet the teams responsible for collecting POIs in the first place had no way to migrate their information to the new process engine—the data structures, the type of data involved, and the overall process were not aligned. The teams were operating as individual units, not as part of a cohesive effort aligned around a common goal.
The coaching team recognized that the root cause of the problem wasn’t so much that Product Owners and ScrumMasters weren’t coordinating enough—they had daily Scrum-of-Scrums and Communities of Practice in an effort to do just that—but that the teams needed a more appropriate Organizational Design. Rather than having teams operate under the existing organizational structure, working for a given divisional lead, the coaching team tried organizing the teams under a common goal instead. For instance, for Next Generation Map Building (NGMB), teams involved in these efforts—regardless of where they may initially belong on the org chart—came together to solve that problem.
The new structure was entirely aligned around value creation; people organized around a common mission that was led by a Chief Product Owner, typically a senior director or a VP. Teams formed more or less organically to help deliver on this goal—and upper management decided the relative size of the program based on the priority of the effort. Mission-critical efforts received more oxygen than initiatives that were not deemed as strategic.
The results were dramatic: teams became more aligned around shared goals, miscommunications and misunderstandings decreased, and customers received a more cohesive product experience. The change was probably more challenging for middle management than anyone else. Reporting relationships was no longer the primary source of power and influence; rather, the ability to deliver customer value became the most important factor managers were held accountable to.
Agility at ING
Another similar example took place in the Netherlands. The Dutch banking group ING was doing great financially, but leadership understood that the business environment around them was changing fast.
“Customer behavior...was rapidly changing in response to new digital distribution channels, and customer expectations were being shaped by digital leaders in other industries, not just banking,” Bart Schlatmann, CIO of ING states in McKinsey Quarterly.9
Leadership realized that ING was no longer competing against other banks—it was being compared to the performance of other, more nimble technology companies. To learn more, ING executives decided to visit tech companies like Google and Spotify to better understand what made these companies so adaptive and responsive to customer expectations. One of their epiphanies was the way the companies organized around value. Schlatmann continues:
“The key has been adhering to the “end-to-end principle” and working in multidisciplinary teams, or squads, that comprise a mix of marketing specialists, product and commercial specialists, user-experience designers, data analysts, and IT engineers—all focused on solving the client’s needs and united by a common definition of success.”
ING was particularly inspired by Spotify, the Swedish streaming service that described its own organizational model in 2012 in a paper by agile consultant Henrik Kniberg and Spotify internal coach Anders Ivarsson.10 After visiting Spotify and admiring its organizational structure in action, ING decided to organize itself through a number of Tribes (to use Spotify’s terminology), each responsible for an overarching business objective. Each Tribe consists of several Squads—small groups of people (no more than nine) dedicated to working on a given end-to-end solution. Each Tribe is led by a senior executive who is responsible for strategic direction and financials. Each Squad’s direction is then owned by a Product Owner. To facilitate cross-team collaboration and alignment on technical capabilities, each Squad has a Chapter Lead who meets with other members of the chapter on a regular basis to keep everything in synch (see Figure 4.8). The Chapter Lead is also the people manager for the persons in the Chapter. A Chapter Lead should have people in her chapter spread across various squads; this prevents the conflicts sometimes seen in matrix organizations. The PO is responsible for what happens (prioritization) and the Chapter Lead is responsible for a standard way of work for the experts within her particular expertise area (the Chapter).
Figure 4.8 ING’s Tribe Organization; Inspired by Spotify
There are quite a few similarities between this and the model described earlier with HERE. Both models share characteristics we noted earlier, with formal and informal structures. For example, this model shares elements of self-organization with both Holacracy and Sociocracy. Embedded in these models are elements of hierarchy as well: although the “how” is left up to the teams (with informal controls provided by the Chapters), the “what” in the form of business objectives is clearly defined from executive levels.
These structures are aiming to strike a careful balance between flexibility and reliability. By allowing people to have more autonomy within clearly defined boundaries of responsibility, Spotify, HERE, and ING are able to respond faster to market changes while delivering on customer commitments predictably. Although none of these companies claims to have achieved a “perfect Organizational Design”—they are always evolving and improving—these structures have proven to be more effective ways of managing work in environments characterized by rapid change. One indication that this way of working is bearing fruit: while the initial efforts at ING described above were limited to the organization in The Netherlands, senior leadership was so impressed by the results that the entire company is moving in this direction, implementing a similar structure in offices across the company worldwide.11