Or

How to be a successful corporate guerrilla

Intro

Most essays just develop out of the logical flow of research and reflection. This one I knew I would write from the start. Because I think it is about most of you and it is about me: the people in the middle. And as promised, this is going to be about what possibilities we have to make ourselves heard and to be able to act. Basically, how to be a guerrilla manager in a hierarchical system. Now, I don’t want you to put up Che Guevara posters, and this is no little red book — but more on that in a moment. See, you probably already do some of the stuff I am going to describe here; you just don’t know it. In the surveys and companies Rensis Likert studied for his 1961 book “New Patterns of Management,” he found that the most successful managers are the ones who do not follow the old-school style. They use different methods, often without knowing how right they are in doing so. And Debra E. Meyerson, an organizational scholar from Stanford, identified in her book “Rocking the Boat” that change agents — the tempered radicals, as she calls them — are often the everyday leaders somewhere in an organization: often not a CEO, not a president or senior professional. So if you ever thought differently or different, had another opinion, you are actually in good company.

Now, there are experts in the field of change management who think that change is only successful if it comes from the top — that authority is needed for any kind of meaningful change. Interestingly, the problem with that was pointed out by one of the promoters of the top-down approach, the author John P. Kotter: if people in the middle sit back and wait for someone at the top to kick off change, they are actually reinforcing the very inertia their organization is trapped in. For him this is still not enough reason to raise the flag of mutiny. And although he points out that some change can come from middle management, he still promotes the old line that change is tied to authority.

But wait: Guerrilla? Mutiny? Didn’t I promise you no posters? You are right. The term corporate guerrilla comes from management enfant terrible Robert Townsend’s hilarious 1970 book “Up the Organization.” Townsend was the turnaround CEO of Avis. His guerrilla is, of course, nonviolent. And I think the term describes the mentality and the situation well. We are inside an organization. We often lack formal power, and we intend to work by indirection. But we do not take over the means of actual guerrillas. Let me add and emphasize that this is about business, not politics. And we are not at war in any way. Not literally — which is hopefully obvious to you; otherwise, if you have trouble following the ethical checks in this essay, please go seek some advice or help, seriously — nor figuratively. We want to discuss ideas, not win an ideological war. And this is not an insubordination manual. Generally, I do not advise going against orders, but rather influencing and working within the limits that we have. We will see later to what extent, and look at the limits in more detail.

But enough of the legally required disclaimers. You and I, we have an opinion growing out of education, experience and reasoning. This is what makes us valuable for an organization — and for society, but that is another subject. All that we looked at in the first set of essays was about what does not work, what makes no sense: the absurd. So you can say that using what gives us value is also a rebellion against the nonsense. Or, to quote Camus — whom we will hear more from in this essay: “Je me révolte, donc nous sommes.” — “I rebel, therefore we exist.”

Why the middle

But why, and why the middle? First, we have to clarify what we are talking about: it’s about change — specifically from within organizations — in any form. New processes, new products, solutions to problems, plans and strategies: all are changes. Change is necessary for all organizations, because they exist in an environment that is constantly changing and they consist of people who are also — you guessed it — changing. Now, who in an organization is best suited to drive change? Nobody, actually, is the most honest answer. The top of an organization usually has the most reach, so they are heard throughout the organization. Unfortunately, they are not very well equipped to ensure that change is actually implemented throughout the organization. The middle of an organization has poorer reach, but what starts there is more likely to take hold. It has good uptake. You can go to whoever has looked at organizational mechanics, and they support this, each in their own way — Barnard, Hackman, Mintzberg, Weick and Follett are just some. So for change, top and middle just have to work together. Good answer, and mechanically it is the best solution. If there weren’t the thing with the content of change: the idea.

1,000 vs. 10 brains. You know it by now. The more people you have, the more brain power — all employees vs. just top management. Plus, in this version of the picture it is about organizational knowledge, or the issue of the particular and the generalization. See, all your employees have the knowledge of all the particular points in your organization. Susan knows the amount of grease the hydraulic press needs. Dan knows which cost center spends how much on express shipments. Maria knows which customer always calls before placing a large order. Ahmed knows why the new delivery schedule works everywhere except on the route through the old town. Sarah knows which spreadsheet really runs the monthly planning process, despite what the official process map says.

Top management is farther away from the particulars — both in the sense of problems and of ideas. It will therefore have to work in generalizations. Now, every generalization necessarily creates distance from a reality that consists of all the particular points. This gap can easily become a distortion. Plus, as Jackall correctly pointed out: the more abstract — his word, not mine; I like generalization better — the more plans and ideas can become hostages in implementation. Nothing can be manipulated with “Grease the machine very well,” but “Infrastructure maintenance is key” can also give somebody a pass to replace perfectly good equipment, renovate the executive floor or fund a pet project nobody actually needs. Don’t get me wrong: generalization is nothing bad or stupid. You cannot steer based on particulars. This is the same principle we know from law. Legal texts are written to match many particular cases at once; that is why they are generalizations. But applying them to particular cases can produce wrong results. That is where Aristotle’s epieikeia comes in: the law has to speak generally, and equity bridges the gap where the general rule fails the particular case. Epieikeia is not some legal loophole — it is the recognition that justice requires judgment, not just rule-following. The middle manager is exactly that: the person whose judgment bridges the general rule and the particular case. Because middle management is the point where the particular and the general meet. So they can be the translator or the bridge. Of course, and to be fair, they can also be the hostage-taker. And that is why we will talk ethics later.

But first: we know that change is needed, and the particular problems are known by everybody — as a collective — within an organization. And in all the essays I have tried to show you that the destination has been described for a century. Follett, Likert, Laloux, even Kotter and many others see the need to change organizational structure and culture. We have also seen that culture follows action. A message from the top is not change; it has to be founded on concrete, observable examples. The lived values. To be honest, Kotter has a different view on who is enabled to start it — remember, he is the top-down guy.

Loyalty and its limits

Before we go to the actual guerrilla tactics manual — just kidding — we have to talk loyalty. And my apologies for taking Swiss law as an example — it’s what I studied, after all. In the Swiss Code of Obligations there is Article 321a, which requires the employee to protect the employer’s legitimate interests in good faith. So that means you owe loyalty to your organization. This is nothing particularly special to Swiss law. Employment law in many countries imposes some form of loyalty or fidelity on employees: you don’t merely owe your employer the performance of specific tasks; you also have a duty to safeguard its legitimate interests — legitimate as in: this wouldn’t work for an employee of the mafia doing crime, and some other stuff — just wait. Article 321d specifies that the employer can exercise the right to give instructions regarding work stuff. Now, your boss acts as a proxy for the employer — remember, boss and employer are not the same thing. Hans might be your boss, but Gummy Bear Inc. is your employer. This idea of a proxy seems logical — especially because most employers are legal persons. So Hans is just the proxy, usually a good enough one, but not always.

The problem starts when he isn’t. What if your boss is a bad actor? He might do stuff he is not allowed to do, or that is not in the best interest of the organization. Or he might just be plain incompetent. Is he then still a proxy for the organization? It depends — the typical answer of a law person; I have a master’s in law, but I am not a lawyer. See, the employer — the organization — delegated the authority to give instructions to the bad boss. At the same time, you have the duty to uphold the interests of the employer. These have been assumed to match — with the emphasis on assumed. Fact is, this is often not the case. Jurisprudence’s answer to this dilemma is that unlawful or immoral instructions are not binding — yes, it is clear about that — but a lawful yet bad instruction still has to be followed. However, if the employee recognizes that following it is likely to harm the employer, their duty of care and loyalty may require them to warn the employer. Well, then all is clear: you will have to follow all instructions anyway. See, that is where the discussion starts.

When in doubt, you would have to follow but scream — or warn. Now, my international friends will scream whistleblower law. And I have to say, good for you, but the legal foundation for that in Switzerland is not really comprehensive and mainly stands on the already discussed general duty to the interests of the organization in Article 321a. Luckily, jurisprudence has set a cascading rule: report internally first, then to the authorities, and to the public only as a last resort. The catch is that this is meant for illegal activities, decisions and instructions, not for bad management. Bad decisions are everyday business; the law only says follow and warn. We will see in a moment what that leaves us with.

So you see that the law only covers so much — the extremes, the really negative stuff. If we look at it again from an organizational standpoint, the organization has, or should have, a purpose beyond itself — also not something given by law. We saw in past essays that it is this purpose that gives the organization its reason to exist. The middle manager, however radical, inspired or guerrilla, will always act in accordance with that purpose. To act against the purpose would negate the organization itself. But this is not only a limit; it is actually the description of the role, because the manager will go against everything that contradicts the organization’s purpose. The guerrilla manager is supposed to be the most loyal part of the organization. Townsend calls this the “loyal opposition.”

My fellow Kant fanboys and fangirls will argue with his essay “Was ist Aufklärung?”, where he clearly states that in your office or function you are “part of the machine” and must obey. Only as a scholar addressing the public are you free to reason and criticize. This might be one of the rare cases where I think Kant might not be applicable to modern business organizations. This is no correction of Kant, but we cannot use him in this case. See, he spoke of the state, the army and the church. Modern companies are different in two regards. First, a company has a much narrower purpose. An employee can judge whether an instruction serves it. With the state, that’s much harder. Second, Kant speaks of organizations as machines — you might remember the image of Taylorism from a century ago. Modern organizations won’t work with passive members; they require active thinking on the part of their members. So our manager goes further than Kant’s officer, because the machine that Kant assumed no longer exists.

Tactics I: Your own realm

But Kant has another point right, and it starts the tactics part of this essay — no manual on how to build a glitter bomb, though, more on how the glitter dazzles. But seriously, Kant says to the taxpayer — which we all are — pay your taxes without much ado; the criticism can be made in the right forum, which for him is the public space. For us in business this translates to: carry out the decision — even if you think it is wrong — but make it clear that you do not agree, and never run down a decision in front of your team while implementing it. Camus has this idea in “L’Homme révolté” as participation without legitimation. Within certain limits you can go along without affirming that something is okay. Contrary to Kant, I extend the forum where you can speak up, because there are limits to never running it down in front of your team: that does not mean you have to pretend that you agree. We are bound to truthfulness — that is actually part of proper information management, which we’ll discuss in a future essay. So toward your team you should be honest and fair, explain the good, the bad and the ugly — they are grownups, they can take it — and say that you raised your concerns. Faking agreement is a lie, which will also be discussed in ethics later. What we never do is mock, invite resistance, sabotage or make the implementation harder. And Kant actually supports this with his example of the clergyman, who does not have to pretend to fully agree with the church’s doctrine. It is fine to separate it from your own opinion. You may run into a limit at some point, and then there is still the exit option — again, to be discussed later.

What is even clearer is your area of authority: your room, your team, your department. For decisions that don’t need permission, don’t ask for it. This can be your team’s rituals or working conditions, or rules for your meetings. Go wild, try out stuff. There are so many innovative ideas. Townsend recommends holding secret ballots every six months on whether a certain meeting is still needed. Or set up a two-meeting structure: the first day for the quick reactions — or reactors — and a second one the next day for the more deliberate reactions and discussions. I sometimes do my fake meditation routine — Stop. Breathe. Think. Breathe. Think again. — to calm people’s minds before a meeting. I know of other workshop leaders who do the same. And it’s not just meetings. Everything that is cheap and easily reversible — some of you might know these as quick wins — just do it; don’t ask anybody. In all our job descriptions there is a certain room to move, sometimes explicitly stated, sometimes just apparent from the absence of written rules. You just have to actually fill that room. This is very much your job. And often you will have even more room you can give to your people. You might have superiors who are very restrictive, but you don’t have to be the same way with your team. We saw in the last essay that the most liberating leaders are those who enable their teams to self-manage.

In our own realm it is also very important that you are a role model, that every action you take is an important part of changing culture. We have seen that culture follows action. Now, values, which are sort of the seeds of culture, do exist before action, but they stay dormant until acted upon. Or you can say: “we become just by doing just acts.” And that’s not me; that is Mr. Aristotle himself: the potential and the actual. And you should also not think that certain actions or decisions are too small to matter. Meyerson correctly pointed out that details are the raw materials of culture. It’s all the small stuff that makes the big fancy culture thing.

Tactics II: Lead with B

Once we leave our own realm, we are in the open wild of the organization, and we have to discuss and negotiate. We already talked about communication, but this is narrower: how to deal with opinions that oppose our own. Just disagreeing is not going to achieve much. We will want to come up with an alternative to a decision or plan we don’t agree with. The organization needs the ability to make a move. If you don’t like a move toward A, then give it B as an alternative. Standing still is not going to work. The best tactic when discussing alternatives is to emphasize your preferred one; let it lead. Don’t spend five PowerPoint slides arguing against A; actually, hardly mention that there even is an A. Let the logic and beauty of B speak. Arguing against something invites defense; discussing your alternative lets the people behind A save face.

Speaking of A and B, it is important not to tear down the existing solution before the alternative is ready. This again is inspired by Kant, who said about gradual reform that one should not break the existing bond before a better constitution is ready to replace it. The same goes for business: don’t dismantle the old report or the old process — however bad they are — before you have something new that works. Otherwise you will leave a gap, which creates chaos and will actually discredit your new alternative.

When it comes to building your idea — new or an alternative — goals have to be concrete and inspiring. But be aware that goals cannot be fixed forever. We constantly test an idea within the PDSA loop: we plan, we do, we learn and we adjust. This is not a weakness — we never lose sight of the direction — but it strengthens the result.

I also mentioned the small wins already. They are fine and can set a good example, but tactically they are most useful if they are built upon. So if you gave a team the possibility to approve customer refunds up to CHF 200 without manager sign-off and it works without abuse, you should immediately try to delegate other customer decisions to that team. Small wins are just first steps.

Eventually you also have to decide which battles to fight — you probably will not be able to fight them all, due to time and resources. And it is also true, as Wrapp points out, that organizations will only tolerate a certain number of proposals. In his article “Good Managers Don’t Make Policy Decisions,” he has a good three-criteria test to make the decision easier: (1) Would it move me toward my objective? (2) Do I have enough support — who supports me, who opposes and who is neutral? (3) How does it combine with what’s already in motion? Especially the last point is really important. Don’t spread your actions out too much, so you don’t lose momentum. Concentrate on actions that accumulate or create synergies.

Tactics III: Words and channels

When it comes to content and discussing opinions and ideas, we always have to connect our proposal to what the organization expresses as its values and purpose. So you need to hold the organization to its own words. This is generally the right thing to do, because — remember — we want to go with the organization’s intent, not against it. But this is also a tactic. Organizations utter so much when it comes to values and purpose — all the slogans, the mission statements, the policies. These can easily be used for your ideas and alternatives. “Safety first” enables us to address all risks. “The customer is at the heart of everything” can be used to enhance value creation in processes — and against cheap efficiency gains that reduce value for the customer, but don’t lead with that. “People empowerment” is an argument to enhance self-management in teams. And these are just a few examples. And it’s not a cheap trick; this is actually waking up the dormant values of the organization.

I discussed Kant already, and that he thinks we should always be allowed to criticize in the public realm. But business does not have a public realm, right? Actually, it sort of has. It is backed by the management system, specifically management systems according to ISO standards. They have audits and management reviews and, in some cases, even worker consultation. These are forums where criticism should be raised and is often sought. The same goes for idea management systems, complaint processes and risk analysis.

And sometimes we may just be missing the right order in which to tackle a change. Townsend helps us there with his very no-nonsense tactics. It starts where it should — with yourself — and then goes outward: build the objective yourself, let the team own it, get the top to agree, then defend it against distraction. Easy enough, right?

Tactics IV: Reach

But what if you are missing what we call reach to the top? There are enough authors promoting the use of individual allies or groups of allies to improve your reach. I do agree; I only want to add a word of caution here. Getting an individual benevolent top manager or some like-minded managers on your side is absolutely fine. But be careful not to let it slide into a power-building exercise à la House of Cards. This is business, not party politics, meaning there is a real danger of getting distracted by building the channels and not the content. And the content, as we have discussed, is the real advantage of the middle manager. Kotter has four criteria for coalition building. Three of them — expertise, credibility and leadership — can all sit in the middle. You just need some position power, and that, I agree, mostly has to be “borrowed” by finding people who agree with you. If your idea supports the organization anyway, chances are high that you will find them.

Not to forget: all of this takes time and persistence. Hackman and colleagues have this line that you need to prepare and lie in wait for the right moment — be organized before the window opens. While I agree with the preparation part, I think persistence is needed more than just waiting. If you don’t try to open it every once in a while, the window may stay shut. But persistence also doesn’t mean that you should try the exact same thing over and over again. Saul D. Alinsky, who was a community organizer and wrote the very cool — although only partly applicable for us — book “Rules for Radicals: A Pragmatic Primer for Realistic Radicals,” said that a tactic that drags on too long becomes a drag. So you have to switch up your approach. Read the reactions to each one — don’t provoke them for the reaction’s sake — and learn from them. And don’t give up. Delay only with a fixed return date.

Tactics V: Downward

I mentioned before that, as fancy as upward “rebellion” may be, it is even more important to enable your people to come up with new ideas and better ways that go against the majority. The most important aspect is certainly providing what Amy Edmondson calls “psychological safety,” by permitting and encouraging different opinions — yes, also ones different from your own. When I am wearing my manager hat, I love it when people working for me argue against me. It brings something new to the table: more information, more content, better solutions. It just helps everybody. That also means you have to admit when you’re wrong. This is part of the game, so to speak. If you cannot do that, nobody will discuss anything with you — why should they? But more than that, you will also have to secure this right of your people toward the rest of the organization. When one of your employees expresses her opinion to another manager, back her right to speak up. You will also have to ensure information sharing, because people can only form opinions if they are in on things — as said before, more on that soon. You’ll have to teach them methods: all the things we discussed here and more. Teaching and learning never stop. And eventually, of course, it all starts with hiring and promotion. Look for people who are open, who can form an opinion, and please hire people who are not a copy of yourself.

Ethics check

That brings us to the safety net — at least the internal one: ethics. Sounds grand, but we have discussed stuff like that before, and I try to keep it quite simple. First, we have a settled formula: we want to go with the organization’s values and purpose. Our goals are concrete and inspiring, and we always check them with PDSA. Our way or methods are flexible but stay within limits. These limits are clear and uncompromising: we follow the truth — no lying; all people, especially all members of the organization, have to be treated with respect — no humiliation, no destroying of people; and their right to speak is always guaranteed. This is where we follow Camus and are strict about our means but approximate about ends — meaning the goal is concrete but corrected with PDSA as you learn. Yeah, and also Kant with his “klug wie die Schlangen … ohne Falsch wie die Tauben.” But that comes from Matthew, and I don’t like to quote the Bible.

Yet reading rules is simple; more interesting are the times when you have to test yourself and your actions against a principle. This is, of course, when we go to Kant again — this is his area of expertise. The first test has to be the universalizability of your actions. In this case that would be: what if every middle manager did this? So, for example, what about hiding bad numbers to protect your team? Well, that would fail, because if everybody did that, no number would be correct anymore. What about building your own KPI set with your own targets bottom-up and reporting them honestly? That would be fine. It might lead to a lot of different numbers if all parts of the organization did that, but more numbers also mean more information. Running down bad decisions while implementing them would not pass the test, because if everybody did that, it would become very hard or nearly impossible to implement anything in an organization. But stating your honest opinion passes the test. Everybody needs to have the right to express his or her opinion. So you see how that works, and it should always be the first question you ask yourself.

And Kant even has a second test that is also important: the publicity test, which consists of two formulas that give us three cases. (1) If something only works when the action is hidden, it is wrong. This would cover all cases of manipulation, passing on confidential information and so forth. (2) If an action could be open but doesn’t need to be, it is permissible. So discretion and some backstage work are allowed, if not absolutely required. (3) If the action needs to be open and public to work, that is the positive case, and it is of course the right thing to do. So working in formal channels and presenting your alternative openly is the way to go. Be aware that part one alone does not prove that you are right; it is a negative test. Failing it proves an action wrong, but passing it does not prove anything — you need part three for that.

I also want to point out here that effectiveness does not count as an ethics test — that should be clear, but I want to mention it. Just because an action achieves its intended result doesn’t mean that it’s okay.

Our methods or means also matter in a different respect. They actually define our ends; they create what we are. This goes back to Camus, who said — and I am paraphrasing — that you can’t reach a good end through bad means, because the means become part of what you build. For me that means a change pushed through with manipulation produces an organization that runs on manipulation. Or, as my grandfather, who had read Gandhi, taught me at a young age: you cannot fight war with violence, Punkt. This actually carries over to business. Participation and cooperation — you know I am a fan of those by now — cannot be ordered. They have to grow out of supportive relationships, not instructions. So it isn’t only ethics. The wrong means fail on their own terms. If your goal is more cooperation, coercing it destroys the very cooperation you want.

Kant has another image that serves us well: the moral politician vs. the political moralist, in “Toward Perpetual Peace.” The moral politician uses prudence within his moral limits. The political moralist bends morality to fit his advantage. In business that would be the manager who writes the value statement after already having made the decision. So the self-check for us as guerrilla managers would be: do your limits shape your tactics, or do your tactics produce your limits? The first is good, the second bad — just in case you were wondering.

Eventually this discussion leads us to a structure of ethically sound action. You argue before you refuse. You refuse before you leave, which is your last option. Start as Kant’s officer: do your job and argue. But doing your job does not require you to pretend that a bad decision is a good one. Nor does obedience absolve you when the act itself crosses the line. At some point you may have to refuse to become an instrument — and, ultimately, walk away. That is, of course, easier to write and considerably harder to do. The middle manager with a mortgage, two children and twenty years invested in a career has a lot to lose. Ethics should not pretend otherwise. Plus, there is another argument against the ultimate step: you will no longer be able to change anything in that particular organization. It is a hard call, but sometimes it might be better to accept that you will lose some confrontations but still have the opportunity to win others.

Darkness

After all these tactics and ethical checks, let me share some points that are a bit more personal. I know a lot of people who have just stopped arguing and pushing for better alternatives. Far more than were fired or resigned. And it is usually a slow process of erosion: too many times you lose an argument, then private life cuts in, and the baby is more important than whether the KPI set actually reflects reality or whether we are empowering employees enough — I mean, you run on five hours of sleep, so what the heck. But there is also chosen conformity, when the next career step becomes more important than the right thing to do. And not speaking up is often better in that regard. But it was Camus who said: “Si le révolté ne choisit pas, il choisit le silence et l’esclavage d’autrui.” — “If the rebel makes no choice, he chooses the silence and slavery of others.” So it is not just about myself. What I choose affects others. And I hope I have made it clear that doing nothing does not produce anything positive, especially not for the organization; actually, it will eventually make it less able to change and adapt.

Why people stop, or never start, going against the stream is an important topic for me. In a past essay I told you about intellectual laziness — not stupidity, but the simple failure to use a capacity we already have: our ability to reason. And since — even in the bad case — I like to think the best of people, it is still the main reason I see. Kant pairs laziness with cowardice — Faulheit und Feigheit — which is of course more damning. “Es ist so bequem, unmündig zu sein.” However you look at it, the reasons may be structural, but the responsibility remains individual.

The risk is real. People get fired. And the system usually punishes the rebel first. All your work of improvement will mostly go uncredited. Your team might not want the freedom you give it. Middle managers get exhausted by yet another role. And you always run the risk of failing an ethical test. Do I want to rebel, or do I resent? Am I doing this to make a change for the better, or do I just want to punish people?

I see people stopping, but I also see people continuing. It takes a lot of strength, because it is not a nice place to be; nobody wants to swim against the stream. It is mostly a place of isolation, which is not easy to take. We are social animals, after all. We need affirmation and connection. Without them we end up without energy, burned out, in a dark place.

And what about me? Well, I keep arguing internally, and I have the luxury of writing. I write about all the absurdity I see, but I get enough input from friends and colleagues to ensure general applicability — you know, Gummy Bear Inc. Which is exactly what Kant meant when he said you can always critique in public. This is my right, and that of anybody else, at least since the Enlightenment. And that’s exactly what it does — at least a bit — it lightens up my dark place.

Light

But I don’t want to end this all dark and gloomy. Middle managers have possibilities, and as promised in the intro, it’s not about being Che Guevara. We want lasting change, and that is often found through moderation. Camus said that moderation is not the opposite of rebellion; rebellion is moderation. Tempered is the real thing. See, the yes-sayers are the ones who stop or never start to go for change. The no-sayers are the cynics, refusing everything. You want to walk the fine line of having both the yes and the no. Then you can achieve something. Not for money or glory — forget about that. But you can achieve something better than what is. And even more important, you can make room for the next ones: by example, by enabling, by creating a safe and nourishing environment for new ideas, different ideas.

So, to summarize very briefly — if you got this far: the middle is uniquely positioned to drive change, because it connects the base of the organization to the top, the particular to the general. Use the room that you have, act within your authority, create small wins and build on them, build alternatives, use allies and formal channels, and hold the organization to its stated purpose and values. But never forget that means do matter. Truth and respect: no lying, no manipulation, no sabotage or sacrificing of people, because the methods you use are part of the organization you are trying to build. And please keep arguing. Obey where you must, refuse where you ethically cannot, but don’t choose comfortable silence.

And if there is ever a moment when you feel you cannot do anything, at least pick up a book. It’s the first step, not the last, because, in Kant’s spirit, a book is not a replacement for thinking.

“Sapere aude! Habe Mut, dich deines eigenen Verstandes zu bedienen!”

Don’t just take it from me, here is some good stuff to read:

Meyerson, Debra E. Rocking the Boat: How Tempered Radicals Effect Change Without Making Trouble. Boston: Harvard Business Press, 2008.

Townsend, Robert. Up the Organization: How to Stop the Corporation from Stifling People and Strangling Profits. New York: Knopf, 1970.

Camus, Albert. The Rebel: An Essay on Man in Revolt. Translated by Anthony Bower. London: Penguin Modern Classics. (Original: L’Homme révolté. Paris: Gallimard, 1951.)

Kant, Immanuel. “Beantwortung der Frage: Was ist Aufklärung?” Berlinische Monatsschrift 4 (December 1784): 481–494.

Kant, Immanuel. Zum ewigen Frieden: Ein philosophischer Entwurf. Königsberg: Nicolovius, 1795.

Alinsky, Saul D. Rules for Radicals: A Pragmatic Primer for Realistic Radicals. New York: Random House, 1971.

Kotter, John P. Leading Change. Boston: Harvard Business School Press, 1996.

Jackall, Robert. Moral Mazes: The World of Corporate Managers. New York: Oxford University Press, 1988.

Wrapp, H. Edward. “Good Managers Don’t Make Policy Decisions.” Harvard Business Review 45, no. 5 (September–October 1967): 91–99.

Wageman, Ruth, Colin M. Fisher, and J. Richard Hackman. “Leading Teams When the Time Is Right: Finding the Best Moments to Act.” Organizational Dynamics 38, no. 3 (2009): 192–203.

Edmondson, Amy C. “Psychological Safety and Learning Behavior in Work Teams.” Administrative Science Quarterly 44, no. 2 (June 1999): 350–383.

Aristotle. Nicomachean Ethics.

Or

How to become one

So you want to become a good manager and have stopped scrolling. Good. I will tell you, but you have to read the whole essay. Promise? I know it’s 4,500 words, but some effort is required. And that’s only the first step. But reading is an important one. Like the first time I read something by J. Richard Hackman 20 years ago, when I was a rookie at leading teams. I had totally forgotten about the guy until I re-read it recently, but his lessons have stayed with me ever since. He was a brilliant social and organizational psychologist working at Harvard. Sadly, he has since died, but his book “Leading Teams” is very much alive.

Just to be clear, I don’t think I am a good example of a Good Manager — everybody can be a good manager, by the way; it’s learnable. The reason is that I only barely fit the definition of what I mean by manager here. The manager — let’s call him Tom — in this essay is the one with authority to shape conditions: one of the top guys, or the head of a part of the whole thing with real room to act. The manager without that room goes into the next essay — spoiler: it will be called Managing out of the Middle. I have always fit more into that second category. And if you are in the same boat, don’t despair; we have possibilities too. You will see.

To get an even clearer picture of what a manager is, we have to look at Hackman’s authority matrix. There are four functions when teams and managers work together: (1) executing the team task, (2) monitoring and managing process and progress, (3) designing the team and its organization, and (4) setting overall direction. If you now look at how many of these functions Tom does himself, you know what kind of a manager he is or what kind of team he has. If he is doing everything but the execution of the work, he has a manager-led team — that is not micro-managing, by the way, that would be even worse. If he leaves the execution and the monitoring to the team but does the designing and the direction setting, his team becomes self-managing. If the team also takes over design, it becomes a self-designing team. This could go on until Tom wouldn’t have to do anything and the team becomes self-governing. We have seen examples of this in past essays, but they are rare. Some communities described by Ostrom work like this, or Laloux’s self-organizing companies from the last essay.

In this essay we will look at the most interesting and fruitful forms of managers and teams. But first, to show you the diversity among the authors in this essay: they have different views on hierarchy and organizational structure. We will hear from Elliott Jaques and his “In Praise of Hierarchy” — the name of the paper matches his thesis, but I only refer to him for some specific points, not his general thesis. On the other side of the aisle is Isaac Getz, a business professor and a strong advocate of what he calls “Liberating Leadership,” by which he means that leadership needs to liberate organizations, teams and employees from the shackles of hierarchy. And Frédéric Laloux we already know for studying examples of self-organizing companies. Hackman is somewhere in the middle, carefully analyzing different forms. He is very critical of organizations that are too free or too self-organized, but at the same time he sees the problem of too much hierarchy and bureaucracy. Karl E. Weick, an organizational psychologist, is also in the middle with his analysis of loosely coupled systems. The collector of studies and numbers is Rensis Likert, a social psychologist, with his book “New Patterns of Management” from 1961. He will provide us with many studies that support what he calls participative management. And to complete the cast: Mary Parker Follett will provide some of the theoretical concepts we need.

But why all this discussion? First, I refer to my essay “Josef KPI,” where we saw that the manager is a chaotic, overloaded, unfinished profession. It is still very much in the process of becoming an adult version of a job, so we need to talk about it.

Usually, and correctly, we look at purpose or direction when we search for guidance in organizations, but purpose can be turned into work in many different ways. Somebody can tell you — hierarchically — or you can do it by yourself — self-managed — or in a team — cooperatively. What is always true is that organizations themselves don’t do anything; they don’t decide or set purpose by themselves. That is always done by people. Which also means that structure is immensely important, because the design of an organization determines who takes part, and how, in doing, deciding and setting purpose. The manager will then have a major role in controlling or enabling that structure — depending on what that structure actually is.

Another issue is that this controlling is often wrongly taken as a final authority over everything, which Follett already showed us it is not. Decisions by management are based on the people below — who defined the problem, picked the data, framed the options, controlled the dashboard. The top approves a version of reality built by others.

That doesn’t mean that authority is an illusion. It exists and is needed in organizations. It just follows function and knowledge rather than rank. That’s what Follett calls functional authority. The more honest version is that the manager doesn’t simply decide, but finds out whose knowledge carries authority in a certain situation.

We also have to rethink management because Likert can show us that superiors often misjudge what drives the people below them — money and easy work are usually overstated, while interesting work and good relations with colleagues are underrated. At the same time, managers often overestimate their own practices — praise and recognition or listening skills.

And there are many reasons to doubt classical command-and-control management theories. My main point is what I call the 1,000 brains against 10. If you only took your top ten managers for any task, they would have to be a hundred times better than a thousand employees to match their ideas, knowledge and just general intellectual capability. Even the selection and filtering of all the ideas coming from a thousand people is better done by people close to the matter and not by the 1% at the top. This is just my own view but the math holds.

Another point that we discussed in the last essay is that of information availability. The manager usually has the complete list, the budget and the plan. Jaques ties this to the greater mental capacity of the people becoming managers. This is of course a confounded claim, since as long as all information stays at the top, you can’t tell whether it is about capacity or about actually having access to that information.

What is also problematic about hierarchical managers is that they get systematically overloaded. See, in vertical structures most things — information, decisions, etc. — run through the top, which again justifies additional layers in the organization to distribute a workload that was mainly created by the structure. You don’t need an additional sub-boss; you just need people to talk directly to each other.

And what we certainly need to clear up is what coordination means: the hub vs. builder issue. Our manager Tom can distribute every task to his team members. Then he is a hub, and he is basically the only thing the group shares — Hackman, for example, calls this a co-acting group. But this is not really a team, right? It’s just a bunch of people having the same boss. A builder will form a group that can coordinate itself. For Hackman, this is real teamwork. Follett describes this as coordination by creating direct relationships.

More important is the question of accountability and responsibility. Jaques has a point here that clear individual accountability is needed, which means a clear hierarchical structure and a clear manager role. He is making the decisions and he is accountable for them. That seems to be the most efficient way of handling it. But is it also true and effective? Remember the many hands problem from Dennis Thompson. He said that pushing all responsibility to the top wrongly removes accountability from everybody below who also took part in the decision, as we have seen in Follett. It makes people complacent, which makes the organization less effective in achieving its purpose. As Thompson argues, responsibility has to be traced person by person, in degrees. So my working rule is something like: each person answers for what they could reasonably know and could influence, no more and no less. And this does not contradict our belief that no one in an organization is without responsibility. Also not the frontliner, the worker, the cleaning personnel — tough luck. So a street sweeper also has a certain accountability that goes beyond cleaning up dirt. If she or he, for example, sees a dangerous number of potholes in the street, it is their responsibility to report it to somebody.

This brings up another problem. What if management pushes down responsibility and keeps all the decision-making at the top? The general rule still holds, but this way of managing will be perceived as unfair, because responsibility is not defined by organizational structure. Follett points out that the way that works is to keep authority — to make decisions, for example — and responsibility together at the same level. This means for our manager Tom that his job should include understanding the whole, listening to issues raised, routing them to the right people, and protecting people from overload by not making them responsible for things they cannot influence.

This brings up another important point, which is dignity as the real boundary and guiding principle for how to treat people in order to achieve the desired results. Kant said — allow me to paraphrase — that people may be means, but never merely means. Treating people as equals is also what produces the extra ideas and effort, so the two are linked.

Two things are left to say before we look at the manager’s actual tasks. One is that leadership is more a function than a person or position. Hackman found that as long as it produces the right conditions, you can never have enough leadership. Laloux, meanwhile, has examples showing that even in very flat organizations roles are not identical. There are always leaders — in many variations of course.

The last point is that decades of research found traits to be only modest predictors of good leadership, and they predict who gets chosen more than who leads well. Management is a craft. People just start at different points.

For the tasks, we will generally follow Hackman’s approach and see what it takes to create a real team, compelling direction, enabling structure, supportive context and coaching. And we will mix in some good stuff from other people along the way.

First off, you need to create a real team or teams. According to Hackman, real teams are defined by four features: a team task where people depend on each other for a shared outcome, clear boundaries — meaning who is a member and shares accountability — explicit authority over the work process, and stable membership over a reasonable period.

See, we already mentioned that some teams are merely co-acting groups sharing only the boss. That’s not a team, and Tom will not leverage any of the team superpowers like combined knowledge, experience and workforce. So Tom has to turn these groups into teams by giving them a shared outcome. For example, all customer requests for a certain area, not individual tickets under the same boss.

But it doesn’t stop there. Even better than just a shared outcome is to organize teams around processes and customers. That is quite a far-reaching thing to do, since a lot of organizations have formed teams around functions — operations, customer service, sales, finance. We know that organizations work along processes; this is where value is created. Management system standards also tell us this, but org charts have still stayed functional — probably because historically specialist know-how was thought to be the leading factor. But hey, history is a shitty excuse for bad structure. In the end we want to produce great products and services, not great specialist departments. In some fields, specialists are highly educated and few. That means they may have to be shared between teams. But it is also true that some companies have developed innovation without much R&D, just by enabling working teams to solve problems innovatively.

Now, generally, this approach will lead to a certain number of parallel teams, which is fine for a lot of simple production or service companies. Laloux found them in a lot of the examples he looked at.

There are also more complex forms of non-hierarchical organization, like the web of individual contracts: yearly negotiated one-on-one contracts between colleagues upstream and downstream, some written down, some rather informal. This continuous process of committing results in a very fluid organization. Nice, but probably too much for most organizations.

Nested teams are another possibility. As with parallel teams, they are mostly autonomous, but they are also part of a nested structure. So team A, team B and team C, together with person 1 and person 2, make up a bigger team while still being their own teams. This is a way to incorporate some special functions into a team form. This approach is mainly used by organizations with more complex creation processes.

In any case, teams always need to be connected in some way. Likert proposes that teams should overlap through a linking pin. Such a connection can be the manager who is a member of two teams. He leads one and sits in the one above — the group of managers.

If we talk teams and connections, then we have to mention Weick’s principle of loose coupling, because it tells us what options are available. Managers and teams, or teams on their own, can have tight coupling or loose coupling, meaning they work very closely together or only loosely. Obvious, right? And that this is a spectrum should also be clear. Weick argues that tight coupling works best in organizations with people working together on the same activities — a real team, as seen above. That team is then loosely connected to other teams. That is because attention is finite: you cannot have tight coupling with everybody, so tight coupling in one place means loose coupling elsewhere. This also goes for managers. The manager has to choose the few things to hold tight — for example, his own team of supervisors.

Finally, and not to be forgotten, not everything is teamwork. Some tasks are better done individually, like writing the first draft of a novel or a report. But the manager should decide this deliberately, to avoid frustrated team efforts leading nowhere.

By the way, management teams — the big guys — also need a real shared task and outcome. That can be — honestly admitted — just information sharing, which is okay, as long as that team and its activities are set up for exactly that and nobody tries to fill it with more fantasy content.

Stability of teams is the easiest to understand but often very hard to guarantee. Teams need time to form and perform well. Rotation every two years destroys this. Likert’s studies point the same way, and I have witnessed it firsthand many times.

After the teams comes their direction. As with the whole organization, purpose and how to get there are what bring teams together. Now, we are not going through the whole shebang of vision and values and mission — it is overrated in my opinion anyway and often confuses more than it helps. What is needed is a purpose concrete enough that people can decide by it. “Saving a buck” might be great for a budget airline and give its employees a guideline, but if you are a life science group, something like “For the health of the world” is too general. This is a slogan. It might inspire, but it is of no help when employees need to decide business matters. Plus, should you have multiple purposes — same as with stakeholders — they need to be clearly prioritized: great service first, cheap service second.

Specifically for team direction, Hackman names three qualities. Direction needs to be challenging, which gives teams energy, but within the limits of the possible. It needs to be clear — as just explained. And it needs to be consequential, which means that it matters to customers, colleagues or society, so it engages teams.

Even more important is what Hackman connects to direction in his view of authority: how ends and means are distributed to teams. If a team is given neither ends nor means, that is anarchy. The team has no rules and no direction — that makes for an interesting experiment but a bad organization. If the team is given both means and ends, it ends up just executing, and its abilities are wasted. The worst case is being given means but no ends. The team has all the rules but no purpose. This is where much bureaucracy — and bad companies — end up. He recommends giving ends but no means. Rules should be reduced to a minimum — you know, safety and legal and so on — and direction should be clear. This is a key condition for a self-managing team. By the way, Kant would recognize this: giving ends, not means, treats people as capable of reasoning their own path. And respect if you notice something there: you are a true Kant-nerd. 

And for that, teams also need information. Within an organization, information should be shared generously and smartly. We talked about it before, and I will bring it up many more times because it is needed. For people living in the Information Age, we do a piss-poor job of managing it.

Purpose and direction are also closely connected with membership of an organization — or employment, if you want to call it that. Meaning new — and of course also existing — employees always need to be made aware of it. Now, this doesn’t mean a lot if it is very hard for employees to change organizations when they don’t agree with the purpose. Getz found an example of a company that made it a statement that everybody was always free to leave, and backed this by helping people start their own companies. This doesn’t solve the complete issue, because of course the economic reasons go beyond just needing startup help. Yet I really think this is a central point; otherwise purpose can become something we endure to have a job, rather than a guiding star.

Purpose almost always comes from the very top of an organization. Only an absolutely self-governing organization would make it a common task. Even — and especially — most NGOs are started by a person or a small group setting a purpose. It comes first, so it sits at the governance level of an organization.

OK, we have teams, we have direction; now Tom needs to enable the teams by creating the right kind of structures. You know, the whole who may decide what, and where.

Authority always needs to be set explicitly from the beginning — be honest and clear, and avoid a lot of inefficient meeting discussions. The guiding principle is the purpose of a specific team. This can also be found in Jaques, who correctly connects it to the amount of accountability a person or team should have.

Changes to this clear guidance can be made later on — usually it is a widening of authority once teams have become fully performing, but it can also be the reverse. Getz actually has a checklist for authority: willing — does the team share the purpose, and is it treated as equal? — capable — does it have the skills? — and authorized — is it allowed to act?

Once equipped with authority, managers and teams can make decisions. The manager primarily has the role of the “meeting door opener”: to get the team in there and stay out of the content. But since there are multiple teams, there are other meeting rooms. He is the linking pin between them.

The forum always matches the matter. Each team handles its own level’s problems.

Special procedures should be agreed on for emergencies. A crisis doesn’t need new structures — actually, it is the worst time to use something new or different. What you need is a clear plan for how to act.

All of this is set up by a manager. A team, on the other hand, can now deal with the rules and the content. Ground rules should always be set and maintained by the team. That is a prerequisite for acceptance. And content, yes, that is the real deal; that is exactly what we want them to do. Even the hard stuff, like setting a budget. Laloux has a couple of good examples showing that this is possible. This avoids inefficient haggling between bosses and teams. But what about wasteful spending? Peer pressure is a much better way to handle that in an organization than trying to negotiate budgets as if they were a selling price. And if the total of a budget is over a reasonable limit, it just goes back to the teams’ representatives to sort out.

Once Tom has set up the basic enabling structure, he has to enhance it with a supportive context. In plain English: Tom has to create freedom.

The most important part of this is competence. Qualify everyone and ensure they are ready for as many activities as possible in advance. This means a lot of training and a lot of thinking ahead from management. But planning is better than reacting — as in having to give people individual permission case by case.

The next one is important, and most managers are not going to like it, so I’ll make it a command myself: remove any controls that express distrust, like time clocks and locked supply closets and reserved parking spaces — you do have to ensure that there are enough spaces, though. Keep only the functions that have a direct connection to the purpose of your company. So you also immediately have a test at hand: does this mechanism or process serve the purpose or the distrust? Does quality control try to find product failures or badly working employees? Does the travel policy protect employees or the environment, or does it make adults ask permission to take the train? Legally required functions stay too, but they can be run in a trusting way.

Now, freedom does not mean no rules, no standards, no methods. Having worked a lot with best practices and standards, I know that they have a real benefit. They give us examples of how to do it well, and they can help keep quality at a certain level. For example, a management system is a framework for organizations to work well and consistently. Nothing wrong with that. We should always plan our activities and, once we have executed them, check and improve them — ha, PDCA. But I think we should get away from thinking of this as a separate function within an organization. Every team should do it. Let’s not waste time on all the controls, and instead enable teams to improve.

The same goes for KPIs and measuring. They are a great tool for learning and improving. They are really bad for controlling people. Just read Deming for that. This is his big thing, not mine, and he was a statistics nerd. Teams have to find out what kind of data they need to learn, to become better and to avoid mistakes. The manager still has a certain assurance role here, because there is an overall purpose of the organization that also needs to be achieved — and thus also measured for improvement. But managers have to take this as a leading function, not a controlling one.

Now, freedom has its costs, certainly. More training, more enablement, more workshops, more measuring. And that cost shows up in a budget. But let’s not forget that controls also have running costs; they are just usually hidden within pre-existing hierarchical structures. So far, in a structural sense, I have never seen a study comparing them like for like.

The manager will also have to link a team upwards. That usually means ensuring information flows upwards and downwards, and representing the team upwards. That seems obvious, but without influence upwards, good team practice will backfire.

One last point is how people are paid, or should be paid. I have my opinion on this, and of course some of the authors mentioned have too. But in order to only partly set off a glitter bomb, I will discuss this in a future essay about motivation.

Finally, what our manager Tom will always have to do is coaching. One caveat, always, is that design comes first and coaching second, because no amount of coaching can compensate for a badly designed or badly set-up team.

Ruth Wageman — as cited by Hackman — studied the effect of coaching on different types of teams. Well-designed teams self-manage even with little coaching, although they are also the kind of team that profits most from good coaching, and bad coaching barely hurts them. But bad coaching hurts poorly designed teams most. So good design and setup of teams in an organization partly protects them against a weak manager.

Hackman identified three functions that coaching can work on: effort, performance strategy, and knowledge and skill. Coaching on effort usually comes at the beginning of the team’s life cycle and is mainly motivational. Coaching on performance strategy comes at the midpoint and is consultative. Towards the end of the cycle come knowledge and skill, and this is educational coaching.

Before coaching, you should always check the design conditions for each type of coaching. For effort, that is energizing direction, task design and the reward system. For performance strategy: focused direction, norms of conduct and information systems. And for knowledge and skill: talent-engaging direction, team composition and the educational system. If these are not properly set up, no amount of coaching will help.

Timing is crucial: performance coaching at the beginning will just dissolve into thin air or even do harm. This is also the case if a manager is called in late and was not involved in either the design of the team or the first direction setting. It is then best to check the design of the team and then relaunch the effort.

Hackman also states that some studies have found that coaching on the work is more beneficial than any kind of interpersonal intervention. An experiment conducted by social psychologist Barry Staw even suggests that performance shapes how people describe their relationships, rather than the other way round.

Usually the manager is in charge of coaching, and it is one of the most important tasks, if not the most important. But coaching can also be shared. Peers coaching each other is used successfully by many companies. Obviously, it is then the manager’s responsibility to train the team to coach itself.

So that’s it? A manager’s job is to create real teams, give compelling direction, provide an enabling structure and a supportive context, plus coach? See, after reading 4,500 words, you know it. The problem, of course, is actually doing it. As promised, we will cover learning and knowledge in a future essay, but spoiler: it has to do with trying things out — which takes some bravery — and learning — which takes honesty and the humility to admit that you make mistakes. So to become a good manager, you also have to learn those two things. And that will take time. As with transforming an organization, the role of the manager will change slowly. Eventually, if this approach is followed, the role will become more stable, less chaotic and less overburdened. There will be setbacks; old habits are hard to shake off. 20 years after reading Hackman for the first time, I still catch myself playing the boss sometimes.

But eventually this new kind of manager will make the team able to manage itself. Or, as the former CEO of a French automobile parts maker put it, quoted by Getz: “Creating conditions in which things happen by themselves.”

Don’t just take it from me, here is some good stuff to read:

Hackman, J. Richard. Leading Teams: Setting the Stage for Great Performances. Boston: Harvard Business School Press, 2002.

Likert, Rensis. New Patterns of Management. New York: McGraw-Hill, 1961.

Jaques, Elliott. “In Praise of Hierarchy.” Harvard Business Review 68, no. 1 (January–February 1990): 127–133.

Getz, Isaac. “Liberating Leadership: How the Initiative-Freeing Radical Organizational Form Has Been Successfully Adopted.” California Management Review 51, no. 4 (Summer 2009): 32–58.

Laloux, Frédéric. Reinventing Organizations: A Guide to Creating Organizations Inspired by the Next Stage of Human Consciousness. Brussels: Nelson Parker, 2014.

Weick, Karl E. “Educational Organizations as Loosely Coupled Systems.” Administrative Science Quarterly 21, no. 1 (March 1976): 1–19.

Follett, Mary Parker. “The Illusion of Final Authority.” In Dynamic Administration: The Collected Papers of Mary Parker Follett, 2nd ed., edited by Elliot M. Fox and L. Urwick, 117–131. London: Pitman, 1973.

Thompson, Dennis F. “The Moral Responsibility of Public Officials: The Problem of Many Hands.” American Political Science Review 74, no. 4 (December 1980): 905–916.

Or

How we should talk with each other

This was not an encouraging opening for a conversation. Alice replied, rather shyly, “I—I hardly know, sir, just at present—at least I know who I was when I got up this morning, but I think I must have been changed several times since then.”

What changed in the last essay was that hierarchy is not the only — always right — answer to how organizations can work. We can also work together. And we looked at different ways this is possible. But it was all about the setup, the structure and what is needed to make the setup work. What we haven’t looked at is people actually doing it. That means first and foremost people engaging with each other. And most engagement at least starts with communication. Secondly, doing cooperation means using all the mechanisms, the design principles Ostrom and Benkler have explained to us. Now using these requires individual people to reason. Individual reasoning only becomes useful for cooperation if it is transformed into information being shared with others. Again, communication or — because it is the most apparent manifestation of communication — talking with each other. So how we should talk with each other is the topic of this essay.

OK, easy, we all know that one. Especially all the managers will go, yeah, I can do that, I talk all day, no problemo. The issue is we use this skill that most of us have a basic understanding of very casually and without real knowledge of its depth. Frederic Laloux in his 2014 book “Reinventing Organizations: A Guide to Creating Organizations Inspired by the Next Stage in Human Consciousness” looks at quite a lot of it. Things like cooperative decision-making, constructive conflict resolution, and a safe environment are things he found in studying next-level organizations. Now, Laloux, although not a classical academic — he is a former McKinsey associate partner and executive coach — is a brilliant collector of good practices. He can in his reflections sound a bit new-ageish sometimes, but we’ll stick to his concrete, useful observations rather than his more new-age framing, and that starts a bit further down. One of the best things he found was the minute of silence that is a custom at the beginning of meetings in some companies. Not a form of spiritual meditation but just a practical way to have a fresh start to a discussion with the help of a one-minute absence of discussion. Works every time, believe me.

Let me start exploring “good” communication in organizations with my favorite early management theorist Mary Parker Follett. In the collection of essays “Dynamic Administration,” posthumously published in 1941, she has this beautiful sequence: Conflict → Integration → Giving of Orders → Power-with → Coordination. For all Follett fanboys: I know Integration is not in the “official” sequence, but damn, it is the coolest takeaway from the whole thing, so it needs to be named.

Follett knew already a hundred years ago that conflict is not the problem. And I hope we today are all aware of that. Differences in opinion are to be expected and should not be suppressed, because they can be potentially productive. She distinguishes three ways of dealing with them. Domination is when one person gets what they want and the other person just gives in or is forced to do so. Compromise — and since I am Swiss, I am the expert on this one because our whole political system is basically built on this — is when both parties surrender something in order to create an agreement. Both domination and compromise are basically loss situations. Either it is win-loss or loss-loss. In a cooperative organization this is not a favorable way to progress. That is why Follett came up with a third option. Integration is finding a solution in which the underlying interests of both parties can be satisfied. She has this famous example of her working in the university library. She wants the window to be closed so there is no draft. Another person wants it open to get some fresh air. The practical solution is to open another window further away, allowing air to come in but far enough not to produce a draft. This is of course only possible if the two people talk to each other and find out what the other’s interest is. Only this allows them to find a solution that satisfies both interests. So communication is a prerequisite for integration. Follett thus described what we would today call a win-win situation, something that has often been misused since — you know, by just adding it to your preferred solution without going through the whole integration work.

Follett then goes on to the situation in organizations when a boss would usually order a subordinate to do something because “I say so.” Since we have discovered that we can talk with each other, we should try to discover what the situation requires, what’s best according to the circumstances. She calls this the “law of the situation.” So it is basically the situation that decides or gives the order, not the boss. But situations of course cannot really decide, so it is rather the knowledge of the situation. Knowledge is not just in the hand — or head, or heart — of one person. So the boss might know something the subordinate does not and the subordinate might know things the boss doesn’t. So only by taking their knowledge together can they make the best decision. Authority becomes tied to knowledge, not position — or it becomes functional.

Now it gets even cooler, because this means that the traditional conception of power — you know, the whole hierarchy thing — can be overhauled. Because if people do come up with stuff together, the traditional model of the boss making the subordinate do something — A → B — will become two people together enabling the act of doing something — A + B → capacity to do stuff. She describes this as not power-over — the hierarchy — but power-with — the collaboration. So two people together understand the situation, integrate their different knowledge and determine what the situation requires. And nobody needs to dominate the other.

Cool, isn’t it? But the caveat is that this only describes one individual interaction. Organizations contain hundreds or thousands of these activities and most of them are somehow connected to each other. This is where Follett describes how organizational coordination works in four easy principles. First is coordination by direct contact, meaning people whose work affects each other should communicate directly — horizontally — rather than everything having to go through the vertical channels of hierarchy, because that just takes too much time and is too complicated. Second is coordination in the early stages, which means that coordination or communication should start from the beginning, while plans are being developed, before everything has already been decided and basically shit has already happened, and untangling it would be complicated again. Third is coordination as reciprocal relating — OK, what? This just means that causality in organizations is circular, not simply linear. Or if that doesn’t help: The work of A affects B, B affects A. C affects both and they affect C. And don’t get me started with D. It is a concept that should be logical, but we still think too much in those damn linear process diagrams, which have a start and an end. But that is just an incomplete fragment. Activities in organizations look more like circles. We will talk more about this when we look at Systems Thinking. Lastly, the principle of coordination as a continuous process states that coordination is not a meeting or a management workshop; it is happening all the time everywhere in the organization. How this coordination changes the practical work of managers we will look at in the next essay. With this sequence — Conflict → Integration → Giving of Orders → Power-with → Coordination — Follett gave us the backbone for how interaction and communication in cooperative organizations can work.

Now Follett is so important to this whole discussion thing — heck, to modern management theory as a whole — that I want to add some detail points and tidbits — and yeah, I am kind of a fanboy of hers, I admit. First, a really good real-life example for integration: the Smith/Panotti case. Smith sued Panotti, who was a small fruit dealer, for a 40-dollar debt. Having to pay the whole debt would have put Panotti out of business — remember, this is a hundred years ago. So the judge ordered a repayment in installments from Panotti’s earnings, allowing both Smith to recover his money and Panotti to remain in business. A perfect example of how opposing interests do not need to be incompatible.

When it comes to gathering information on the situation — remember the law of the situation — cooperative fact-finding becomes really important. Pre-existing knowledge is fine, but as we have seen in an earlier essay, gathering information in a joint process resolves real disputes.

Also important to Follett is the distinction between muddle and difference. The latter is a real difference in interests, needs, perspectives or purposes between people. Muddle, on the other hand, is confusion created when apparent differences have not been clearly identified and separated from assumptions, misunderstandings, emotions or poorly defined issues. Muddle can be cleared up in order to make the real differences visible and concrete. This should happen before integration is started. Remember the example that we have to first be clear whether what we see is a snake or a fallen branch on the ground before us.

A bit more complicated is the concept of X and X’. That is Follett pointing out that when we interact with each other we change. And every subsequent interaction is with these changed participants — and in a discussion that means mainly the opinions that have changed. So, in my shorthand, it is not A → B → A → B but A + B → A′ + B′. Now this gets really complicated when it is not just A and B but also C, D and E — and therefore also C′, D′ and E′. What can happen — and this needs to be avoided — is that people refer back to an opinion that has already changed, therefore hindering the progress of the discussion and just confusing everybody in the room.

There is also what I would call the mis-scoped question. This is important because Follett found that in discussion — and she was sort of a professional mediator, especially between employers and unions — people often do not argue about the exact same point; they are mis-scoping the discussion topic, which of course leads to misunderstandings.

Another issue is that it is difficult to impose standards and data on all the participants in a discussion if these were handed down from the top. How can we possibly think that things like that just get accepted? Why should they? They first have to be integrated into the reasoning of all participants — or even better, collected and created cooperatively, but that is not always possible. OK, example needed, I get it. KPIs and targets, for example. People need to understand what they measure and why this is important and why a certain target value is worth achieving. Or even harder: a certain process standard. We all have them and we kinda think they are important. But you cannot just post them on a wall. Everybody involved needs to understand why it is important to do something the way it is described. There might be quality, safety or environmental reasons.

Eventually it comes down to two things. First is from Follett and is just that fighting is lazy: it is the easy way of doing things, same with domination; integration is genuinely hard. So I don’t want to oversell anything here. But remember intellectual laziness. It’s again that. And this also brings me to the second point, and it’s from Laloux, not Follett: This needs to be learned. Nobody can expect good discussions and good communication to happen without training. This applies to transforming companies and newly created companies, and to CEOs, frontliners and new hires, of course. Many of the organizations he studied train every new person in conflict resolution and in their ground rules.

Information sharing is also a point that Laloux looked at, and in the companies he uses as examples, this works quite differently from what you might be used to. See, in the absence of classical hierarchies we have seen that decisions are made by many — or all — not by a few managers. That means the team also needs access to full information. You and I, we are used to the need-to-know way of sharing information. But the point is that this is part of our vertical power structure; horizontal structures don’t need that to function. This also brings us back to Benkler’s analysis of how badly this restrictive information handling actually works: a lot of it is simply lost or not available when it is needed in hierarchical organizations.

I can hear you think: “But, but, this is also about organization. How should we organize information sharing if the managers don’t filter?” I understand you, and it goes along with another fear that goes something like: “But I am already overwhelmed with information today. So many emails and newsletters. Now I have to deal with all information within the organization.” The solution is quite simple — and it comes from my own experience, not some book. The problem is largely caused by today’s poor sharing of information. Information is usually pushed into the organization by memos — or emails or newsletters or Teams messages, nicely worded by a communication expert with some nice pictures or maybe even in a video. The main misunderstanding here is that information within an organization is a good that needs to be marketed and sold to people. That is of course wrong. Information is a resource for work, like steel or electricity or capital. It needs to be in a usable form. That means we can skip all the nice fluff and persuasive niceties. Information needs to be true, reliable and to the point. We are not newspapers trying to sell copies. We just need the thing, the information. That also means that organizations should stop pushing information. In organizations where people are trained to communicate and decide — and as we have seen need information for that — they will get — or pull — it when they need it.

There are two potential issues with that. First, how should people know what information is available — for example safety points or what kind of data has already been collected? On the one hand this is knowledge — sort of knowledge about knowledge. People need to be trained in what information to look for and how to look for it — and what not to forget to look for. And then all information needs to be organized, indexed, possibly also bundled with a simple alert system. The second potential issue: How would that work if they are not filtered and prepared nicely? Nowadays there are really more than enough technological possibilities for that. I mean, when was Wikipedia founded? 25 years ago? Nowadays we would have even more advanced methods to do the same. I recently read a user comment asking Microsoft if they could not fully integrate Copilot into Teams and make it sort of a record holder of all knowledge in an organization, including all chat histories and documents and data published in SharePoint. I am sure this is actively being worked on and would of course make information pulling a lot more efficient.

We already saw Follett’s proposal for decision-making from a hundred years ago, and in general integration is still the best approach. In Laloux’s observations there are additional ideas from organizations that are worth looking at. Especially one from a software company where the governance meetings, in which the roles and rules of the organization are decided, follow a clear process and serve as an example for Follett’s integration. One person will present a proposal followed by a round of clarifying questions — the facilitator will block any hidden reactions. Then all participants can individually react to the proposal but without a discussion. The presenter will then amend and/or qualify his proposal. Only then follows a round with real, very specific objections, should there be any. If any objections surface, the facilitator leads the group in reworking the proposal to remove the objected parts while preserving the original intent. This is almost what Follett was aiming at with integration; what is still missing in this example is finding the true interest behind people’s objections and working that in.

Now let me just be transparent: Laloux selected some very special organizations, not the normal ones. His sample is not representative in any way, but it shows what is possible, and that is exactly why I like his work. 

And where Laloux’s observations really become interesting is when decision-making faces special circumstances such as a crisis situation. In crisis we might not have time for all that communication and cooperation stuff, right? Actually this might be underselling people’s ability to adapt. A car-part manufacturer, which was very cooperatively organized, was faced with reduced orders in 1990, in the wake of the First Gulf War. The CEO decided not to go all boss man and fire all the temp staff but literally stopped the machines and stepped on a soapbox in the middle of the plant. He then explained the situation to the staff. Within an hour the employees had proposed a temporary collective 25% pay cut to save the temp staff, put it to a vote, and unanimously agreed on it. This was possible because these people were used to making decisions jointly in normal circumstances. So they were enabled, so to speak, to also do it in a crisis.

There are other examples, usually involving organizations having to take shortcuts like giving teams options, not the possibility to discuss the whole problem in a crisis situation. But organizations that are used to communicating, and to discussing with the intention to solve or decide, can do the same in a crisis. And why not.

Another area where it is interesting how communication is used in next-level or cooperative organizations is conflict resolution. An American tomato-processing company has a very detailed structure for this — but elements of it can be found in many of these organizations. First of all, they have ground rules — as many of these companies do — on how to work in general and together in particular. In their case it’s “never use force against other people and honor your commitments.” Another example is “People are of equal human worth; people are essentially good unless proven otherwise; there is no single right way to manage corporate issues.” Both work as very simple guidelines that can be understood by everybody.

That of course doesn’t mean that conflicts do not emerge. The tomato company has a clear escalation ladder. But escalation does not mean I report you to your boss; it means who gets involved to solve the conflict. Firstly, the parties have a private conversation in which the complaining party makes a clear request, not a judgment or demand. This can then be met with a yes, no or a counterproposal. If they cannot resolve their issue, a mutually trusted colleague is asked to act as a mediator who supports but cannot judge or rule. If that doesn’t help, a panel of topic-related colleagues is asked to do the same as the mediator did. As a last “escalation,” the company founder will be added to that panel, further adding to the moral weight, but still he cannot impose anything. In this way, more and more assistance is provided for conflict resolution, but the involved parties still have to agree on a solution themselves. Sounds like a lot for a bit of interpersonal conflict. That is because every kind of conflict in this company is handled that way, whether it’s technical, personal, performance-related and so on.

Other companies have found that making team assignments more flexible and giving people the possibility to switch teams easily works as a relief valve. When, for example, your team leader starts to act too bossy, you can just leave that team. In other organizations, team leads or other hierarchical positions are elected by the team members.

This is all fine and well, but you might interject — as did I — this sounds like all the people in these organizations are nice, holding hands constantly and singing Kumbaya all the time. What about people not playing nicely? What about having bad actors in the organization? This is an important point, and there are three types that we have to distinguish. First, opportunism — or free-riding and shirking — which Ostrom, whom we know from the last essay, covered in detail. A lot of her design principles for successful self-organization address exactly that point. In short, it is important to remember that self-organization is not the absence of control; it is a different allocation of control. Meaning the members of the organization do the controlling themselves — you know, like the Japanese villagers themselves patrol the woods, and the violators’ fine of good old sake goes to the patrolling villager as compensation.

The second category is more malicious behavior that we can describe as destructive and/or non-rational. This often cannot be handled as gracefully. It is when controls or fines don’t really take effect. That is why even in Ostrom’s commons and in next-level organizations there are still exclusion or firing mechanisms. Sometimes, if all reasoning fails, you just have to let people go. Making this short statement is sort of an exclusion itself and we will come back to this in a bad things happen essay, I promise.

The third type of bad acting is power-seeking and manipulation by individuals within the organization. This would be stuff like controlling the agenda, coalition building, and monopolizing airtime within discussions. This is a tricky thing. So much so, I have to bring up one of my least favorite German philosophers and sociologists, Jürgen Habermas. He actually has very good ideas, I just have an early student trauma of having to read his forever-ongoing arguments in far-too-complicated gibberish. One of his central subjects, his social theory, touches exactly on this manipulation problem in particular, but is relevant for the whole topic of communication in cooperation.

Habermas — kidding aside, his ideas are truly brilliant — first of all gives us the names for the different communication methods we were looking for. Communicative action is what we want because it tries to reach an understanding or agreement. Strategic action, on the other hand, is communication aimed at getting the outcome I want rather than at reaching understanding. Done openly, like haggling over a price, that’s fine. Done in disguise, it’s manipulation — the manipulative bad actor. The disguised kind is quite difficult to spot because both look like communication. Manipulators do not need to scream and shout or have a title; they just have a hidden agenda — see, hidden as in hard to see. My practical version of this is a test in the spirit of Habermas, which you can use. If in doubt about manipulative elements, you can ask yourself: Would this argument have the same weight if somebody else had made it?

To understand the issue a bit better, we have to look at the structural source of it. So let me quickly explain Habermas’s social theory. He divided the social realm into the lifeworld and the system. The lifeworld is the part that is coordinated primarily through communication and mutual understanding. People talk, interpret situations together, come up with rules, build trust and arrive at shared meaning. Ha, sounds exactly like our cooperative organizations. The system, on the other hand, coordinates via impersonal mechanisms, mainly money and power. This is what markets and bureaucracies mainly do. Now the problem he saw was that mechanisms belonging to the system increasingly invade areas that ought to depend on communication and mutual understanding. In our case it is status, money or power creeping into a discussion that should be about understanding, and the manipulator is the person who uses it while pretending to reason.

Habermas was mainly thinking of whole societies, with their economies and states, but his solution can still be applied to an individual organization. Yet it is not a “perfect” or absolute recipe. He doesn’t want to just get rid of the system. He proposes roughly — again I have to apologize to all the hardcore Habermas fans for some simplification — two responses. The lifeworld should be protected by not replacing communicative processes with impersonal mechanisms like power struggle and bureaucracy. And communication has to be strengthened by making decisions and rule-making part of a process that involves the affected people, with claims questioned and justified through reason rather than authority or status or money or coercion.

This will prevent some of the manipulative forces, but the danger always remains. But it also answers the big question behind all of this cooperation talk: Can social order exist among autonomous individuals? Individuals who are not told to or forced to do something. Habermas’s answer is clear: Yes, by using communicative action and discourse as the mechanism for social integration — rather than money or power.

I’m not a big fan of summaries, but let me try it this time to hammer these really important ideas into your heads: Cooperation does not just magically appear when job titles are removed. It requires a different coordination mechanism, which we all know: communication. But constructive communication is not just talking. It requires us to learn about information sharing, integration, decision-making, conflict resolution and safeguards against manipulation. And by learning how to really talk with each other, we as autonomous people can create order together.

But remember, with all the talk about teams and cooperation, we have seen again that it all depends on reasoning. Reasoning is what keeps communication and cooperation from failing. And who does the reasoning? Exactly, you do, and so do I, and Alice, and the caterpillar, although with all the puffing maybe not.

Don’t just take it from me, here is some good stuff to read:

Carroll, Lewis. “Alice’s Adventures in Wonderland”. 1865.

Follett, Mary Parker. “Creative Experience”. Longmans, Green and Co., 1924.

Follett, Mary Parker. “Dynamic Administration: The Collected Papers of Mary Parker Follett”. Edited by Henry C. Metcalf and L. Urwick. 1941.

Laloux, Frederic. “Reinventing Organizations: A Guide to Creating Organizations Inspired by the Next Stage in Human Consciousness”. Nelson Parker, 2014.

Habermas, Jürgen. “The Theory of Communicative Action”. Vol. 1: “Reason and the Rationalization of Society”, translated by Thomas McCarthy. Beacon Press, 1984. Vol. 2: “Lifeworld and System: A Critique of Functionalist Reason”. Beacon Press, 1987.

Finlayson, James Gordon. “Habermas: A Very Short Introduction”. Oxford University Press, 2005.

Or

How cooperation works

“Come back!” the Caterpillar called after her. “I’ve something important to say!” This sounded promising, certainly: Alice turned and came back again.

“Keep your temper,” said the Caterpillar.

“Is that all?” said Alice, swallowing down her anger as well as she could.

“No,” said the Caterpillar.

No, indeed, we saw in the last essay that value creation most of the time comes from horizontal cooperation, not vertical hierarchy in business organizations. Cooperation is not a new concept of course, although in the world of business it is often considered like an anarchist in a well-ordered firm, a left-wing fast food-idea at a fine business lunch, a screaming activist at a shareholder meeting. Look, this is all stupid talk. Cooperation can be — as already shown — found everywhere and is fundamental in business. What you can do with it, is something we will discuss in this essay.

Easiest to discover is cooperation in your everyday business organization — your company, or mine: Gummy Bear Inc. for example. It’s easy because it is in the definition of an organization — every, not just, but including business. Communication, willingness and purpose are the only things that are really needed to form an organization. Purpose is clear and we talked about that before. You need something you want to do — like making gummy bear or achieve world peace. And there needs to be the intention to do something — unless you are the world organization of nihilism but then you also want to do something, promoting nothingness. The first — communication — is probably the most interesting one because it doesn’t just have to be there it also sets the organization’s boundaries regarding size. It’s actually mathematics. An organization of 5 people has to deal with 10 relationships (A-B, A-C, A-D, A-E, B-C, B-D, B-E, C-D, C-E, D-E) if you double the number of people you are not just doubling the number of relationships; it jumps to 45 (A-B, A-C, ha, thought I would write that out, didn’t you). Just to show off: as a formula it would be n(n-1)/2. Chester I. Barnard discusses this problem in his 1938 book “The Functions of the Executive”. He did a great analysis of cooperation in companies and he was no egghead. He was a manager, self-educated in social theory.

Barnard pointed out some important characteristics of how people interact within an organization. And he also made clear that although there is interaction, individuals always exist outside all organizations. Individuals do not come into being once they do something within a company, they are there before and after and also during. Individuals have relations to the organization, but they are never fully absorbed into it. That is important because it does away with the machine-like picture that we discovered and disregarded in the last essay (input, process, output). But my organizational setup seems very machine-like, you might say, we have org charts and processes and such. That is the formal organization and it is true, but beside each formal organization there is also an informal one. Don’t know how to work the printer? Just ask Steven. He can tell you. That is possible because both you and Steven exist also outside the — formal — organization. So you can act outside of your role in the machine. Of course Steven’s printer advice is just one of the gazillion things that an informal organization does. And see that is not a workaround, it’s just always the case, each formal organization also has an informal counterpart. And just to be clear: both organizations are needed and require each other. The formal organization would collapse if informal activities were not present. And the informal organization cannot exist in a vacuum. Now what does all that have to do with cooperation, well especially informal organizations work almost always on a horizontal level, because hierarchy is something formal.

We already heard from Margaret M. Blair and Lynn A. Stout when we looked at what business organizations are. Their team production theory emphasizes that the horizontal or lateral interaction — as they call it — is crucial to business organization. According to them hierarchy alone cannot explain how value is created. Most of what it does is coordinate or enable, not create. So while it might be needed to keep the interaction running smoothly, the lateral interactions themselves are creating the value. Even some of the coordination — the classical hierarchical function — happens cooperatively.

This begs the question: what are some of the mechanics in the hierarchical structure that make it problematic and drive horizontal activity. For that permit me to introduce you to the law scholar Yochai Benkler and his very cool paper “Coase’s Penguin, or, Linux and the Nature of the Firm” from 2002. One of the things Benkler describes is how information flows can explain some of this whole horizontal thing happening in vertical organizations. See all formal organizations handle information really badly by design. Benkler calls it lossy, but I call it badly. What happens is that hierarchy-based systems discard all information that did not arrive through the authorized or formal channel from a formal source. Problem is of course that a lot of information does not travel in a formal way — have you heard what is happening in Finance? Steven, about that printer. So the hierarchy is bad at transmitting information within the organization. It is like a person who only listens to information delivered in exactly the right format, and ignores good advice just because it didn’t arrive that way. So a lot of information has to flow via a horizontal structure outside the hierarchy.

Another point he is making is about what the role of management is in a business. Whenever there is uncertainty about what to do — and there often is in companies — the manager tells us what to do, he commands. But where does this command come from — the content I mean. It actually comes from the market. See the manager’s job is to translate the market into business internal actions — Produce more red gummy bears, they sell good. But this is not different than what the market alone would tell us, management is basically just a very well paid translator. This is important to note because this translation function can also be taken over by other roles than the manager and is not necessarily tied to the activity of command. After all, a translation app is not commanding me, it’s just giving me information. So that if I feel really fancy I can call my company Societas Ursorum Gummosorum, but I don’t have to. Just to make this clear: That last point is my reading of Benkler, not his exact claim.

The way cooperation can improve classical hierarchy-based structures can best be shown with one example, although there are many. Xerox used to have a simple setup for informing its technicians about its photocopiers. Have the engineers write a manual with all the possible failures they could think of and give it to the technicians to read. Problem was that the engineers did not think of everything that could go wrong. All technicians together knew everything that could happen with these machines, but each individual technician was on his own. They lacked a way to pool their collective knowledge. So Xerox invented an electronic system where technicians could post all the cases, and specialists and other technicians could support them and discuss the cases. By integrating a horizontal cooperation — or peer production as Benkler calls it — Xerox made the instruction-following technicians into knowledge creators, they added value to their support process which was still nested inside a classical company.

Thinking back to Coase and how markets — individual interactions with high transaction costs each time — and firms — cheaper bundled interactions — work, peer production or cooperation gives us a third possibility to use in traditional business organizations.

So it works within normal companies that are mainly hierarchically structured. But are there any forms of organization that are structured on horizontal cooperation exclusively or mainly? Yes, there are and the idea comes from very different directions.

When Václav Havel in 1978 discussed political philosophy he was not just critical of the system he lived in — communist authoritarianism, he actually called it post-totalitarian — and the systems he observed elsewhere — democracy —, he also proposed a sort of post-democratic structure as solution to both. He imagined that a system that was run by small organizations, small enough to ensure personal trust and responsibility would be more open and flexible instead of the strong bureaucracy that would lead to a dehumanization and eventually a collapse of the system. Organizations would also have to be temporary, existing only as long as their purpose had to be fulfilled in order not to be consumed by the mechanism of the form. They would have to be plural — meaning with the intention of there being many doing the same rather than monopolistic. And most importantly these organizations would have to be task-driven — purpose-driven as I call it — not being loyal to the organization itself. That also means leadership would empower those who are best suited for the task, not the ones holding an office or being at a specific position on an organigram.

Interestingly what Havel had envisioned was already existing at the time — at least partly — and still does today. We will get to that in a moment, but first let me sort of give you a framework for organizations. And I promise to stay really simple and fundamental. We already heard from Benkler and how he continued Coase’s work on the nature of the firm. For Coase there was either the market or the firm — either individual contracts for every transaction or bundled transactions. Benkler adds a third possibility which is called peer-production. You and I all know that. It is how development of most open source software — ah Linux, the penguin, now you get it — is organized. A peer production system falls into the category of self-organization — at least more or less, but details will follow. Now I promised to give you a framework. Here it is. There are two axes.

The first axis concerns how something can be organized most efficiently: Market exchange, organizational — meaning classical hierarchy business organization, the firm — or via peer-production. Now getting an apple is something so simple that we probably leave it to the market exchange — I go to the grocery store and buy one. But if I want to have apples delivered each day to all of my office locations I might want to either buy them wholesale or if I think I can organize it more efficiently within my company, I tell HR to organize that — then it’s organizational. Now if I want to make a website with pictures of all the different kinds of apples — and there are more than 7’500 — I might ask the community of apple experts and lovers to contribute pictures that can then be seen on the website. This is certainly more efficient than me traveling all around the world or founding a company for the sole purpose of making apple photographs. Clear so far?

Good, the second axis evaluates the property system and there are two possibilities. Is it worth owning a resource — so building/buying it or defending my claim to it — or would that cost more than what it’s worth and is it better to share it? Did I lose you? Buying an apple is a simple task. So the implementation costs of me owning an apple are relatively low. I have to go to the grocery store — that always bugs me a bit. I wish we had replicators at home — Apple, Granny Smith, cold. But if I own a farm in California, getting water for my apple trees is a different story. I have to get it from a water reservoir. Building my own or buying one might be too expensive or complicated which also means too expensive. So I have to rely on what Elinor Ostrom defined as common pool resources in her 1990 book “Governing the Commons”. I don’t own the water, but I have the possibility to use it.

Ostrom is quite important in this discussion about cooperation because she laid out a lot of the groundwork. And again a framework — so sorry, but also kinda important — which again takes Coase’s firm as option 1. It is a voluntary contract with at least partly the aim to give power to the entrepreneur, the boss man. Ostrom adds a second option which is the state, which can also organize and will do so by a monopoly of force — that doesn’t have to be actual violence but it can include that. The third option to organize work is then the one she analyzed and described as self-organization, and the reason for it is that, if it can be made to work, it would need neither the state’s force nor the contractual power of the firm to guarantee such things as supply of rules, commitment of the involved parties or monitoring of the rules. Therefore, it would be more cost-efficient. Also interesting that self-organization differs from the firm or the state not just in that it does not use coercion from the top, but also in that it is organizing in the sense of a process, not as the result of a process. Firms and states, on the other hand, are results. This is of course another reason why it is less formal and often simpler and less expensive to self-organize.

Rule-compliance is a universal problem every form of organizing has to solve somehow — firm and state solve it with authority from the top, and now we’re about to see how self-organization solves it without that. There has actually been quite a lot of theoretical thinking put into this. And we know just out of episodic experience that it is a real issue when people work together. There are even some formulas — C > B – S which just means compliance beats breaking rules only if the sanction once detected is greater than the gain. Another issue is if you put in place an external enforcer you would need an explanation for what motivates the external enforcer to enforce. You could monitor the enforcer but who then monitors the monitors and so on. Same issue with the free-riders — people who just want to profit from an organization without contributing. You would need to monitor them externally and someone would… again ad infinitum. And there is a connection between the creation of rules, commitment and monitoring. As Ostrom describes: Without monitoring, there can be no credible commitment; without credible commitment, there is no reason to propose new rules. Meaning if you have a problem with one, the whole thing collapses.

Ostrom has developed some design principles for the cooperative or self-organizations but she approaches the problem by studying real-world examples of successful self-organizations. The examples all have to do with natural resources because remember the thing with ownership not being worthwhile and sharing being more realistic. That happens a lot in these cases. Like the fishery in Alanya, Turkey where on-shore fishing — especially the fishing locations — have to be assigned to the different fishers. Over centuries the fishers organized themselves and found an elegant rotation system that was then self-enforced between individuals and within the community.

In Japanese mountain villages, the common pool resource is plant-based — wood and thatch — and is again distributed by a self-developed system. Again these rules were not given from the outside but developed within the communities themselves. And again the enforcement happens from within. Rotating community members have the duty to act as detectives roaming the forest for any rule-breaking harvesting. The punishment is often in the form of cash and sake going directly to the detectives and so also works as compensation for this monitoring work.

A bit more complex system can be found in Valencia, Spain where the resource is water for the irrigation of fields. Depending on the subregion, individual rules were defined on how the water is distributed. For any disputes there is a centuries-old water court. Without lawyers, very direct and public decisions are made on the day of court.

There are dozens more successful examples that have been studied in the Swiss Alps, California, the Philippines, and there are countless more. But Ostrom also looked at the cases where self-organization and cooperation did not work out or were fragile. There is again a fishery in Turkey, this time Bodrum where the local cooperative collapsed by 1983. Before that the government had encouraged the purchase of larger off-shore fishing boats by locals but then did not enforce the 3-mile onshore limit. That plus the intrusion of tourism and hobby fishers overcrowded the fishing locations which had been allocated between the fishers. Another example is a fishery in Mawelle, Sri Lanka, where wealthy entrepreneurs accumulated multiple shares of the fishing rights and disrupted the original equilibrium between the fishers. Or water irrigation in the region of San Bernardino and Mojave in California where a single uniform self-organization was attempted in a very large diverse region. External interference is the most visible threat — but self-organizations also fail from within, when commitment wears down or monitoring becomes captured by insiders.

Out of the good, the bad and the ugly Ostrom extracted what works in the setup of the cooperatives and what does not and eventually came up with eight design principles — seven plus one actually but isn’t that… whatever: (1) Cooperatives need clearly defined boundaries both in substance and membership. (2) Rules have to match local conditions — no funny ideas from the outside. (3) There needs to be a forum for collective-decision or choice. The ones affected need to be able to modify the rules. (4) Monitoring has to be directly done by the cooperative members or at least the monitors have to be accountable to them. (5) Sanctions should be adjusted to how serious or repeated the violation is. (6) There needs to be a fast and easy mechanism for conflict resolution. (7) The right to organize needs to be recognized — at least minimally or at least not contested externally. (8) If the self-organization is very large it needs to be organized into nested enterprises. Sounds kind of detailed? It is, mainly because these are not theoretically developed — they are the common success and failure factors that Ostrom discovered in actual cases.

Benkler also has his design principles for peer-production and they differ a bit. But why? Shouldn’t this be the same? One was just looking at common pool resources and the other at information technology. But that is exactly the difference, the resource type. See Ostrom’s resources are things like water, fish, wood and other natural stuff — also called rivalrous resources. So her organization has to deal with scarcity and depletion and boundaries. Sanctions against overuse are really important. Benkler’s information is nonrival. So his design principles have to cover things like coordination and quality – developers integrating their stuff into a system not fishers overusing a location. He — and he is a lot less nice and doesn’t give us a straightforward list — also describes eight points. (1) Modularity means to break work into components that can be worked on independently. (2) Granularity describes that contribution should be possible in different sizes, from a couple of lines of code to month-long efforts. (3) You need to have a low-cost way to integrate the individual contributions. (4) People need to at least partly be able to self-select their contribution — otherwise it is not self-organization — but self-selection also has a strong influence on the next point. (5) Motivation comes from many different sources like interest, learning opportunity, social connections and other intrinsic sources — money, if at all, is only one possibility. (6) Who does what should be largely guided by decentralized information — trying to centralize all information should be avoided. (7) Where prices and managerial commands are unnecessary or costly, social norms, trust, reciprocity and reputation should coordinate work. The next point is a really interesting one and solves the quality issue I — as a quality person — was always wondering about. (8) Participants themselves will evaluate, correct and improve each other’s contributions. That is the peer review model and it is also found in the world of science.

For me this makes so much sense because when I first read Ostrom I was constantly thinking, well that does not apply to my — mainly information-driven — work, how can I apply this fishery cooperative mechanism to my way of working. And then I discovered Benkler and all of a sudden that works a lot better. Which is in itself an important point. In practice, hardly any real activity is purely one or the other — most work involves a mix of both kinds of resources. Time and budget often behave more like Ostrom’s natural resources whereas information or knowledge — Benkler — are present in most business processes anyway.

So if self-organization sometimes makes sense and if there are good real-life guiding principles, the logical next question is how to get from here to there — from a traditional hierarchical organization to a cooperative or self-organized organization or part of it? First it is important to remember that there is no true empty or fresh slate. There truly isn’t, believe me. Even if you found a start-up you will find yourself in a certain environment with certain structures in place. In addition we cannot just erase what was and start fresh. It is always just a change or — what I like to call it — a transformation of what is into what you aspire to. Ostrom found this in many cases she studied. Organizational change is successful if it happens in incremental, sequential small steps which involve low initial costs, therefore also being low risk. Eventually changes will accumulate and the commitment risk will rise but this happens gradually. It is also important that each step needs to build the incentive structure for the next. You cannot build a monitoring system if you are not sure how the monitors will be compensated. A central precondition from both Ostrom and Benkler is information sharing and discussion. While a forum for discussion alone is no solution it is required for self-organization. We have already discussed and will again how important information gathering and information sharing are for meaningful — or better solution-oriented — discussion.

So eventually after the transformation everything will be self-organized cooperatives? Sounds fun, doesn’t it? No, that’s not the point and it is not the point of any of the named authors. Self-organization, peer-production, and cooperation are additional ways to organize business activities or complete business organizations. Please note that I didn’t say option because I think it is more sort of a logical decision what is best suited than any choice of flavor.

See we had Benkler’s framework with the axes. On one side the way something is most efficiently organized – market, firm or self-organization and on the other side if it is cost-wise better to own a resource or share it from a common pool. So we have five different options. Ha, got you again, it’s six of course. Permit me to use my all-time favorite example of making gummy bears to fill the options.

If many people own gummy-bear machines and recipes and resources are easily available on the market, we are in the market and owning category. Because resources are cheap and easily available and the process is simple, organizing internally or peering would just add unnecessary cost and complexity — so market wins by default. (Other examples: a farmer’s market or Amazon Marketplace)

Now if there are still many people with their own infrastructure and simple process but they have to use a freely available resource like the forever known recipe for gummy bears or the red color extracted from the common red jumping spider, the best setup would be that of market transactions and sharing resources. (Or freelance photographers or street musicians)

In a situation where making gummy bears is a really complicated, multistep process that requires a shitload of R&D and machines and a tightly coordinated team effort and one mistake might turn the gummy bears into cement rabbits, but resources are still available on the market, it seems best to go all classic firm and owning resources structure. (Or Tesla or Zara or a lot of other traditional companies with a lot of in-house activities)

But what if under the same stupidly complicated but needed process you still need those damn common red jumping spiders and they are just jumping around everywhere. Or, more realistically, the gummy bear machines have to be located in a community-owned huge hangar-like building because they are as tall as thirty houses. Then you would still need a firm to coordinate your process but you need resources that are better to be shared, because your company cannot build a huge building, or because it’s stupid to buy spiders on the market when you can just collect them off the street. (Or an airline sharing airports)

Now what if a lot of people have gummy bear leg machines and a lot of people have gummy bear belly machines and a lot of people have gummy bear head machines and so on. These people, of course, also know how to make these parts because they have their recipes. Given that you’ll need one central machine — owned by a company — that can combine the legs, bellies, and heads to give you a complete gummy bear, and resources are available on the market — you know the stuff the gummy bears are made of — that would then be a logical example of self-organization and owning resources. (GitHub or knowledge-sharing platforms that are owned by a company like LEGO Ideas)

What if you have individual gummy bear parts contributors and the parts can be easily and cheaply combined — not with a machine but just with some spit. And resources — like recipe and gummy bear material and spit and the spiders — are mostly common and freely available. This would then be a form that is both self-organized and sharing resources. (Wikipedia or the Linux Foundation)

Some of these categories might sound niche to you, but that is not because they have no potential. The problem is rather that certain structures have become predominantly used and are often misapplied. The classical hierarchy/firm and ownership structure is often applied when conditions would encourage a more self-organized structure. In the field of information technology there have been examples like Microsoft Encarta trying to capture knowledge the way Wikipedia did later. AOL trying to organize the Internet — by using a walled-garden, centrally curated portal. Other times it is not a question of failure but a question of what would work better — cheaper or more efficient. Take Unix — originally developed and owned by AT&T, then fragmented into various commercial variants often being complicated to buy and operate for customers — and Linux — free and readily available in many versions. Often we don’t see the cheaper alternative because that structure has never been applied to that business activity. In the book business it was forever thought that the production and distribution system made publishing houses a requirement. But then self-publishing was enabled by modern printing machines able to produce single or small quantities of a book. Online booksellers made distribution more accessible and e-books shook up the old structures again. Now that does not mean that self-organization is perfect for everything. Self-publishing has its fair share of problems — quality control being one of them, marketing another. But it opened up new possibilities. See also gummy bears can be made by Gummy Bear Inc. They can hire 5 new food designers to make new flavors and forms. But you will only get what 5 people could think of. If you take all the 500 million gummy bear lovers on the planet with their own kitchens and knowledge and experience you might get a lot more ideas for new gummy delights.

Again cooperation is not the exclusive form for business organizations, it is a third possibility. It is certainly underused although clear criteria are available. Eventually what matters is what makes most sense when looking at costs, both for organization and for ownership. The logic itself is fairly straightforward once you know to look for it. Meaning we should just look at our organizations and be honest if we really have found the most cost-effective structure to use and set aside fixed stereotypes. We need to experiment more with more open structures. These experiments can be very confined to specific areas of our business — remember small incremental steps are important — but they do need a real fighting chance which brings me to one last point that Ostrom makes on probably the biggest threat any self-organization faces, and that is undermining from the outside. Self-organizations need an environment in which they are accepted as a valid option in the business world, not a marketing gag that often gets misused and destroyed.

Another important point is that large numbers of people, with diverse knowledge, often benefit value creation which is — as we have already seen — central to business activities. You know the 500 million gummy bear lovers. You might get that whiskey-infused grizzly gummy or the “Lil’ Nitro” or the 5-kg gummy bear.

With more open business structures we have the possibility to accomplish what Havel wanted for the political system: a more flexible, trustworthy and social way of doing business. Havel was talking about political resistance, not business. The pressures and incentives are different, but the basic principle still applies. But just to be crystal clear, as he made that point himself: no institutional design can guarantee moral actions. Structures alone cannot save us. We might get a better vessel — for some weather conditions at least — but we still need to get the sailing right. That’s what we’ll talk about next — how we work together better.

Don’t just take it from me, here is some good stuff to read:

Carroll, Lewis. “Alice’s Adventures in Wonderland”. 1865.

Barnard, Chester I. “The Functions of the Executive”. Harvard University Press, 1938.

Blair, Margaret M., and Lynn A. Stout. “A Team Production Theory of Corporate Law.” Virginia Law Review, vol. 85, no. 2, March 1999, pp. 247–328.

Havel, Václav. “The Power of the Powerless”. 1978.

Ostrom, Elinor. “Governing the Commons: The Evolution of Institutions for Collective Action”. Cambridge University Press, 1990.

Benkler, Yochai. “Coase’s Penguin, or, Linux and the Nature of the Firm.” Yale Law Journal, vol. 112, no. 3, December 2002, pp. 369–446.

Or

What are we looking at

“One side of what? The other side of what?” thought Alice to herself.

It is a good question that needs to be answered before we can proceed, because often we investigate and analyze and change and improve and mess up without ever asking ourselves: what is that thing that we are looking at?

Most of us — and I include myself in this, at least a former me — carry around a picture of business organizations as machines — inputs in, outputs out, run by whoever holds the controls. That picture is logically, historically and legally wrong. And it’s worth a little of our time to look at this more closely, because what it gets wrong shapes everything else we do.

And where better to start than with Ronald H. Coase and his 1937 paper “The Nature of the Firm.” Now this is hardcore economics, but we’ll get other points of view later. For Coase the definition of a firm — as he calls it — is given by the reason it is formed. Using any kind of resources — labour, capital, knowledge, steel, microchips or anything else — in the market has a certain real cost: finding it, finding the price, negotiating, making contracts. These are called transaction costs. A firm now replaces a number of these contracts with one broad contract. Thus the transaction costs should be lower. Now, costs here are meant in a total sense. Because if you are like me you probably think, well, it’s not just about cost, it would also be incredibly complicated — to the point of impossibility — to make individual contracts for all different resources with all the different providers of these resources. But see, for an economist like Coase, complicated just means more costs, which of course is true, we are just not used to calling it that. So firms are lowering transaction costs. Point. Of course firms generate their own internal costs too — that is what sets their limits — but that is a different discussion.

Again we want to raise our voice and say, but what about coordination as a reason for firms? So many people, so many resources cannot just work without it being coordinated. Agreed, coordination is real, but it’s not what actually explains why firms exist. See, coordination would also be possible in a multiverse of contracts, it would just be so complicated that transaction costs would explode.

So it is still about lowering transaction costs primarily. I know, sort of flat, but hey, it’s true. Something else driving up transaction costs is one of these obvious facts of life. We already discussed that time and change are aspects that are always there. In the last essay we saw that we are constantly playing a game of catch. Reality changes the moment we look at it. And because of our changes in perception, even the past changes. Therefore the future is necessarily uncertain to a degree. Now most contracts are made for the future. Making detailed, individual contracts for the future exchange of resources is problematic — to be clear: costly, not impossible. It is kinda hard to contract for future microchips whose workings I don’t even know today. Then a better possibility is to hire the knowledge and the labour and get it developed thus again lowering transaction costs.

So firms come to be because of costs and — permit me to single that out although it is connected to costs — because of uncertainty. Or: they replace individual, specified contracts with broad contracts that are just directional within limits — I hire you as a microchip engineer, not to make a specific, predetermined future microchip.

But if I learned one thing in law school, then it is that there are always different approaches to interpretation or definition. Another would be the historical approach. Now we have done our homework and already discussed the history of the corporation in a past essay. Colin Mayer in his book “Prosperity: Better Business Makes the Greater Good” has a good summary too — the whole book is of course interesting. In essence the historical approach shows that business organizations were not meant to be a simple machine, but more of an independent entity or a persona — yeah, that’s why we call it legal person today. See the need arose for an entity that was independent of the individual people, that could take over and act continuously on its own. The idea was built gradually — a development — starting with the Roman collegium to canon law to English municipal charters to guilds. These organizations were meant to perform a public and/or institutional function — for example tax collection, ocean trade, standards for craftsmanship. The whole shebang with the raising of capital was added later, by merging the corporate organization — whose focus was on the performance and the governance — and the partnership, which was the traditional fund-raising form. So out of their historical development, business organizations as we experience them today still have these two aspects: raising of capital and public purpose.

The funny persona thing was also a historical development. And it connects to something I quoted Barnard stating: that an organization would destroy itself if it had only a fixed purpose in the sense of a goal line — we want to make running shoes. In the historical original form, that was the case for business organizations. But already the Romans saw that this limits the usefulness of these organizations. They needed to adapt to needs. In one year, provide provisions for armies; in the next, collect taxes. And that is the difference between a machine, which is built for one purpose, and a persona, which can change. Not a hammer — which can only hammer nails and fingers — but a handyman, who can fix all kinds of things.

Now if we also want to look at the question of what is a business organization from a societal perspective, we need to establish who contributes to these organizations. Because remember, Coase doesn’t go there. He just said there are a lot of interactions or contracts, and it’s cheaper if we make a bundle out of these instead of having a hundred different contractual agreements. So who provides this input, this investment? Because that’s what these contributions are, investments in the broadest sense. Margaret M. Blair and Lynn A. Stout describe this in a paper called “A Team Production Theory of Corporate Law.” Yeah, I know, corporate law is not the most thrilling of subjects, but they have some neat ideas. So it is generally accepted that besides financial investors there are other contributors, and that a lot of them are actually essential for a business to exist: managers, employees, community, and of course the capital providers. So all of their contributions are needed investments or resources in the organization. The act of incorporation — sorry, strange way of saying the forming of a business — is that some contributors form a team by giving up control rights over the input, process and output to a new, separate legal entity that “takes on a life of its own and could, potentially, act against their interests.” Why do they do this? Well, good question. One is that once working together in a business, their contributions become firm-specific, meaning they would have less value outside their collaboration. This is common to all the contributors, so it establishes their connection. Then there is the problem of control. All forms where only one or a small number of contributors would hold control result in a risky situation for the others. So control is given to that formed entity or more precisely, to its mediating body — the board, because no single contributor group can be trusted to hold it alone.

Ok, we had the economic view, the historical and the societal one. We briefly mentioned purpose but we still need to discuss that in more detail. Purpose is actually philosophical, but don’t be scared, it’s going to stay rather practical.

But what does purpose have to do with definition? Well, even without building too much of a thought construct, we can all agree that knowing the purpose of something gives us a good idea of what that thing is. Specifically regarding business organizations, Mayer found that the purpose determines the inner workings — he calls it structure and conduct — which determine performance measured against purpose — in ocean trade you need ships not a server farm or a chemical lab. The first part here is interesting for us, because it means once we know the purpose, we know how a business organization is truly structured.

Now let’s go back to Blair and Stout, or better, their recounting of Alchian and Demsetz’s definition of team production. It is actually quite simple and consists of three parts: there are multiple resource types involved in production — the usual suspects: labour, materials, knowledge, things like that. The finished product cannot be split meaningfully into separate outputs — yes, you can take a smartphone apart, but usually the individual parts are then not very useful anymore. That’s also why there is no single owner of all these resources — no one contributor can claim the result for themselves, which is why it falls to the new entity to hold and administer it. 

Now let’s take that idea from Blair and Stout and add that the whole product actually becomes bigger than the sum of the parts. Think of it. Tiny speakers, a good touch-based GUI and microchips and batteries that you can recharge thousands of times, and all the other components of a smartphone are quite cool, but only when you put them together do they become something really useful. Or, to take their example, that by writing a paper together the two authors could come up with a lot nerdier ideas than either of them could have alone — meaning of course that collectively they hold more knowledge than individually.

Something we need to clear up now and here is why money is not the perfect answer to the purpose of a business organization. Profit — right? — is what is most important in business. The whole discussion of profit is detailed and we need to be very precise. So open up your reading ears wide. Profit is really important in any business activity. No author writing about economics or management would disagree. At least I haven’t found one. But what about the more critical ones? Authors like Mayer or Porter or Freeman and Mintzberg, also the epic D&D — Drucker and Deming — still think that profit is important. And they all basically agree that profit is important mainly because it is a requirement for the survival of the organization. It is a necessity, an economic condition of survival. Mayer goes as far as calling it actually an input in the process, not an output. But these more nuanced thinkers also agree that profit is not the purpose of a business organization. Making money is cool and needed, but you need to do something to make it. That something is your purpose. Even top managers built their businesses around something other than profit: Paul Polman (Unilever), Konosuke Matsushita (Panasonic), Akio Morita (Sony) and Steve Jobs for example. We can describe this something as value and the process is called value creation — and yeah, we will discuss it in a future essay.

Now, let me add one bonus point, because it comes out of this discussion. And that is how structurally value is created in a business organization. We saw that teams create value, the bundled effort, out of the many needed resources. This is nothing new. The team in Blair and Stout created something that was more than the sum of all inputs. It added value. Now the truly interesting part is that these value-adding processes work mostly horizontally in businesses. We know that from processes in our daily experience. Most process flows that add value are horizontal. One person or team adds to what the last added. My colleague in procurement buys the material. Somebody in R&D designs the speakers. I put together the microchip, then another team puts the microchip into the smartphone — sorry, remember I shorted a hard disk once — somebody develops and loads the software. It’s generally speaking all a horizontal process, or let’s call it a horizontally oriented flow. What is vertical structure — or hierarchy — supposed to do then? It’s needed too, because it coordinates, since otherwise processes would constantly run into problems. You can say that hierarchy is a cost of coordination, but not a source of value. The value comes from what happens between people at the same level, it is based on horizontal cooperation. So hierarchy is problem-driven and cooperation is value-driven. Did this just blow your mind? Mine certainly did when I first found this.

And since we found that value adding is at the heart of what business is, so must cooperation be too. We found a bundling of relationships as a reason for firms, the need for a flexible, persona-like organization, the team forming, and value creation as the purpose of business organizations. Now we can also add that they are about cooperation.

Which means we have to discuss cooperation next, and we will. But for now, we certainly have a good picture of what we are looking at.

Don’t just take it from me, here is some good stuff to read:

Carroll, Lewis. “Alice’s Adventures in Wonderland”. 1865.

Coase, Ronald H. “The Nature of the Firm.” Economica, vol. 4, no. 16, November 1937, pp. 386–405.

Mayer, Colin. “Prosperity: Better Business Makes the Greater Good”. Oxford University Press, 2018.

Blair, Margaret M., and Lynn A. Stout. “A Team Production Theory of Corporate Law.” Virginia Law Review, vol. 85, no. 2, March 1999, pp. 247–328.

Drucker, Peter F. “The Practice of Management”. Harper & Row, 1954.

Or

How we look at things

For some minutes it puffed away without speaking, but at last it unfolded its arms, took the hookah out of its mouth again, and said, “So you think you’re changed, do you?”

Yes, I think I have changed. Have you? The objects I look at have changed, too, or is it just my perception? And ah, again, it just happened, new colors, new meaning. My gosh, again. Or is it not the coloring but my eyes, my brain? And how about you? Do you see it, too? Because otherwise what good does it do? We will get into all of this. Step by step, but first let me assure you we are still talking about business, management theory, systems and organizations. And we are talking about people, me and you and them. The subject today is perception and perspective, framing and assumptions. How we look at something. Why is this important? Well, it is pretty much the first step. Without looking at something — or perceiving it in another way — how should we do anything with it? So we have to capture IT somehow, make IT our own. Sounds logical? Yes? The problem is that this might exactly be the start of our problems.

When we look at reality we build fixed, isolated and often fictional snapshots and think it’s the real thing. A KPI this afternoon. Three months of data. Option A, B, or C. True for a fraction of a second, but treated as permanent.

We do this for three reasons — time pressure, missing knowledge, and something harder to admit: intellectual laziness. Not stupidity. The simple failure to use a capacity we already have, our ability to reason.

It matters because it’s the same failure that lets evil become ordinary, that lets a bad decision go unexamined, that lets a strategy run without ever checking its own facts. This happens all the time — quiet, everywhere, and costly. Let’s talk about how to avoid it.

That is what we always do. We take fixed snapshots of reality. We calculate a KPI this afternoon, we make an analysis of the last three months. We evaluate options A, B and C. But these are just aspects; they leave out the majority. We look at the front but forget what is behind or behind the behind. Plus these facts are true only for a fraction of a second. After that reality has changed again, even the past. What we capture are fictional, fragmented versions of reality — sometimes that is necessary, sometimes this is flawed and sometimes downright catastrophic. Or as in Köhler’s experiments it can cost us the banana. Understanding that coloring alone is not what points you to the food. He discovered that apes can learn that relation matters. Dark grey and light grey are not just important because of their individual colors. The fact that when an even lighter grey is placed beside the light grey, the even lighter grey is the correct path, is an important lesson to learn. It is about relational perception. We tend to view stuff in isolation when the relationship is really the meaningful part. And if apes can learn it, we should be able to as well.

Why then do we often stop before we have a good enough perception of what it is that we are concerned with? One reason is what you might think right now: “I don’t have time for this!” Time or other resource pressure leads us to take shortcuts. To think that a superficial idea is enough. Something that we also encountered when we looked at decision-making processes.

Second is that all of this sounds rather complicated. Stuff behind other stuff behind other stuff. And how should I know how two things relate, let alone a dozen different aspects? It’s like when I look at the inner workings of my computer and don’t have a clue what each shiny piece of hardware does or if it short-circuits when I put the cable in the other way — a hard disk in 1998, if you wonder. That is me not having had enough education or knowledge or experience. Both of these things — resource and knowledge scarcity — will be the topic of future essays.

Today we talk about the third reason that keeps us from having a good enough perception of reality. And for lack of a better term I have to call it Intellectual Laziness — sorry for the bluntness.

Let me make one thing clear, because I’ve had that discussion many times. By Intellectual Laziness I don’t mean the opposite of being an intellectual — Rodin’s Thinker, sitting around pondering the great questions of existence. I mean something much simpler: our tendency not to think when we could. We all have the ability to reason, to draw conclusions, to question assumptions and to follow a thought from one step to the next. Intellectual laziness is simply the failure to use these abilities when we should. Even apes can do it. If you cannot reach the banana with one stick, maybe two sticks will do the trick.

We talked a lot about the problems of missing moral reasoning when we looked at organizations and totalitarianism, with Havel, Jung and especially Arendt pointing out what happens if individuals cease to reason, to think for themselves. It leads to a void, a sum of a million zeros, that is often filled by bad actors, or the worst possible, as in the example of Nazi Germany. The same failure that makes evil become ordinary lets a bad decision go unexamined or keeps a strategy without a foundation of facts.

This shortcut to snapshots, to the superficial, is something we already found with Boorstin when he analyzed the media in the 1960s. The difference between the ideal and the image. The complex actual thing, the complicated idea of it, gets covered up by a simple image. Quick to draw, easy to market. It’s the slogan that replaces the mental health program, the RCA sticker that replaces actually getting dirty, digging beneath the surface and looking for the roots and worms and bacteria.

One particularly nasty problem with this kind of fixed snapshot thinking that we so often find in organizations is that it is self-destructive by logic. See, if you perceive your organization as something fixed, its most important characteristic — its purpose, what it does — is also fixed. We make shoes for runners. We make computers for normal people. If Nike or Apple had stuck to that initial purpose — as the story is usually told — they would have done that, and once accomplished, there would be nothing left to do. That would mean that in the end, accomplishment would lead to failure as an organization. What successful companies like the ones mentioned did was expand their purpose — to other athletes and customers, in Nike’s case, or to other product categories, in Apple’s. Of course, if an organization fails to reach its purpose it also fails — stating the obvious. In both cases purpose should be more understood as a direction not as a destination. Barnard already described this mechanism in his book “The Functions of the Executive” from 1938. He also notes that this giving of purpose is a tricky thing anyway, because it is a generalization which “is normally substituted in our minds for the concrete performances that are the real purposes.” Just think of it like a slogan, the kind organizations often use to make their purpose more general but also easier to understand. This is not actual performance. It’s marketing. The image, not the ideal.

A similar issue is described by Follett, whom we already know from the discussion of the manager’s role, in her book “Creative Experience” from 1924. In organizations we often work with concepts in various forms. Concepts can be strategies, process descriptions, org charts and such things. The problem is that on their own they are not more than the words they are made up of. See, a concept is something that was created in the past that is supposed to describe the present or future. What connects it from the past to the future, what gives it meaning, is the described activity. A concept is the materialization of an ongoing discussion that can be used in putting something into action. Most importantly, it stays viable only as long as it’s connected to that activity. Viewed on its own, it loses its sense. That shifts how we quality people should look at process descriptions, for example. They need to be directly connected to actual work — a living document is a good description of this. If a concept loses its connection to what it describes, it’s losing life, and it cannot tell us anything about reality.

Follett describes some more issues, but let me just point out one that is very impactful and close to all of our experiences. The problem with specialization. Jobs and tasks have become more and more specialized — another obvious statement, sorry. The connection to perception is that people are trained with the intended purpose of only seeing a very narrow part of the whole. She gives the example of a celery-disease specialist who is asked by the farmer how to improve his cultivation of said plant. The specialist replies that he has no knowledge of how to grow celery, only of its diseases. Compartmentalizing knowledge to such a degree makes his knowledge useless the moment the farmer asks a real question. The farmer would need a person with experience of at least everything that has to do with growing celery — probably even wider. Somebody having only a narrow view cannot give him a good enough idea of reality.

So we saw the issues relating to fragmented and fixed snapshots of reality. Now how do we catch ourselves and others when this happens? We can split this up into individual perception errors and group perception errors. Be aware that individual perception errors can of course also happen in groups, since groups are just a bunch of individuals.

The first thing we individually do is forget which level or which part we are looking at. Or just the fact that we are looking at only a part of the whole thing — which, by the way, is always the case, but we’ll get to how to solve that later. First it is important to acknowledge that we look at parts, and to know which one. Goffman described this famously in his book “The Presentation of Self in Everyday Life” from 1959. He uses the metaphor of a theater play to separate the frontstage and the backstage. The audience is shown the act on the frontstage, but all the preparation and the actors’ real selves are only visible backstage. Frontstage is not a lie. It is just the part that the audience sees, or the customer in an organization. It’s the finished product — not the R&D, not the sourcing, not the production. That is all backstage. For an auditor this has always been important to distinguish. He is shown some facts but not what leads up to them. His job, or mine, is then — if he wants to do it well — to discover the backstage area. The most important thing to remember is never to mistake the frontstage for the complete thing. Already realizing that will give you a better idea, because knowing that you don’t see everything is the first lesson.

The second important point comes again from Follett. And it sort of brings in time as an aspect — where Goffman was more about location, if you want to put it that way. And it’s even more important, because situations and facts always change over time. The constant is that everything changes. Usually we just think, well, let’s update the numbers, but the situation is still the same. That’s actually a logical shortcut — you know, laziness again. Because the situation consists of the facts. If they change, the situation also does. Or, as she puts it: “And when a situation changes we have not a new variation under the old fact, but a new fact.” You might say that if this is true, we never have any consistency. Yes, that is kind of the point. Change and time are the only constants. Too philosophical? Well, if you are making an important decision it is worth checking how accurate your information still is so you don’t act on outdated data. So this is a very practical issue.

Location and time perception are issues we all have as individuals, but when we get into multiplayer mode, there are more. And most important analyses happen in a group environment, although I still prefer a good single-player RPG. We stay with Follett and the problem of muddle vs. difference. When we individually gather facts we will have other perceptions of reality. But these other perceptions can be two different things. They can be differences that are based on accurate information. This would be absolutely normal and a good basis for discussion. But a lot of the time, different perceptions of reality are based on inaccurate information. Follett calls that muddle, and it should be detected. If I am looking at a black snake, my perception is different from yours if you think it’s a fallen branch. We are then in completely different situations. And one of us — or both of us — clearly has inaccurate information, because they cannot both be correct. And any collaboration in this moment would be very difficult. Should we throw the thing for our dog to catch, or should we run away? And even if we had the same information that it is a snake, we could still differ in what to do. Kill it? Feed it? Or cuddle it? This is then something we can sort out together, because we have the same accurate information. Follett calls the first — snake or branch — muddle, and the second kind — kill or cuddle — difference.

What also happens a lot in groups is that we don’t even talk about the same question. Sounds familiar? I know it does. As a moderator in countless meetings and discussions, I often realize that people have somehow drifted apart — not necessarily in their opinion, but in the subject itself. We call this the mis-scoped question. In auditing training I often put beginners in a role-play situation in which the auditee never talks about the points he or she has been asked about, but about something completely different. And I do this not because it is funny to watch — although it is — but because it happens a lot when two or more people discuss a point. The reason behind it is our tendency to come up with these fast snapshots of our situation, or discussion topic, in this case. And since they are not based on real facts and not well reasoned, different people develop different snapshots of what the topic is – freestyle perception. So it is really important to always ask whether you are arguing about the same question or sub-question — the same scope — or whether you have drifted apart and are needlessly spending effort discussing different topics.

So after catching issues with perception based on location, time, accuracy and scoping, we have to ask ourselves how this can be avoided from the start. The solution needs to address both individual perception and group perception. In my opinion it is a combination of two ideas from two thinkers. Follett has already shown us a couple of issues, and she also has some neat ideas — actually so many that some will be discussed in later essays.

Specifically regarding the identified issues, she offers the idea of cooperative fact-finding. Some issues arise because in groups we look at different levels, from different time periods, and scope issues differently. But we can avoid that intellectual ego-odyssey when we collect facts together as a group. Imagine this practically — not some lone analyst crunching numbers in Excel or watching a process with a stopwatch, but groups of people looking at reality, observing it together. This solves the issue of agreeing on facts, because we find them together at the same time, and it also saves the time of presenting facts to others. If you feel a bit uneasy, you are not alone — most people do. For now I would say this is mostly because we are not used to really collaborating on that level. I try to encourage these moments in more formal ways when I promote people doing gemba walks — groups of managers together looking at processes in real life from an improvement perspective. When I join them, I often feel like a herding dog, keeping the sheep together, catching any strays. Improvement workshops and coaching sessions can be more informal moments when you look for facts together, and the experience is truly inspiring. It is just thrilling, digging into a system, getting data out, analyzing it as a group.

Finding facts together does not solve all differences — I mentioned that before in the muddle vs. difference part — but it gives something worth the effort to discuss. And because facts have been collected together, everybody feels entitled to discuss them. It enables individuals in groups to confront each other’s perceptions of those facts more, and you get less crowd unanimity. Because they are not just somebody else’s facts, they are your facts too. You own them too, and you can have an opinion about them.

What we are still missing is a clear instrument, a looking-glass with which we can — hopefully together — perceive reality. Luckily for us, there is already sort of a looking-glass in management theory. It comes from W. Edwards Deming, the four-lenses approach, as I like to call it, or as he called it: PROFOUND KNOWLEDGE. Sorry, it just sounds so epic. Jokes aside, he addressed a lot of issues we have with looking at organizations. But his approach works for other areas too. The four lenses are System, Variation, Knowledge and Psychology.

It is really important to understand that these are not the “what” we are looking at, but the “how” we are looking at something. Yes, a system is something. Deming defined it as a number of components which are interdependent and have an aim. And that is essentially the same definition we also find with systems thinkers like Donella Meadows. Elements plus interconnection plus purpose. These must always be present, and they always are. For now this is important for us insofar as it gives us a list of must-sees. When we look at a process, for example, we usually look for the main elements, we usually know the purpose of the process. What we often forget are the dependencies on other elements. Take, for example, a thermostat and a heater. Those are the elements. When the room — another element — gets too cold, the thermostat turns the heating on. The heating raises the temperature of the air — again, an element — which causes the thermostat to turn it off again. The purpose, of course, is to keep the room at a certain temperature. The interconnection is the temperature affecting the thermostat, the thermostat affecting the heating, and the heating affecting the air. Now, if we are interested in room temperature, there will be other interconnections that affect it. The temperature outside the room, the insulation of the room, and any windows and doors and how they are opened or closed. Which brings in humans as an element in that system. Their behavior will be dependent on other elements. Time of day might be an element, or a schedule. These depend on more elements. This raises an issue that has come up before. How much detail is enough in our perception of reality? Can I live with seeing only part of it? Can facts be slightly outdated? We will certainly discuss Systems Thinking in more detail, and how it can be used practically, in a future essay, but to answer the question briefly: there is something close to what I call — inspired by Meadows — the boundary-drawn-for-purpose principle. We have to set limits to what we look at based on why we are looking at it. Otherwise we would go further and further into details and get lost. The decision about what is enough is not easy, but we will experience it once we use applied Systems Thinking.

Deming’s next lens is variation — or data and its most important characteristic. We all know data, right? Yes, but that is exactly Deming’s point. We look at data all the time. We compare it, we manage it, we get bonuses based on it, we get fired or promoted based on hitting a certain number. But we often forget the most important thing about it: its variation. See, data, numbers — they are just that when viewed in a fixed and isolated manner. Take the number of smartphones sold. Okay, I have a quarter with one million sales. I can say that’s more than last quarter. I can say that’s more than my target. But that is still limited information about my business. Now I can put it in relation to other data — let’s say how many people work in my stores, or how much I invest in marketing. That is useful if I build a system view with that — the previous lens — because I see how different elements influence each other. And I can look at the changes within my data, the variations. Because there are always smaller and larger changes over time, and other factors (location, department, etc.). These variations have different reasons, or causes. It was actually Shewhart — an influence on Deming — who defined that there are common causes and special causes, and how to tell them apart statistically. Common causes come out of the system itself, how it is set up, the elements, the interactions. Special causes are triggered, well, by special events, usually from outside the system. This is important to determine, because the actions to change them are fundamentally different. I know it sounds like all nerdy statistics stuff — it is — but it is also really important, because it gives you more than the comparative view you usually get when looking at data. It’s so important that we will also look at it in more detail in a future essay.

Next up is epistemology — ha, see, I can do big words too — or just: how does knowledge work, and how is it acquired? Deming goes back to C. I. Lewis for this, and his 1929 book “Mind and the World-Order.” Sounds grand, but he was a pragmatist philosopher, so it’s not that far up in the sky. And it actually connects to the data bit we just discussed. The whole common-cause and special-cause thing is a well-thought-through theory which makes sense of the data variations that we see. Without the theory we would just see spikes in a graph and would not have the slightest idea how to look at it and gather meaningful information — which essentially is what knowledge is. C. I. Lewis said that is always the case. Facts are important, of course. They are what we see, what we observe. But without a theory they do not contain more information than just a value. Five hundred thousand smartphones sold. Okay, nice, but what does that mean? If this is 10% higher than last year, and this year I lowered prices, I can form the theory that sales increased because I lowered prices — and yes, something as simple as this is also called theory, it doesn’t have to be something super complicated. With this theory I have more information than with the simple sales numbers alone. And I can make predictions for the future — maybe lower prices more to increase sales even further. The tricky thing is that this is not a one-time process. As I collect more facts I have to constantly compare them to my theory. Maybe I find out that this year my main competitor went out of business. Could this also have influenced my sales numbers? I could look, in this case, at the sales development from when he was still operating. It can go either way. It can support my theory, or it can cause me to revise it. Either way, I learn more, I gain additional knowledge, and my perception of reality becomes better. The value of theory is not just in explaining facts. It is about what more I need to look at, and sometimes it means I just need to change my mind, because I learned something better.

One additional point regarding facts and theory. And it comes again from Follett. She points out that theories are never neutral. She has a good example of this regarding flour. She says that when white flour is scarce, experts talk about how much more nutritional value brown flour has. When white flour is abundant, they say that white flour is more easily absorbed by the body. Neither of these theories is wrong. But it is obvious that they are used when one of the two results is more convenient. This does not mean that theory is bad altogether, but we should always ask ourselves about the intention behind a given theory. This gives us then even more knowledge.

The fourth lens is that of psychology. I admit I would not have mentioned that lens separately if I had made the list myself. For me, psychology is just another source of facts and theories. But Deming — maybe because he was an engineer by training — treated the field of human behavior and the mind as something special, worth naming as a different way to look at an important element in systems: the human being. Feelings like fear or love, motivation, learning abilities — these have an important role to play in systems. In my opinion these are just facts — however complicated to explain — that have to be incorporated into any perception of reality. That being said, I admit that humans in systems are something a bit odd. They are not just actors or agents — elements of the system — but they are also us, the observers, the analysts, the theory formers. Sometimes one person can even have both roles at the same time. Furthermore, psychology matters because motivation and feelings influence the perception of facts — they distort them, you could say. So it is really important to be aware of them when trying to understand an interconnected social system. And almost all relevant systems are at least partly social systems. That is also why I would change this last lens into psychology-slash-sociology, to incorporate human mechanics even more broadly, because systems are rarely purely individual-psychological — they live from the interaction, the social. That being said, Deming was most interested in two aspects of psychology: motivation and learning, and — you guessed it — both will get their own separate essay, because any perception of reality would be truly incomplete and truncated if these two aspects were missing.

So, summing up — sorry, I love it when authors use obvious formulations like this, it gives me an anchor — in order to avoid fixed and fictional snapshots of reality, we need to follow Follett and collaborate on fact-finding and discuss its meaning. In addition, we need the discipline of a multi-lens approach suggested by Deming: thinking of reality as an interconnected system, seeing the variations in the data, using theory together with the facts we find, and taking psychology and sociology into the picture.

Sounds like a lot?

“You’ll get used to it in time,” said the Caterpillar; and it put the hookah into its mouth and began smoking again.

Don’t just take it from me, here is some good stuff to read:

Carroll, Lewis. “Alice’s Adventures in Wonderland”. 1865.

Boorstin, Daniel J. “The Image”. 1962.

Goffman, Erving. “The Presentation of Self in Everyday Life”. Doubleday, 1959.

Barnard, Chester I. “The Functions of the Executive”. Harvard University Press, 1938.

Follett, Mary Parker. “Creative Experience”. Longmans, Green and Co., 1924.

Köhler, Wolfgang. “The Mentality of Apes”. English translation, 1925 (German original 1917).

Deming, W. Edwards. “The New Economics for Industry, Government, Education”. MIT Press, 1993.

Lewis, Clarence Irving. “Mind and the World-Order: Outline of a Theory of Knowledge”. 1929.

Shewhart, Walter A. “Economic Control of Quality of Manufactured Product”. 1931.

Meadows, Donella H. “Thinking in Systems: A Primer”. Chelsea Green Publishing, 2008.