Twelve years ago, one of us was sitting with a senior executive at a telecommunications equipment manufacturer. The firm had been a high flier with demand driven by the global explosion in wireless networks and the Internet bubble. But demand had dropped rapidly, and this firm, along with most of its peers, faced rapid contraction. The executive looked out at a stalled turnaround effort and wondered why none of her plans to address the issue had been implemented.
“I told everyone what they had to do,” she said. “Why didn’t they do it?”
Essentially, she was asking how leaders get followers to follow. No matter how well constructed a leader’s plans, they count for naught if they are not enacted.
Consider the experience of a global supply chain and logistics company. A few years ago this organization needed to deploy operational enhancements quickly to meet competitive challenges. The problem this posed to the organization’s leadership can be put simply: How do you get large numbers of people to act in a coordinated, sustainable fashion while leading a global company that has grown through acquisitions that have not been fully integrated?
This simple question masks a tall order for leadership. As a global company, its workforce was geographically dispersed and highly diverse. What exists, other than gut instinct, to inform leaders’ decisions around organizational engagement?
For several years, we have been studying this issue: first by analyzing research data on organizational success and failure, then through research carried out in cooperation with global companies facing this challenge. Our project has sought answers to that telecommunications executive’s implied next question: What should I have done as the leader of this organization?
In focusing on strategy execution and large-scale changes that depend on large numbers of people working together,1 we have identified three factors that are present when organizations achieve their goals. At least one of these factors is missing when organizations in the study struggled. Further, we have observed that as leaders learn to use analytically verified models that shape effective decisions, they tend to attain greater degrees of success in coordinating the efforts of their workforce toward a defined goal.
The As One project began as a research effort, focused on studying collaboration in organizations. The project started with three fundamental assumptions:
Collaboration among large groups of people is a requirement for meeting many organizational challenges.
- There are multiple paths to organizational success based on the nature of the organization and on its goals. There is no one “right” type of organization.
- The project findings should be based on rigorous data analytics. It is easy to look to charismatic leadership and willpower as explanations for collaboration done well, but ultimately neither is verifiable or especially useful as an input to constructing a model for improved collaboration.
- The project developed a series of case studies of organizational success and failure.2 The cases were selected to represent every continent and many industries as well as government and nonprofit groups.
Organizational collaboration is an inherently broad topic, and it was not immediately apparent how to approach it through quantitative methods. Initially, team members tried to categorize case studies based on personal judgment, but these assessments were subjective and therefore unreliable. People who read a case would place it in different categories. For example, was a top-down structure with few layers the same as a top-down structure in which directives were passed through many layers of a hierarchy? To be more objective, the project team collected numerous organization assessment instruments and scored the cases on 87 variables. A substantial effort went into establishing inter-rater reliability. This process allowed statistical techniques to be applied to the data generating an analytics-based view of collaboration, a view whose results were then used to develop measures of the conditions that underlie collaboration in organizations. Our intention was that an understanding of those conditions would afford insights that would assist business leaders in unraveling the collaboration riddle in larger organizations.
While there are many things to consider in looking at an organization, three conditions emerged consistently in our study among organizations where large numbers of people collaborate effectively. One (or more) of these three conditions was absent in every case of organizational failure in the study. These conditions are:
- Belong: People collaborate on behalf of organizations they feel connected to.
- Believe: People collaborate when they commit to carrying out specific actions.
- Behave: People collaborate when they share a common understanding of how things are done.
While employee engagement measures are widely used to assess connections to the organization as a whole, collaboration depended to a large degree on which parts of an organization people feel connected to. This notion of parts helps capture the complexity of modern organizations. People might feel most attached to their local work group or their product line, their function, their geography, or even the organization as a whole. Matrix structures increase this complexity by asking people to hold two or more loyalties.
Belong: Who Will People Listen To?
Brambles, a large supply chain and information management organization, was embarking on a new global strategy to bring together its dispersed business units to achieve the benefits of being part of the whole. The challenge they faced was that their people identified more strongly to the business units and countries they were part of and did not identify strongly with the organization as a whole. As a consequence, the pace of execution on organization-wide initiatives was going slower than expected. According to one senior executive: “The people element in the ‘how’ component of the new strategies was critical to our achievement of goals.”
Based on knowledge of where their people felt they “belonged,” Brambles changed the way they communicated with employees, shifting from an organization-wide approach to one that involved greater focus on communicating specific organization-wide messages at the country and team level.
Brambles has seen their organization-wide initiatives gain great traction, they have increased people‘s commitment levels to group initiatives and this has translated into strong business results. They have also seen an increase in the level of belonging people have to the organization as a whole. By understanding where the strongest levels of belonging were and having local leaders become the initiatives’ primary drivers they have achieved their goals.
Collaboration works notably well among people who feel loyal to the same parts of an organization. Our research indicates that those who are loyal to different entities with separate agendas struggle to collaborate. Additionally, it shows how a sense of belonging influences organization change. Most initiatives involve shifting the parts of an organization that matter most to people: encouraging functions to collaborate more effectively, merging two or more separate entities, or asking people in independent organizations to adopt shared processes or data definitions. These change efforts often founder when the need to shift where people’s loyalties currently lie is not considered.
Rank and file employees in the more than 100 projects studied tended to feel most connected to local work groups, whereas more senior people tended to be more connected to the larger structures of the organization. In organizations where the senior leaders resembled the rank and file in being locally focused, the organization usually found it hard to achieve broad goals. As organizations adopt more virtual and global forms, creating a sense of belonging becomes all the more challenging.
Specific takeaways about the Belong dimension:
Importance: A group of people will work together well when they feel they are part of the same organization.
Warning signs: The groups whose performance is measured are not the groups that leadership wants to make the focus of people’s efforts. Which group evaluates performance, sets compensation, decides promotions? People often lack connection to others whom they are expected to collaborate with.
What can be done: Either connect changes to the groups people care about or raise the importance of groups by changing organizational structure, measurements, responsibilities, structures, or effective team building.
Lack of a shared sense of belonging is often a problem in postmerger integration situations, as people move slowly to shift their sense of belonging from their prior company to the new entity they have joined, but this issue is much broader. We see numerous initiatives around the world with “One” in their title, usually “One Company” or “One Unit”. Leaders often call for these initiatives to achieve economies of scale, to make their organization easier to do business with by being more consistent and/or simpler in how it goes to market, or to promote more cross-selling. Repeatedly this sort of initiative runs afoul of what essentially is the tribalism that persists in our globalized world. That sense of the local team being the one that matters, often reinforced by the local team holding its members’ performance reviews and compensation, is a persistent barrier to “One” initiatives that are often more compelling logically than emotionally.
Specific takeaways about the Believe dimension:
Importance: Defining the behaviors that people have to commit to helps to drive collaboration and sometimes highlights weaknesses in strategies and plans.
Warning signs: You are communicating, but things are not happening. People are sitting on the fence, unengaged by organizational goals.
What can be done: Ask whether people know what is expected of them. If they do, are they motivated? If motivated, are they encountering barriers to new behavior: inadequate processes; unavailable data; measurement systems or organization structures misaligned with goals?
In the Brambles example, the issue of connectedness was addressed by aligning who was communicating with people’s loyalties. Other ways to address “Belong” issues include changing organization structures so that people who need to collaborate belong to the same organization, as well as changing where both profit and loss and individual performance are measured to the parts of the organization that the leaders want employees to give their loyalty to.
Woody Allen once quipped that “80 percent of success is showing up.” Perhaps that should be amended to “showing up and doing something.” Collaboration requires action, doing things that sometimes are challenging. The more successful cases we observed were marked by people taking action, and analysis of patterns of action and inaction suggested a new approach to thinking about change readiness—assessing the propensity of people to act in new ways.
Change readiness has typically been operationalized as people’s attitude toward changing behavior, measured on a scale that runs from favorable to opposed. This approach has at least two shortcomings.
First, individual attitudes are poor predictors of behavior.3, 4 People hold many attitudes and are subject to numerous other influences, so frequently the influence of any one attitude is weak. For example, one may aspire to provide a creative solution but also to deliver on time, deliver within budget constraints, and deliver a solution that is less risky than something novel, and so aspiration to be creative may be overwhelmed. An improved assessment approach addresses this by measuring intention to carry out actions, rather than attitudes toward actions. This was suggested by the observation of Kurt Lewin, the seminal social psychologist, that intention is a much more effective predictor of behavior than attitude.5
The notion that people either overtly favor or oppose an action is often inconsistent with what one sees in contemporary organizations. Opposition is more often passive than active, expressed as inaction rather than as vocal opposition. People publically state a strategy is wrong, but they can be slow to take action or they leave the work to be done by others. Attitude measures are not designed to capture the possibility of passive resistance.
There are five ways people respond to behaviors that can achieve goals, as shown in figure 1.
This approach provides a more conservative assessment of an organization’s ability to act than traditional change readiness because “supportive” people are no longer counted as proponents of action. These five categories help identify what interventions should be deployed to change behavior. If the majority of people are “undecided,” this is usually a problem in credibility; people need to be convinced that the need to act is not going away. Many of us will recognize “initiative fatigue”—the cynical but all-too-often realistic assessment that organizations have carried out numerous change initiatives that have petered out, and hence sitting on the fence for a while is a rational response. Early adopters are usually a minority compared to those who wait and see. If the majority of people are “supportive,” that often suggests people do not recognize a way to take action from their own positions. Sometimes they are right; but often the high-level goals of an organization need to be translated into specific actions that people may not be willing or able to develop on their own.
Consider a consumer products company in South America that had experienced several failed process improvement initiatives. What it found was that passive resistance increased dramatically among managers with more than four years’ experience. The managers’ sense of belonging with the company as a whole also dropped. The effect showed up when analyzing demographic details of the Believe scale’s passive resistance measures. This was not something that was identified when the company compared who favored and who “opposed” initiatives. What became clear was that a large group of managers were blocking the company’s initiatives. The company responded by putting a program in place to work to reengage these people.
We hypothesized that “opposed” would be a very small category in modern organizations, and the data have borne that out. The studies found that active opposition is seldom expressed by more than 10 percent of an organization’s members, while the various forms of passive resistance are much more widespread.
Contemporary observers of organizations see a shift underway from command and control structures that had been a historical hallmark of the Industrial Age to networked structures that take advantage of the global and virtual possibilities of modern work.6, 7 While this may be true, there are several extensions to this view suggested by analysis of organizations in our study:
- While networked structures are emerging, they are not replacing command and control, which often continues to work well.
- While “command and control” and “networked” are useful terms, they can mask a great deal of variability within each category.
- No one organizational type was superior; many different types worked well in different circumstances and in pursuit of different goals. What stands out in cases of effective organizations is that there is a high degree of agreement across an organization on what type of organization is in place.
Absence of agreement can undermine organizational effectiveness.
Specific takeaways about the Behave dimension:
Importance: There are a lot of ways for people to work. People who agree on “how things get done around here” spend more time doing those things rather than tripping over inconsistent expectations and approaches.
Warning signs: Is critical work getting done? Done well? Decisions made? Decisions implemented?
What can be done: Drive to clarity. Agree on governance processes and decision rights. Agree on what work will be done in a consistent way and what will be done in independent ways. Reconcile the differing models of leaders and followers. If two organizations with different models need to collaborate or merge, spend time to resolve their differences.
The organizations in our case studies were compared using an analytics technique—the Self-Organizing Map (SOM). The SOM extracts patterns from complex data. The SOM solution that identified eight types of organization in the data provided an effective balance of complexity and clarity.
Each organization type was assigned the name of a leader-follower pair. The resulting set of patterns is shown in figure 2.
The models are arranged on two dimensions. The vertical dimension describes how power is exercised: from top-down forms such as Landlord and Tenants organizations to Community Organizer and Volunteers organizations where the power resides with the members. This dimension conveys variations in the degree of top-down control. The horizontal dimension conveys how work is carried out: from the Conductor and Orchestra organizations at the left, whose work is scripted and repetitive, to the Producer and Creative Team organizations at right, in which individual practitioners act in distinct ways. Again, the dimension as a whole conveys gradations from scripted to creatively emergent.
Four organizational types occupy the end points of these two scales, and four more represent hybrids that combine characteristics of the organizations at the end points. Thinking of this as a clock face, the three organizations from 9 to 12 represent varieties of Command and Control while the other five organizations are varieties of Networked organization.
Each of these eight models can be a path to success. One observation from comparing success with failure is that organizational success can be more a function of agreement on the model in use than of which model is in place. Data from client diagnostics bears this out: Organizations with lower agreement on the model they use have lower Believe and Belong scores. Disagreements on the model can arise between leaders and followers or between separate parts of an organization whose models reflect their tasks or that have come together in a merger from different backgrounds. These disagreements often are hard for people who experience them to describe; putting them in the structure shown in figure 2 allows models to be compared in detail and processes to be put in place to reduce areas of disagreement.
Each of the eight organization types has distinguishing characteristics that enable it to achieve its goals. By understanding what type of organization is pursuing a goal, customized approaches can be generated for that particular organization.
Consider the experience of a large commercial bank in Europe that was struggling with the results from several acquisitions it was forced to take on during the 2008 global financial crisis. The senior leaders of the bank saw the organization as one of Community Organizers and Volunteers; their approach was to allow individual entrepreneurs to develop books of business, with the most profitable efforts ostensibly leading to personal success in the organization. However, the junior staff were not seizing such opportunities, and this was reflected in poor operational results. Essentially, they wanted leadership to show them the way, while the leaders were saying: “Figure it out yourself.”
Given the need for a rapid response, it was more effective to convince half a dozen senior leaders to become more directive in addressing the staff than to convince thousands of staff members to accept their leaders’ view of how they should become independent, empowered actors. But what emerged in our project was that the most important aspect was to establish agreement so leaders and followers could work together effectively.
Thinking back to the beleaguered telecommunications executive at the start of this article, this perspective might have helped. There are plenty of organizations that will be responsive to top-down leadership—“I told them what to do”—but effective engineering cultures often have strong strains of Producer and Creative Team in them, and Creative Teams want the authority to solve problems on their own rather than being told the solution. The executive’s approach of “I told them what to do” probably planted the seeds of failure. What if she had laid out a set of measurable goals and then encouraged empowered teams to generate solutions?
Most of our research on organizational failure found substantial lack of clarity in the organizational model. In what has emerged as something of a modern classic case study, the lack of clear decision-making authority in the failed response to Hurricane Katrina immobilized ample assets that might have been focused on the situation sooner and to much greater effect. Even a somewhat tongue-in-cheek analysis of the breakup of the Beatles pointed out that their successful period had been under the leadership of a landlord manager (Brian Epstein) but that after his death the band’s inability to agree on a governance approach undermined their unity.8
These organizational models build up over time and are part of the organization’s culture. While the thought of moving to a different model may seem attractive, it is not as simple as swapping out one program for another in a computer’s memory. A full-scale change in model is generally a multiyear effort, requiring consideration of the structures and processes that support the current model and how those need to be changed. This only can be undertaken by a leadership team with a long view.
As a consequence, as in the bank example above, organizations often adopt the quick solution of having leaders work in the mode their followers want. Alternatively, there are times when leaders want an organization to work in a model different than the one it uses. A frequently seen example is regulatory compliance on part of the work of a group of Builders or Citizens who generally set their own work routines. This requires “organizational judo,” essentially techniques for getting people who are in one model to act in accord with another model. Akin to martial arts, organizations usually fail if they try to dictate such changes; it is far more effective to try to channel the actions of employees in the directions the organization needs. For instance, Citizens want to make their own decisions, and while that can be respected, the judo approach can constrain decision making by introducing requirements, reporting obligations, and deadlines.
Levers to pull
W e developed the As One model to encompass the implications of three factors that enable people to work together effectively:
- Belong: As organizations grow in complexity, that complexity can either encourage or discourage action. Action tends to be taken on behalf of those elements of an organization that people feel strongly about. A simple call to action may not exert influence unless it represents or comes from a part of the organization that people belong to.
- Believe: In considering how to encourage people to act, consider the several ways people respond, including passive resistance, rather than a single dimension that runs from support to opposition.
- Behave: Many organizations pick an organizational model that is appropriate to their culture and the tasks they have to accomplish. Organizations seldom struggle from having implemented the wrong model but rather because they have not made a decision about what their model is or they contain groups that hold to different ways of working together. Understanding the ways organizations operate enables tailoring approaches to strategy deployment and change that reflect the organization itself rather than applying a universal model.
While these three factors do not in themselves amount to an answer key to the complex problem of collaboration across large organizations, they provide a basis for charting a course in what is typically a journey of ambiguity.
Consider a products and services company that wanted to improve its revenue and profit in global markets. The company had gone to market primarily by business unit and, in an informed strategic shift, had announced a greater focus on going to market by country. Even on the surface this looked to be an ambitious change in direction.
An assessment of the Believe dimension found that 50 percent or fewer of the company’s managers were committed to collaborating in ways that would drive the new strategy. Many of those managers were undecided fence-sitters who expressed concern that the new strategy was “not in their organization’s interest.” An important goal became to generate more commitment and less fence-sitting; but to be done effectively, this would require more than exhorting people to do the right thing.
As assessment of the Belong dimension shed light on managers’ concern about “the organization’s best interest.” In this company, function and work group command higher loyalty than business unit or country. The results suggested strongly that managers have pursued their function’s interests above those of the business unit, especially since individual results are measured by function. This in turn suggested that the issues that have hindered going to market by business unit will have a comparable impact on going to market by country. A concerted effort may be needed to make country more important, perhaps by putting more emphasis on country results, perhaps through leadership exercised by country leaders.
The assessment of the Behave dimension further suggested what is needed to achieve this. The organization has a strong preference for one model—Architect and Builders. When this model works well, top-down leaders create blueprints of inspiring goals that followers work on in creative ways. In this instance though, there are major issues to resolve. Is there a single global Architect of the strategy, or is each country leader an Architect? If you decide on the latter approach, how do you manage the variability in country leader capability? How do you create engaging projects for each country, and how do you manage the unequal distribution of Builder talent across countries? These are anything but straightforward questions, but the recognition of Architect and Builder provides a framework for crafting an appropriate solution.
Building an effective organization, as in this example, is a work in progress. Solutions may solve particular problems but do not make an organization permanently capable of resolving all challenges. However, the notion that there are three variables to address with levers to pull to address them is a step toward moving what is often seen as the “soft stuff” of organization effectiveness to an analytical and rigorous plane where measurable progress can be achieved.
- This does not cover every instance of transformational or strategic change, but most of them.
- The core model was built using the success cases. Failure cases were developed later as validity checks.
- G. Maio and G. Haddock, The Psychology of Attitudes and Attitude Change, Sage Publications, 2010.
- E. Aronson, T. Wilson, and R. Akert, Social Psychology (7th Edition), Pearson, 2009.
- K. Lewin, Field Theory in Social Science, Greenwood Publishing Group, 1975.
- T. Malone, The Future of Work. Harvard Business School Press, 2004.
- J. Hagel III, and J. Seely Brown, The Power of Pull: How Small Moves, Smartly Made, Can Set Big Things in Motion. 2010.
- P. Norman, Shout! The Beatles in Their Generation, Touchstone, 2005.