How Emergence Works in Organizations

Emergence isn’t about collecting a bunch of fine people and letting them do what they want with no structure of any kind. In fact, emergence needs structure–just not the same kind of structure traditional organizations use.

Functionally speaking, the traditional top-down structure is a method of augmenting one person’s ideas with the minds and bodies of others–like a human Voltron.

Voltron with an overlay of an org chart

This is familiar to our way of thinking–a reflection of Industrial Age ideals of order, rationality, and efficiency. But Voltron is ultimately a machine, and as we enter the Second Machine Age, organizational structures that mimic machines are going to fall by the wayside: they are less efficient than actual machines, and less adaptive (in an evolutionary sense) than organic communities.

Consider the example of Microsoft’s .NET versus the open-source Ruby On Rails: the closed, controlled, heavily planned product has consistently lost market share for years, while the decentralized, agile developer community has turned out a consistently more popular and more useful product. In terms of external incentives and long-term planning, Microsoft should have the better product. Instead, Microsoft is planning to make .NET open-source because of the spectacular failure to keep up. The organic community has unequivocally trumped the traditional model.

Let’s take another stab at that organization chart. Take the person at the top and place her instead in the middle. Connect her to all her sources of information, her co-workers, the people who report to her, the people to whom she reports. Do the same for each of those people to whom she’s connected, outward until everyone in the organization is accounted for, and connect each of them with their points of contact to the outside world. What you have now much more resembles a neural network:

Map of the neural network of a mouse

This isn’t an aspirational image, it’s a reflection of reality–a reality that the former image attempts to control and confine into a linear hierarchy. For comparison, here is a map of Twitter employees using the platform:

Network map of Twitter employees

Seeing an organization in this way–as it organizes itself, and not as the linear design into which we try to squeeze it–reminds us of the first principle of emergent structure: the number and quality of connections improves emergent behavior.

This isn’t a call to increase the number of networking events at your business. Human beings will create connections on their own if given the right environment.

Consider Google’s practice of mixing functional groups on the floor. While this would be anathema to a linear hierarchy–I would have to walk to another part of the building to talk to a co-worker–at Google, the people around you may become your co-workers. Some businesses have replicated the design without taking into account the hierarchy and culture, resulting in marginal improvements at best, and often employee frustration. (Even at Google the practice is not without its drawbacks, but it is an intentional culture play.)

The quality of the connections is likewise based on trust; the traditional organizational structure makes trust unnecessary, and by making it unnecessary, undermines it. This leads to a second principle of emergent structure: where traditional organizations depend on chain of command, emergent structure depends on social contracts.

A social contract is, at its most basic, an agreement held in common between any number of people, either written or implied. Top-down structures choose policies and impose them. Leaders of adaptive organizations make a case that a policy is in the best interest of the organization, and modify it based on the needs and applicability to a particular population. We all agree that we need to follow a certain regulation or we will face prosecution, therefore we promise to follow that regulation and hold one another to it.

In this way, instead of imposing a rule that employees follow like the speed limit (seven miles per hour over isn’t really speeding, you know), employees are accountable to one another for their actions. We all know we have made the promise with one another; when I violate the promise, I have broken my promise with everyone in the organization.

Social contract underpins the entire structure of a culture. When I expect that I have ownership over my work and it won’t be taken away from me and given to someone else, that is an implied social contract. When I expect that I can do my work every day without fear of sexual harassment, that is (usually) a written social contract.

Social contract also defines the leadership of an organization. A person can be an owner if he has enough money, and he can be a manager if he has enough connections, but he can only be a leader if he has people willing to follow him.

Finally, in addition to good connections and social contracts, emergent structure requires process. This may seem to contradict what I’ve said so far, except that human processes are approximate–they are always applied using judgment and a “feel” for the particular situation.

Human process is an interesting phenomenon. Machine processes can be made to take variables into account, but machines don’t make judgments. And machines don’t vary processes in ways they haven’t been explicitly programmed to learn, i.e., they don’t introduce personal experience or personal interpretation. Emergent behavior in an organization depends upon human experience and variability, working against an established process, to produce unexpected outcomes. Only processes that are documented and followed can be improved. Everything else is folk knowledge, which is improved by rare individuals and passed piecemeal to others.

This kind of adherence to and improvement of process depends wholly upon social contracts and quality connections: I must know that I am not documenting my changes to the process on behalf of some uncaring management four levels above me, I’m keeping rigid documentation on behalf of people like myself in the organization who are doing their best to cover for one another and keep things moving.

This is why I encourage you to abolish the idea that there is a “top” or “bottom” to your organization. The goal is to have a structure where everyone finds the most appropriate place for as long as that place is appropriate, not a structure where everyone struggles to climb as high as they can. In order for this to work, we also need to rethink the role of leadership in such an organization–because it is still critically important–and what that leadership is attempting to do.

I’m eager to get your thoughts on all this. Did any of this sound like an organization where you’ve worked, either under the traditional model or under a more adaptive approach? Have you seen leaders embrace the “neural network” of organization and truly try to engage all their members in uniquely meaningful ways?

When You Stop Controlling, What Emerges?

In my last post I mentioned emergent behavior is a strength of adaptive organizations.

A definition of emergent behavior: A behavior is emergent if it’s something the collective does without any individual intentionally aiming to  it belongs to a collective but not to the individual components that make up the collective. It’s sometimes used interchangeably (even though it isn’t strictly interchangeable) with “self-organizing.” The movement of a flock of birds, for example, isn’t led or organized, but nonetheless ends up creating a beautiful, organized pattern–so much so that scientists long hypothesized that there was some sort of organization or leadership.

On the surface level, it’s easy for leaders to get excited about emergent behavior. It’s engaging, it’s innovative, and because of its connection to artificial intelligence, it sounds futuristic. But as soon as you start to dive into the details of emergence, it starts to get scary very quickly.

Emergence not only unplanned but unplannable. It doesn’t obey a strategy or market tactics. We’re able to tell the stories of the individual parts of a system, and we’re able to tell the story of the system as a whole, but the individual story and the story of the whole are not clearly and causally related. In other words, it’s not possible to control emergent processes to create a specific desired outcome–at least not without some kind of mind control.

To most business leaders, this makes emergence a non-starter, or at the very least makes it something they would limit to a dark corner of the dreaded Innovation Department. Our objective is to contain what we can’t control and subdue it with everything we can control. Giving up control is the exact opposite of what I try to do every day.

The good news is that, although I’ve written quite a few words already about emergent behavior, I haven’t yet mentioned what emerges.

  1. Leadership. We all know leadership and management are not the same thing. Management is a way of delegating down a chain, percolating down to (one can hope) the right people for each task. By contrast, a leader only exists because there are people who choose to follow. Leadership is visionary and inspiring–not the mushy feel-good kind, but the kind that puts such a strong idea in your head that you will stick with it through disagreements and difficulties, even through loss and sacrifice.
  2. Culture. Culture is already an emergent phenomenon, but a particularly interesting one because it works as a feedback loop: culture emerges between people, but then it is identified and qualified and absorbed by the individual who goes on to influence the culture. But even with that in mind, managers still attempt to control culture as though it were something that could obey their will if only they tried hard enough.
  3. Strategy. Slime mold is so efficient at finding food that scientists spent several years laboring under the hypothesis that they would find “controller cells” that coordinated the behavior of the colony. Emergent strategy adapts to present conditions more quickly than planned strategy because its members are responding directly to stimuli, and due to its parity with the pace of culture it is just as good or better at future-readiness. What’s more, due to its emergent properties, it scales much more effectively than planned strategy. (The flip side of its scalability is that emergent strategy is less effective in smaller organizations; emergence is itself something of a network effect.)

I want to emphasize that emergence doesn’t mean surrendering your organization to chaos and anarchy. I will delve further into the subject, but for now just imagine that your organization is your back yard. At present, at its best, your backyard is neatly mowed, has some nice lawn furniture and a couple of lawn games. I’m not suggesting you should let your lawn be reclaimed by nature–covered in scraggly bushes and weeds–but that your lawn could be a Japanese garden, a beautifully manicured work of nature. Without the self-directed growth of the plants, it wouldn’t be possible, but likewise it wouldn’t have form or meaning without the care given to cultivating and pruning those plants.

I’d like to take this opportunity to do a check-in with you. What excites you about emergence? What concerns you? I have several posts in the works, but I’m interested in where you want this to go next. Let’s have a conversation in the comments.

The Strengths of Adaptive Organizations

While most of my posts can be applied to many different kinds of organizations, and even more can be applied to businesses specifically, I write all my posts with adaptive organizations in mind.

Adaptive organizations are generally loosely-structured, non-hierarchical, and depend on temporary teams to pop up and disband on their own. They’re the primary focus of Frederic Laloux’s book Reinventing Organizations, as well as the driving philosophy behind the consulting firm Undercurrent.

Adaptive organizations are designed to maximize the co-operation of human and machine. Unlike Industrial Age organizations, adaptive organizations (what Frederic Laloux calls “teal” organizations) do not rely on humans functioning as machines. Instead, they depend on the value created by healthy individuals, collected from diverse backgrounds and bonded into communities by a common vision for the future.

The contrast between the two concepts can be so pronounced that some can’t even fathom how these futuristic organizations would work. But the fact that adaptive organizations are already beginning to emerge (as with Spotify, Valve, and GitHub) shows that they aren’t just a philosophical exercise. They are real, they are successful, and they will continue to thrive.

Laloux outlines three principles of adaptive organizations: self-management, wholeness, and evolutionary purpose. While these are incredibly useful guidelines for creating a future organization, they don’t quite explain why adaptive organizations work. I’d like to propose three corollaries to Laloux’s principles.

  1. Traditional organizations rely on planned behavior; adaptive organizations encourage emergent behavior. Traditional organizations are heavily planned: they hire people with specific skill sets to fit into specific roles and accomplish specific tasks that make up a system that’s carefully designed to play out the vision of the entity at the top. This ultimately makes traditional organizations less than the sum of their parts. Adaptive organizations operate at the opposite end of the spectrum: they expect employees to manage themselves and one another dynamically. Fixed hierarchy is counter-productive because it limits deviation from an established agenda; in a fixed hierarchy, I don’t have much room to do anything that doesn’t directly benefit my immediate supervisor, and he in turn has little room to do anything that doesn’t benefit his immediate supervisor. Designing an organization to encourage emergent behavior means opening up to unplanned innovation by anyone at any time. It can be equal parts dangerous and game-changing; the art and science of emergent behavior is to minimize the danger without discouraging the game changers.
  2. Traditional organizations consolidate efforts in an attempt to design the best, most efficient single outcome; future organizations rely on multiple discovery to generate iterative, multi-dimensional innovation. When a traditional organization discovers two different efforts to accomplish a similar goal, it’s seen as inefficient. Duplicative efforts are shut down and/or consolidated into one another, leading to political battles and possibly resentment on the part of the employees who were trumped. These consolidation efforts frequently fail, either in process (they are never completed) or in product (the outcome is too unwieldy or unhelpful). Multiple discovery allows several efforts to reach the same point from multiple directions, or to reach different points from a similar origin. The outcomes of the individual efforts tend to be leaner and more focused, and if one option fails there are others at the ready.
  3. Traditional organizations depend on metric productivity (output divided by hours divided by pay rate); adaptive organizations develop unique value. Metric productivity is the enemy of unique value: it suggests that all products, customers, and employees are comparable and judges each employee against some Platonic ideal of productivity. Metric productivity is what causes us to believe that putting in more hours makes us more valuable to our employer, that what we do to our bodies in our off-hours isn’t important to what happens when we’re on the clock, that our mental and spiritual and social well-being is something we do on our own time and work doesn’t factor into it. But metric productivity isn’t just bad for employees, it’s also a dead end for employers. If your concern is wholly for metric productivity, chances are high that you’re in competition with someone. Competition is a sinkhole. If you’re not digging yourself out of it and creating unique value, you’re bound to lose.

This is how adaptive organizations can thrive in spite of the concerns that keep leaders locked into traditional models. Adaptive organizations eschew the assumptions of traditional organizations–efficiency, competitive pricing, planned behaviors and outcomes–and take the lead because they engage both employees and customers in a way that makes traditional competition obsolete. They also gain efficiencies in unexpected ways–from Buurtzorg spending less time on patients by spending more time with them, to Netflix’s “the best are 10x better than average” philosophy. In the end, adaptive organizations are even better than traditional organizations at traditional metrics, because they focus on purpose and put the future of their organization in the hands of each individual. Instead of focusing on functional planning, an effective leader provides focus, narrative, and inspiration to the efforts of the collective–as Saint-Exupery puts it, she teaches them “to yearn for the vast and endless sea.”

This post serves as the frame for my next several posts, in which I’ll tackle multiple discovery and dig further into emergent behavior to provide some practical understanding of how to apply these concepts to a real organization.

What do you believe about adaptive organizations? What’s keeping you from applying these principles to your own organization? I look forward to discussing with you in the comments.

Feedback: The Motivation Superpower

Intrinsic motivation, left to itself, can be unfocused. This is especially true across an entire organization. There are ways to improve focus through establishing shared values and getting everyone to tell the same story, but there are also mechanisms for improving the focus of an individual’s intrinsic motivations. Few of these mechanisms are more fundamental than feedback.

I don’t mean peer review forms or a semi-annual sit-down with the boss. I mean simple feedback loops that work throughout every day.

Simple feedback works like this: A subject takes an action, there is a reaction, and information about the reaction is returned to the subject, who can then use the information about the reaction to modify her activity. I touch a hot kettle, the kettle burns my fingers, my nerves send information about my fingers burning back to me, and I pull my hand away. This is how fundamental feedback is. But because so much business in today’s world is abstract, we have to construct feedback loops deliberately rather than expecting feedback to happen on its own.

Lack of feedback can quickly erode motivation. And the more entrepreneurial or “self-starting” a position is, the more important feedback is to the person in that position. Feedback is your sight, like a bat echoing its own songs to understand the contour of the world around it. If you don’t hear an echo, how do you know what to do?

Yet for how fundamental it is, it’s surprisingly easy to forget. And then it’s surprisingly easy to chalk up motivation problems to lack of incentives, or poor leadership, or other priorities getting in the way, when really the people around you are lost in a world that doesn’t echo back at them.

How can you create effective feedback?

Feedback must be immediate, contextual, and apparent. Feedback is a behavioral stimulus–it has to fit both the time and the context of the action that caused it, and it has to be clear and concise in order to reveal information that’s useful for subsequent action.

This doesn’t mean feedback is always a result of things that are done–sometimes it’s the result of something that’s undone. Networking sites like LinkedIn and dating sites like Match.com tend to provide feedback in the form of a percentage completion bar to let you know how “complete” your profile is. Of course, your profile on these sites is as complete as you want it to be–but by creating this bit of feedback, such sites are able to encourage participants to improve the quality of information about themselves without offering any incentive other than having a “more complete” profile.

Feedback is a leadership superpower because all feedback is either grounded on some fixed point (values), directed toward some fixed point (objectives), or both. Thus continuous feedback is a way of aligning the efforts of a team toward the same values and objectives. And if you focus on those ends–values and objectives–when providing feedback, you can effectively avoid micromanagement while getting results that both satisfy your goals and represent your team.

Sometimes as a leader, I may have to manufacture feedback. This may require a shift in perspective: rather than believing there’s no feedback available because something is tied up in political limbo, I may need to provide feedback on the work itself–its quality, its relevance, etc. My team member will be able to take that feedback and apply it to other efforts. As a consequence, they’ll also be creating value that better fits my own vision, since it’s directed toward my feedback.

I may also have to generate feedback for myself. One way to go about this is to establish clear expectations with every completed action. After completing something for which I expect feedback–which does not necessarily mean something that requires “notes” or changes–I can mention the kind of information I want to receive and the date by which I would like to receive it, and then follow-up after the appointed time has passed. Remember this information should be immediate (and contextual), concise, and oriented toward fulfilling values and accomplishing objectives; it should as a result be quick and easy for the requested party to provide.

Proper application of feedback can, on its own, stimulate a lot of action without the addition of artificial incentives. It’s the first step in turning intrinsic motivation outward, but it doesn’t yet offer an actual incentive–merely a reflection. The information reflected back at us also implies specific objectives–something that someone outside of us is looking to find, and therefore something we can work specifically to improve, which we do if we have the intrinsic desire to create something useful for another person. Giving feedback without tying it to any extrinsic reward is the second level of motivational strategy.

What are some effective ways you’ve found to provide feedback to others? What ways have you learned to solicit useful feedback from others?

Can We Open-Source the Law?

Here’s a wild idea for you.

When the United States was formed as the first modern democracy, its founders had some pretty crazy ideas about how the democratic system would work. People would be elected and leave their jobs to become lawmakers for several years, and then return to their jobs and let someone else take a stab at it. It was an efficient approach to getting the perspectives of the people who made up the nation to collaborate on laws.

But how would they have written the Constitution if they had today’s technologies and insights? Would it have been different?

Enter Jos de Blok, CEO of Buurtzorg, a nursing organization that turned the business of home-care on its head. Unlike traditional leadership that does extensive strategizing and change planning, de Blok has a blog. When he’s planning a change in operations–usually a small, incremental change rather than a massive treatise–he posts it on the blog and allows 24 hours for response from an organization of nearly 8,000. Based on feedback, he proceeds, amends, throws out, or gathers a task force to hammer out the details.

The specifics of Buurtzorg’s approach will not apply for every organization, but the concept is sound: Centralized (non-hierarchical) management of an open-sourced approach.

Open-sourcing has provided not just stable software but some of the most secure software in the world, despite the early concerns over its transparency and malleability. Yet the approach has yet to make much of a leap to other fields, despite being at its core a knowledge method and not a software-specific method. (Wikipedia is among the few notable examples.)

There’s a lot to be learned from the way open-source groups guide development. Many have small development teams that are able to invest and focus on key features that aren’t getting as much attention on the outside. Some companies identify features that have to be developed in a closed environment in order to be valuable, and embed them within a different version of the product. But there are often so many eyes on every line of the source code that it’s difficult to deliberately sabotage the effort, and developers who comb through the code before a final release are typically able to pull out unnecessary operations and tie up loose ends.

These open-source software projects are at work in the wild, accessible to billions of people, and many of the developers working on them have very little personal stake in the outcome except their own excitement about and investment in the project.

Suppose these same techniques were used to guide your organization, every member of which is deeply invested in your policies and decisions. Instead of investing in a complex analysis of the state of your company, you’d get immediate feedback from the people who know it best. Instead of deciding on a change plan and having to get buy-in from every last member of your organization, you’d already have a good chunk of your organization on board. Instead of having to course-correct in the middle of an enormous project, your incremental changes would correct themselves.

If it sounds idealistic, that’s because it is. Don’t misunderstand: It’s a completely practical approach, and may even become a necessary approach to managing large organizations in ten or twenty years. But it’s also idealistic because it requires leaders to relinquish power, and perhaps just as importantly it puts them on uncertain footing. Leaders not only want power for its own sake, they want power because they don’t trust their people, and this approach to an organization’s policy and strategy relies absolutely on trust.

Fortunately there are ways to develop that trust, some of which I’ve already written about. You can treat your employees as partners. You can remove the stigma of failure. You can establish shared values. And as always, you should approach it in the way that works best for your specific organization. But as with any change, at some point you will have to make a leap without knowing exactly where you will land. It can be terrifying, but it’s the only way we accomplish any meaningful change.

Perhaps the Founding Fathers wouldn’t have open-sourced America’s laws even if they’d had the tools. But on the whole they were the kind of people who were always trying to improve upon the work of others. We’d be doing them a disservice if we didn’t try to improve upon theirs, even if it’s only within our own small organizations.

I’m looking forward to discussing your insights and concerns in the comments below.

Rethinking Your Government

Corporate governments don’t often look to the past to see the evolution in tribal and national governments, and consider whether the next step in management will follow along the same lines. In fact, in the business writing and discussions that I’ve encountered, very rarely if ever have I noticed even the suggestion that a corporate “governance model” and a national government might have deep and important similarities.

In fact, the true genius of capitalism from a human-development standpoint was that it decoupled and abstracted political and economic realities. What were formerly struggles between nation-states (consider Machiavelli’s Italy) could now be struggles between nation-corporations, with the toll counted in jobs lost instead of lives lost.

Another impact it had was to more effectively hybridize government, i.e., I am subject to the government of the United States, the particular state and city where I live, and the business for which I work, as well as any other organizations in which I participate. This hybridization makes the world more efficient because each organization can seek the most effective form of government for its activities.

So when I say that most businesses’ governments are totalitarian, oligarchic, and/or socialist, I don’t mean these are negatives. They are simply characteristics that can be properly aligned or misaligned with your organization’s values.

Consider how your organization’s government works. Is it a top-down affair, like a monarchy or an oligarchy? Are there democratic features, and are they really democratic or are you just pacifying the people to keep them from revolting? Who wields the practical power within your organization? And how is your Constitution holding up? Have you added features (like an Innovation Department or a Chief Happiness Officer) instead of revamping your system to fit your present values?

Forward-looking businesses will have to leapfrog the development of world governments in order to keep pace with the future. They will have to skip from where they are (perhaps a little ahead of feudalism) to a post-democratic solution.

Fortunately, we’re starting to get a handle on how such governments work. Frederic Laloux has a great lecture (with Q&A) on the priorities and operations of forward-looking organizations. I have a few things to add, as you will see later this week, but please speak up in the comments about your own experiences with forward-looking governments.

This One Neat Trick Made Alexander the Conqueror of the Known World

Alexander conquered the known world with a simple innovation. It didn’t require a complex new technology; we would call it a “process change.” Instead of giving men shields to protect themselves, the Greeks overlapped their shields to form an impenetrable barrier, a technique they called a “phalanx.”

In these phalanx formations, a soldier’s shield was (according to Gerard Butler) intended to cover himself and the man to his left. The success of the entire formation rested on each individual, but the responsibility of each individual extended to only one other person. It was a manageable, achievable, even simple goal. What’s more, each soldier was able to give up some of his own shield with the knowledge that the man beside him was shielding him in turn. (Unless you were on the far right–that guy was kinda screwed.)

I’m sure someone has taken this idea somewhere, slapped it on a posted and put the word “TEAMWORK” under it. But this isn’t just “teamwork.” “Teamwork” was taking down a mammoth thousands of years earlier. The phalanx was something a little different. Instead of being something abstract like “teamwork,” it was a very simple, practical invention based on two rules:

  1. Each man covers the next man’s weakness.
  2. Covering the man next to you is as important as covering yourself.

If you could apply this technique to a culture–the culture of your immediate team at work, or of your entire company, or of an entire nation–a culture of covering the person to your left–could that culture benefit as much from its resulting unity as an ancient Greek phalanx?

For it to work, there is a hiring issue and a management issue.

The hiring side depends upon hiring a diverse set of people–not just “diversity hiring,” but hiring people with very different perspectives and strengths within the context of the core set of skills necessary to a team. This requires complementing working skills (such as project planning) with social skills (such as empathy).

The management side depends upon horizontal management methods, which is to say peer management. Rather than being responsible only upward, I am also responsible sideways to and for one or more people. Their success is my success.

There are a few possible approaches that I have not personally had the chance to test or observe in a business environment. One is the “linked chain” approach that is seen in some less-formal organizations: Each individual reports to one peer and is reported to by another. (This is a way of placing responsibility for communication on one individual in the relationship; it doesn’t mean the reportee won’t also be asking for help from the reporter.) Instead of seeing these people as higher or lower in the hierarchy, they are peers who are responsible for one another. In a managed organization, the chain can be more effectively “linked” by pairing complementary skill sets, so that one’s strength can cover another’s weakness. By virtue of linking, additional skill sets can quickly be brought in by the rest of the team, especially on larger teams.

Have you had any experiences trying to develop “phalanx” structures in your organization? I would love to hear about it in the comments.