Don’t Reorganize: Run Your Firm As An Agile Network

by Steve Denning

Amazon CEO Jeff Bezos; General Stan McChrystal; Former SRI CEO Curt Carlson

Executives often ask for advice on designing the structure for a customer-centric organization.

Yet this is the wrong question. When they look for a better structure, they are already on the wrong track. All too often, they are assuming that the solution to their problems lies in re-arranging the boxes in the org chart—the fixed hierarchical edifice of the firm, built on authority, roles and rules, and reflected in the familiar pyramidal picture as shown in the chart below.

Such questions usually come from executives in organizations that have inherited a traditional divisional structure of layers and silos. In the 20th Century, every self-respecting firm had to have one. It was a sign that the firm was serious, not some flaky startup that had yet to master the most elementary principles of management.

Such a firm may, or may not, have been on a journey aimed at instilling a focus on customers and operating through small self-organizing teams. Yet inevitably, at some point in the journey, the firm comes to the question of structure. They can see that their static org chart is getting in the way of achieving their goals. The executives gaze at their org chart and think: is there a more customer-centric structure out there? What would it look like?

Typical Organization Chart

Yet starting from the structure and the org chart is to start from the wrong place. It is to imagine that the solution to the firm’s problems lies in re-arranging the organizational units and the relationships between them. Should Tech report to Operations? Where does Marketing fit in? And what about Strategy? And Finance? And so they launch a major reorganization.

The Grim History Of Reorganizations

As Michael Hammer explainedin his book, Reengineering the Corporation (HarperCollins, 1993), “Not a company exists whose management doesn’t say, at least for public consumption, that it wants an organization flexible enough to adjust quickly to changing market conditions, lean enough to beat any competitor’s price, innovative enough to keep its products and services technologically fresh, and dedicated enough to deliver maximum quality and customer service.”

To this day, organizations often know that they are delivering this and seek to achieve it through a major reorganization. I lived through one some years ago at the World Bank. When the former U.S. Congressman Barber Conable became the president, he immediately launched a “big bang” re-organization. The goal was to make the organization more client-focused, as well as to re-arrange the cast of characters who were senior managers. The boxes were re-shuffled, and every manager and staff member had to apply for a new position. It took several years of effort to complete, as people at every level struggled to find their place and their way in the new structure.

Yet the new cast of characters at the top functioned in the same bureaucratic way as their predecessors, as did the organization as a whole. New processes were introduced but they also ended up functioning much as before. The massive effort achieved practically nothing of substance. It had been a distraction from attacking the underlying assumptions that made the organization so reliably bureaucratic. Until those underlying assumptions changed, it didn’t matter what structures, processes, or personnel were adjusted. The organization continued as before.

Designing an Agile Organization as a Network

The right question doesn’t start with the organization or its structure at all. It starts with the customer. What does you customer and your market need? The answer to that determines everything. Generally, the last thing the customer needs is a new structure. (Ever seen a nimble structure? It’s an oxymoron. It doesn’t exist.)

To meet customers’ needs through enterprise agility, the thinking part of an organization must be run as a network, in which information flows horizontally and upwards, as well as downwards. People may still be located in units for administrative purposes. These units might be seen as their “home base” and these home bases might have names like departments or divisions. But they are not operating like silos or layers, which have priorities separate from the goal of the firm, or arguing in secret meetings for their staff and their programs and their status, or against other departments. Work in a network is mostly done in self-organizing teams and groups of teams that are in pursuit of the firm’s overall goal and that in substance report to the top of the organization, which also seeks to incorporate the wisdom of the entire staff and set priorities from an organizational perspective.

In a firm like Amazon, the top management can be quite numerous. Amazon has three CEOs. And many of the senior managers are called upon to chair planning meetings as if they were CEOs, looking at issues from the firm’s perspective, not from the perspective of the unit they come from. The teams have to justify themselves to the whole top management, not to the manager of their home base.

SRI International under the leadership of Curt Carlson from 1998 to 2014 had a similar arrangement, albeit with different terminology. Priorities emerged from “value forums” led by the top management with widespread participation across the organization. Staff were located on departments as their home bases.

A Network in the U.S. Army

Networks can exist even in organizations as traditionally hierarchical as the U.S. Army. For instance, in Team of Teams (Portfolio 2015), Stan McChrystal describes his experience of operating as “a team of teams” in the U.S. Task Force in Iraq in the early 2000s. His force of crack teams was failing against a network of terrorists: he had to design and create a network on the fly. That meant bringing all the key actors together in a common physical space, to enable horizontal information flows, promote deeper understanding and encourage initiative and decision-making.

This wasn’t just about symbolic egalitarianism. The cultivated chaos of the open space encouraged interaction between employees distant from one another on the org chart. Putting himself in the middle of it enabled McChrystal to keep a finger on the pulse of the Task Force. The open space enabled serendipitous encounters, with the recognition that no one could know in advance what connections and conversations would prove valuable.

Contrast between vertical hierarchy of teams and a horizontal network

The daily give-and-take among people at all levels in the hierarchy was a dramatic embodiment of the fluid communication and collaboration that were needed for success. The daily briefings were a constant affirmation that there was no established script that was being followed willy-nilly. Leaders were engaging others at the level of choice. There was no telling in advance what would be decided. The leaders did not have an outcome that was being imposed. They were open to possibility and surprise. Everything that was said was of potential consequence. The leaders had to avoid acting as though they were pressing for a pre-determined conclusion. They had to be willing to allow for alternative outcomes, whatever the cost to their own egos or the pride of their own units. They had to accept that the best answer might come from anywhere in the network.

Thinking like a Network

Ultimately, an organization functioning as a network requires a different way of thinking, a different mindset, from that of administrators in a hierarchical bureaucratic structure. The principle of the network is one of the three fundamental principles of the Agile organization. It is effective only if the other two principles are in place—the principle of the customer—a pervasive obsession in delivering more value to customers—and the principle of the small team—the presumption that work should to the extent possible be done in small self-organizing teams, working in short cycles.

When designed this way, the organization sidesteps the never-ending battles over priorities and budgets between silos. That problem infects all traditional organizational structures, whether they are simple hierarchies, or matrix organizations, or anything else. You can’t solve or remove those problems by adjusting the boxes and arrows in an org chart. You have to change the way people think about the organization. You have to change mindsets.

So long as there are managers defending their program and their budget and their people, ahead of pursuing the organization’s overall goal, the organization is lost. The organization starts finding that the annual budget process takes up to twelve months and requires an immense amount of time. Yet the process never seems to resolve anything. The battles go on. No battle is ever decisive. Each decision is provisional—until the next battle. A unit might lose a decision today, but it quietly resolves to fight on, at the next opportunity. It awaits the next encounter where it hopes to prevail. In this world, structures don’t budge. These structures are like buildings: they are stuck in the same place. A structure has the value of stability. But it is the antithesis of agility. Just think about it for a moment: an agile structure is an oxymoron—a contradiction in terms.

The correct way to resolve these issues is by way of a dynamic network. In the background, there may be administrative homes for people, but that’s for administration. The real action is in the network. In a network, the entire organization is drawing on everyone’s wisdom on deciding priorities. Priorities are decided organizationally, not by individual units. The home base merely carries out the implications of those organizational decisions. It handles administrative matters such as onboarding, staff transfers between teams, training, career development and exiting the organization. In such arrangements, it hardly matters whether the firm groups its staff together in home bases by profession, or by customer, or by geographical area, or some other criteria. The action lies elsewhere: in the network.

The model of the top-down authority-based org structure rests on the assumption that the top knows best. It enables the top to impose its superior knowledge on the rest of the organization. Roles, rules and authority must be made explicit to reflect that. This was a core belief of almost all big firms in the 20th Century. What it didn’t take into account was that the silos of the traditional organization would become power centers with their own sets of priorities that didn’t always coincide with the goal of the organization. Defeating a rival could become more important than advancing the firm’s goal.

The network model rests on the contrary assumptions that expertise exists throughout the organization, that often others know more and better than the top, and that ideas should be able to come from anywhere. Decisions are made from an organizational perspective.

Conditions of Success for a Network

The contrast between the two models is depicted in Team of Teams, which describes Stan McChrystal’s experience of operating as “a team of teams” in the Task Force in Iraq in the early 2000s, as shown in the chart above.

One glance at the diagram on the right of the chart above is enough to give a traditionally minded manager a nervous breakdown. To such a manager, it is obviously chaos. It couldn’t possibly work. It can’t even be discussed, let alone, explored and implemented.

Yet it turns out that in practice it can and does work. It is, in various flavors, the model that the largest and most successful firms in the world are now using. Amazon is a prime example. Most of the big tech firms have variants of it.

It works better than a top-down hierarchy provided that certain conditions are met.

  1. Goal clarity: There must be a crystal-clear and pervasive obsession of the firm on a common goal. In a private sector firm, that goal needs to be centered on delivering more value to the customer, come what may. In McChrystal’s Task Force, it was defeating the terrorists. Without that, yes, there will be chaos.
  2. Direction: The top must set direction in terms of how and where that goal is going to be pursued. Within the scope of that direction, it should be ultra-easy in the organization for anyone to get resources to explore new ideas. But beyond that scope, however, in the pursuit of new directions, the top needs to be the “chief slowdown officer” who makes sure that the risks of heading in a completely new direction are worth the potential benefits. In other words, if you are a bank, it should be ultra-easy to explore ways of improving banking services, but not to start a restaurant. Similarly, if you are a gaming company, it should be ultra-easy to propose and explore changes to an existing game, but not to create a wholly new game. This distinction is explained here.
  3. Metrics: The organization should ideally have ways of measuring whether and to what extent the ideas being pursued by the different teams are contributing to the overall goal. Ideally, the ways of measuring each activity should be established before the activity starts. Thus at Amazon, the team often spends months devising the measures, before it even begins to work on the idea, as further explained here. In cases like McChrystal’s war against the terrorists, the scene was too hectic to put such measures in place, as things were happening from minute to minute. But the spirit of metrics needs to be pervasive, so as to keep track of what’s working and what isn’t.
  4. Discipline: Where things aren’t contributing to the goal as expected, they need to be stopped or amended, sooner rather than later; if the idea is a good one, but the wrong team was pursuing it, or pursuing it in the wrong way, then it can be relaunched with a new format or with a new team. Thus John Rossman’s team was the third effort by Amazon to create a third-party marketplace, which became a core Amazon business; it followed two prior efforts that were scrapped. The idea was good, but the execution had been poor. So they tried again with a new team. It was a huge success and became a central aspect of Amazon’s business.
  5. One organization: The team reports to the organization, not to any particular unit or manager. Thus at Amazon, staff are assigned to various units for administrative purposes, but the teams don’t report to those units. The activities report to the top management. In other words, you don’t have senior managers protecting, or fighting for, resources for “their activities” in the planning or budget processes. The only question is, is this good for Amazon's mission to add value to customers? That question is decided by a collective group of top managers every six months, not by the units to which staff report administratively.
  6. Self-organizing teams: Once the top decides that the team’s activity is, or continues to be, good for Amazon's mission, the team is free to get on with its work for another six months, without interruption or interference, unless something unexpected shows up in the metrics.
  7. Formal interfaces between teams: Amazon tries to make the teams as independent from each other as possible. Coordination between teams is enabled by formal interfaces of each team’s systems, not by consultation or discussions, which Bezos doesn’t believe in. Thus Amazon Marketplace—almost unbelievably—was developed without much consultation or discussion with the group running Amazon’s own sales operations. The marketplace team used the same formal interfaces as Amazon’s sales operations: there was no need for discussion. (One can imagine in a traditional organization the strife that would have occurred, as Amazon’s own sales operation saw this upstart Amazon Marketplace operation of third-party sellers as a threat to its own status within Amazon.)
  8. Appropriate managerial behavior: McChrystal had to unlearn what it means to be a leader. A great deal of what he thought he knew about how the world worked and his role as a commander had to be discarded.

    “I began to view effective leadership in the new environment as more akin to gardening than chess,” he writes. “The move-by-move control that seemed natural to military operations proved less effective than nurturing the organization— its structure, processes, and culture— to enable the subordinate components to function with ‘smart autonomy.’ It wasn’t total autonomy, because the efforts of every part of the team were tightly linked to a common concept for the fight, but it allowed those forces to be enabled with a constant flow of ‘shared consciousness’ from across the force, and it freed them to execute actions in pursuit of the overall strategy as best they saw fit. Within our Task Force, as in a garden, the outcome was less dependent on the initial planting than on consistent maintenance. Watering, weeding, and protecting plants from rabbits and disease are essential for success. The gardener cannot actually ‘grow’ tomatoes, squash, or beans— she can only foster an environment in which the plants do so.”

Networks are not a Panacea

Let’s be clear: not everything that Amazon does should be regarded as a model. Issues include:

  • Stress levels are high. The rigorous discipline on external performance metrics. There are no free passes for extenuating circumstances or special breaks for trying hard.
  • Sometimes multiple teams are given the same task and compete against each other with different approaches to do the same thing. The losing teams are canned. In theory, the individuals who worked on a canned team can find another team within Amazon to work in. In practice, this doesn’t always seem to work too well.
  • Amazon only applies network thinking to 15% of its staff—the knowledge staff. The rest—particularly the fulfillment centers—have often been managed like 19th Century sweatshops. Some reforms are underway. To the extent that units continue to be run this way, they risk being stains on Amazon’s reputation and will eventually cause serious problems for Amazon unless fixed.

And keep in mind that the management approach is not all-negative for staff who can meet Amazon’s high performance standards. For many, the ability to get on with the job, with a clear mandate and a line of sight to the ultimate customer, without micromanagement, is a huge improvement over 20th Century micromanagement..

Amazon could achieve perhaps even more than it does currently if it could resolve issues realted to stress levels. A less Darwinian approach to handling staff issues might attract and generate an even more productive staff with less turnover.

Networks Are The Future

The progress made by Amazon and other organizations towards running the thinking part of large organizations firm as a network is remarkable. The lesson is clear. Hierarchical structures are for 20th Century bureaucracies. In the 21st Century, the future for market-leading firms lies in networks.

The original article was written to You can read the original at:

Don’t Reorganize: Run Your Firm As An Agile Network
Steve Denning May 9, 2023
Share this post
Sign in to leave a comment
How a new kind of Business Model creates Digital Winners
by Steve Denning