Feature Article: Mapping Our Decision Making

Gareth Powell

Throughout human history, people have made decisions on a daily basis. The decisions they make reflect both the information they have and the choices they perceive. At both personal and group levels, better information and more choices generally lead to better decisions. So the question becomes: how can we get better information and more choices?

I’m going to use two analogies in this article to illustrate how the different decision making styles could play out.

Flying a plane
Flying a plane is a complex endeavor: the primary complexity is due to the fact that the plane exists in three real dimensions, and six effective dimensions: in order to go where you want to go, you not only have to be in the right place, right now, but you have to be heading in the right direction, without rolling or spinning. In order to do that, you have an array of instruments, for example, an artificial horizon, to an airspeed indicator and an altimeter. How do you use this collection of information to fly the plane where you want it to go?
Knights in Europe
In Medieval Europe, Knights were continually fighting and plotting against each other in an attempt to gain supremacy, often to the detriment of their country. How could they have more easily met their needs?


Possibly the simplest form of decision making is autocracy – decisions made by a single individual, untrammeled by process or law. Autocracy is founded on the principle that a single individual knows enough information to make the right decision. Is this true?

Actually, very often it is. Most of the decisions we make on a daily basis are of this type: when to get up, what order to do things in, how to get to work, which kind of coffee to drink. All these choices are personal, mainly affect only us, and, inasmuch as they affect others, usually do so in well understood ways where the constraints can be clearly seen.

When fulfilling a role in a company, within clearly defined limits, autocracy also works well. As a software developer, I write code, choose class names, and decide implementation strategies autocratically – otherwise, it would simply be too slow.

And that’s the great appeal of autocracy in all its manifestations – it is a very fast decision making process; it avoids red tape, communication overheads, argument, discussion, consideration or responsibility on anyone else’s part. The advantage of autocracy in an organization or country is that everyone else can simply abdicate responsibility and leave it to someone else.

The downside of autocratic decision making comes from the very simple observation that when two people are affected by the decision, and only one of them makes it, there are always going to be issues – or at least the potential for issues – regarding whether any decisions made are in the best interests of just the one individual making the decision or in the best interests of the group as a whole. And in a truly autocratic environment there is no established process for even questioning a decision, especially in a larger group.

When flying a plane, this would be like focusing exclusively on one instrument to guide you, probably the artificial horizon. Doing this would hopefully lead to a very stable, flat journey, but failure to take into account the compass heading would lead to drifting significantly off course.

In Medieval Europe, a single Knight could take any actions he chose within his own domain; within that scope he was the single, undisputed boss. He was able to organize his serfs to fulfill his needs. Of course, what the serfs thought of it was another matter.


In The Republic, Plato states that there is a progression of government forms which correspond to a progression of personal states. In this article, I will look at many of the same forms of government, but in the reverse progression, starting with autocracy and ending with “the ideal”.

In any society or organization, there cannot be two autocratic decision makers at the highest level (as discussed above, for smaller decisions any number of autocratic decision makers may co-exist). When two or more individuals wish to simultaneously be the one autocrat, it is inevitable that conflict in some sense will arise.

The Role ofScopeThere are two possible outcomes in a conflict-of-equals scenario. Firstly, they could come into conflict repeatedly – either physically or intellectually – until one of them is the victor. In this case, the victor becomes the autocrat and the governance system reverts to autocracy. In the meantime, the system is most generally described by the label “civil war”.

The other possibility is to build the conflict into the governance system directly. This was the case in the Roman Republic, where the great men of the day literally competed to be consul – effectively autocrat for a year. It is a pattern repeated today in many Western democracies where two parties – with radically different agendas – compete to push their agenda for a limited period of time. In each of these cases, there is no attempt to find any common ground, just an acceptance that there is some validity in both points of view.

Another instance of conflict is “debate”. The notion of a debate is that two points of view do not both have validity: one is right and the other is wrong. Strong enough advocacy of one point of view will expose not only its good points but the bad points of the other and convince everyone that there is only one correct choice.

Conflict was endemic in Mediaeval Europe. Kings and Knights would fight over anything and nothing: territory, honor or spoils. It’s not quite clear who benefited from this (perhaps the armor manufacturers?)

In a plane, it would be a disaster to allow two instruments to be in conflict: as they battled, you might consider the compass heading more important for a while, then the artificial horizon. During the conflict itself, both might be ignored. Sooner or later, the plane would no doubt crash.


From the attempt to avoid conflict arises the notion of trying to see that two opposing viewpoints could – in some way – both be correct. The simplest way of doing this is to try and drive straight down the middle: take each point of contention in turn and try and find something halfway between the actions advocated by one or the other party.

The well-known problem with compromise is that while either one or the other suggested decision might have been good – and even though both of them might have their individual merits – the final decision often has not only the good of each approach but also the bad.

Text Box: The Rule of Law One of the oft-noted problems with autocracy is the untrammeled nature of the autocrat. It is possible to address this aspect without changing the fundamental structure by simply placing limits around what the autocrat can and cannot do: executing citizens without trial is not permitted; spending company money over $100 is not permitted without a second signature. The role of laws or limits does not apply just to autocracy. Laws can be used to constrain the permitted natures of “conflict” – which often leads warlike conflict to stabilize into a more permanent “competition” or “debate” conflict. Limits can also be used in more enlightened organizations to facilitate effective decision making: for example, in an organization run by consensus, a limit around a particular functional area could allow one individual to make rapid, effective, autocratic decisions in that area. A less well-recognized problem with compromise is that generally the conflict that is being resolved is between two points of view that have already compromised many times before – consider the effort that each party puts into choosing a presidential candidate in the US every four years – and the final compromise represents not just a viewpoint that nobody holds but a position halfway between two positions that nobody particularly relates to.

In a plane, the idea of compromise would be simply idiotic: imagine that the airspeed indicator and altimeter both wanted to increase as much as possible. They would attempt to compromise, both increasing at the same rate; but as the plane climbed, it would necessarily lose speed. As they jockeyed to keep the compromise working, it would also lose some height. Eventually, in trying to keep everything fair, the plane would stall and crash.

Mediaeval knights often tried compromise as a way out of continual conflict, but the compromises rarely stuck. Since neither party truly believed that they shouldn’t be the more powerful, it was usually “just a trick” to gain time to regroup.


To address the problems with compromise losing information about how most people really feel, consensus can be used as a decision-making style. Essentially consensus consists of the “veto” – each individual can refuse to accept anything that does not meet their needs; the eventual decision is taken based on what few options remain.

The problem here is that none of the fundamental issues with conflict have gone away; multiple perspectives are represented and come into conflict, with the “best” decisions – the ones that have the most inherent value – often being the first ones to be eliminated.

In the best case, consensus offers an opportunity to share perspectives and to come to a mutual agreement as to the best way forward; but generally it is just a very slow way of reaching a poor decision. Fundamentally, consensus consists of eliminating possible choices, and as such is doing damage to the process of finding the best possible choice.

Compared to the other mechanisms described above, consensus has the advantage in flying a plane that all the instruments are taken into account, and any instrument can hold up a decision by pointing out the problems that will be caused by it: for example, most crashes could theoretically be averted by the altimeter refusing to drop below, say, 1000 feet. But the problem is that there is no threshold on how long the process can be held up: the altimeter could refuse to go along with any decision that left it below 45,000 feet. – even if the plane would stall at that height. Furthermore, in the time that is taken to negotiate a decision, anything might have happened to the plane.

As far as I know, the Knights of Medieval Europe never considered or tried consensus. It would be kind of amusing if they had though. I envision the tournament grounds at Ashby-de-la-Zouch, with banners fluttering and tables stacked with wild boar and moorhens while the knights in their armor attempted to reach a consensus decision on the borders between all their respective fiefdoms. It seems to me that the only way this could ever have worked would have been to incorporate other, simpler decision making forms – for example, two knights who could not agree could have gone and had a quick joust; the loser forfeiting his lands to the winner.

ntegrative Decision Making

It doesn’t take much to do better, and many attempts have been made to improve the process. One such approach is found in Holacracy: Integrative Decision Making. In this approach, a contextually relevant committee of individuals meets to explore, as carefully as possible, all the issues and dimensions associated with the decision to be made.

Having explored the territory, the committee then attempts to express how to make such decisions, in general, by creating the appropriate autocratic roles, defining both accountabilities and limits, and then selecting the most suitable individual or individuals, whether on the committee or not, to fill those roles.

Obviously some decisions, on a case-by-case basis, cannot be made entirely autocratically, regardless of the limits placed on the individual roles. In those cases, the individuals filling the roles that have “part” of the decision need to meet and integrate in the same way to produce a single instance decision.

Text Box: Committees Committees are orthogonal to decision-making styles. Although often blamed for the faults of decision-making (“the camel is a horse designed by committee”) it is usually just the style of decision-making used by the committee – conflict (represented by majority voting), compromise or consensus – that leads to the poor decision; the function of the committee is to use representatives to reduce the number of actual individuals presenting viewpoints. While the word committee tends to conjure up the notion of a half-dozen old men, the reality is that many organizations – and particularly governments – work this way. The British parliament is a committee of elected representatives who are chosen because of the viewpoints they say they intend to represent in parliament. Mathematically, a committee can be compared to a statistical sampling function. Each member of the committee represents the viewpoints of some set of individuals in the overall population. How effective the committee can be is most fundamentally determined by how that group of individuals reaches decisions, but it is also constrained by how representative of the population at large the viewpoints expressed by the members of the committee are. As with mathematical sampling functions, it is important that the represented views are statistically significant. It would be interesting to research how the statistical accuracy of a sampling committee is compared to the population at large on a range of issues. This metric could be used to improve the performance of representative governments by compensating for it. Making the Map

The first step in making a decision is to explore the territory in which the decision exists. Edward DeBono refers to this as “making a map”. Instead of each person presenting a proposed solution to “the problem”, each person brings their understanding of the situation: what benefits could be achieved by a decision; what problems could be caused by a decision; what the impacts of various decisions might be.

By presenting information in this way before attempting to present solutions, it is much easier to reach agreement. It is possible, before any solutions have been proposed to agree on the context of the decision: the challenges and opportunities faced; the possible risks and rewards.

When flying a plane, it is valuable to collect information from all the instruments before deciding what to do: any one instrument could give a false impression of what is happening and cause a crash.

Unfolding the Onion

Any decision has a number of facets. Some are clearer than others; some are closer to the surface than others. As the territory is explored, more and more details may emerge which increase the possible range of decisions.

When two people take opposing viewpoints, it is very easy for a conflict situation to occur. Like the blind man with the elephant, they are seeing the same situation from different perspectives. Even recognizing the situation for what it is may be insufficient to resolve it.

In order to make progress, it is necessary to “unfold” the points of view and investigate more deeply what assumptions and constraints are embedded in the respective viewpoints. As the assumptions are exposed, they add to the richness of the map and give extra “control levers” that can be pulled.

If you are trying to fly at 10,000 feet and you notice that your elevation has dropped, it’s not necessarily the case that you immediately want to pull the stick back to climb. It may, instead, on consultation with the airspeed indicator, be necessary to do the opposite and dive slightly to pick up enough speed to climb back up.

What is the Right Decision?

Before a specific right decision can be made for a specific issue, it is necessary to consider what the “shape” of a right decision looks like.

Obviously everyone wants to make the “best” decision. But not everyone agrees on what the “best” decision looks like. In particular, “best” often has an element of “best for me” or “best for what I need” in it.

The best decision, at any given point in time, must be one which meets all the constraints that are visible and represented at that time. Furthermore, the weight that is given to any one perspective is a function of how tightly that perspective is constrained at that time.

Decisions can, in reality, be reviewed at any time. When a decision is reviewed, it will be reviewed in light of the constraints that are applicable at the time of review. This gives rise to the notion of dynamic steering.

When flying a plane, you do not simply point the plane in the direction you want to go and hope for the best. You continuously consult all the instruments, and make small, incremental decisions that nudge you nearer and nearer your goal: just because you lose height as you reach your destination, does not mean it was a mistake to gain it when you left your airport of origin: just that the circumstances are now different.

Reaching the Decision

Once all the information is out in the open and agreed on, reaching the best available decision is usually fairly rapid. With agreement on how the risks and rewards stack up, it is generally much easier to reach agreement on how to maximize the rewards while minimizing the risks. And, by as much as possible trying not to solve individual problems, but rather assigning the solution of the individual cases to an appropriate role, the decision making process is dramatically streamlined.

Structural Decisions

It is worth emphasizing the difference between individual decisions and structural decisions. Any single decision is most easily resolved autocratically. In order for this to work effectively, the organizational structure must provide for fine-grained “roles” that have clearly delineated areas of responsibility – or accountability – and equally clearly delineated limits on how those accountabilities may be met.

In order to tune the system, the various committees must meet on a frequent basis to discuss the current set of roles and accountabilities, and to consider how they are handling the individual decisions that are occurring on a daily basis. Whenever a particular kind of decision is being made repeatedly in the organization, the committee should consider how it can be most expeditiously handled and create new accountabilities, roles and limits to handle the decision.

Choosing the Right Committee

The right decision can only be reached once the right information is on the map; and the map can only be properly completed by having all the relevant viewpoints expressed, unfolded and integrated, so the selection of the right individuals is important.

It is beyond the scope of this article to discuss the aspects of Holacracy relating to structural organization that enable easy assembly of committees at every scope; however, a few points may be briefly made.

Any given scope, or department, will generally have a “head”: one individual who is filling a role which includes the accountability for ensuring that the department as a whole is effective. This individual must be included in the committee for that scope.

In most scopes, this individual will have a number of “direct reports” who often head up sub-departments. In general, all of these individuals will be included on the committee for that scope.

Where sub-departments exist, each sub-department should elect one or two individuals to represent, as best they are able, the context and constraints of that sub-department. The purpose of having these individuals on the committee is to share information on the nature of the territory during the map-making phase, ensuring that relevant information is on the table before decisions are reached.

Again referring to our analogy of flying the plane, it’s important to have the right information to inform a decision. It is no use taking a decision on banking the plane without consulting the artificial horizon; it could be taken without consulting the fuel gauge. Whether to stop and refuel, however, would have much more input from the fuel gauge that it would from the altimeter.

As with consensus, I struggle to see the Knights of Medieval Europe capable of integrating multiple viewpoints by themselves. But it may well have been possible to do this with a sufficiently talented (and well-armored?) facilitator. In particular, the birthplace of modern democracy – Runnymede, in England – saw an abrasive minor knight facilitate a far-seeing agreement between King and Barons that was just enough to resolve the current crisis while setting Britain on the road to full democracy – a process that would take another six hundred years.


Sometimes there are decisions that simply cannot be delegated to one individual filling a role, but do not require the integration of all the viewpoints of a full committee. In these cases, a smaller committee – the smallest group of people who can represent all the relevant viewpoints – can be formed to handle this accountability.

Note that this is not simply a sub-committee of some larger committee (although it might be), but is a re-sampling of the entire organization to find the smallest group of people who can represent the relevant viewpoints. These groups are called Collaboration Groups (or “Chalupas”). In general, Chalupas are relatively short-lived; they are formed to tackle one decision, or achieve one goal. Their purpose accomplished, they disappear. A common use for Chalupas is to take ownership of deciding some thorny but tractable issue—the redesign of an office layout for example.

Some Chalupas can exist on an ongoing basis however. An organization could have a “Billing Chalupa” which is responsible for deciding issues related to both the billing model and making sure that billing runs effectively. With representatives from Sales, Invoicing, Accounts Receivable and Collections as well as Operations, the Chalupa cuts across the entire organization. But by having only a small group of representatives—and not necessarily the heads of the various departments—the Chalupa is able to rapidly reach decisions affecting the whole company with the knowledge that each part of the company was adequately represented.


In the modern world, leadership is usually viewed as a personal attribute. We look to individuals to provide leadership. What do we actually mean by this? It’s a reflection of our decision-making styles: we say that this individual is one who makes decisions with which we agree and that we can follow. But this is itself an evolutionary process: we have history to suggest that an individual makes decisions we trust, and then we follow subsequent decisions based on that trust. Ironically, it is possible for individuals to take essentially the same actions in very similar situations and be heralded as “a visionary leader” in one case and “a dictatorial tyrant” in the other.

But leadership is a quality of systems as much as it is of individuals. True leadership comes from a deep understanding of both reality and the fundamental mission of the organization—what reality demands of it at that moment; its life purpose, as it were.

This leaves an organization needing a system or process to uncover its own purpose and to interpret current reality in light of that purpose. That understanding is the true leadership of the organization. Holacracy not only offers that process, but because of the way in which decisions are reached produces, as an output, the buy-in and trust of decisions which is normally the required input to implementing decisions. Holacracy moves the focus from following individuals with “leadership skills” to following the decisions that emerge from a transparent, integrative process.



Decision-making is often viewed as intrinsically hard; “political” is often used as a pejorative to imply that making a decision is made deliberately hard. However, shifting perspective to see decision making as a collaborative—rather than competitive—activity unleashes creativity and openness.

Agreeing first on the context of a decision—why it is a decision needs to be made and what the possible outcomes might include—is a good start on the way to agreeing a decision: not only do the parties quickly get to the point of agreement, but the ongoing discussion and search for a solution are based in a common understanding of the landscape.

Decisions are made most effectively by individuals or small groups. Unleashing this creativity by creating properly delegated autocratic roles or collaboration groups within appropriate limits frees the organization from spinning its wheels using heavyweight processes to make simple, repetitive decisions.

Formalizing the decision-making process to ensure that as much as possible each decision addresses a set of decisions and that the relevant viewpoints are represented in each decision streamlines the process, enabling rapid resolution of even knotty issues.

Gareth Powell is a technical lead at Ternary Software Inc. Along with twenty years of developing software systems and tools, he is an experienced Holacracy facilitator. He was the facilitator of the committee representing the whole company for over a year, and has made a number of contributions to the development of the practices of Holacracy during his eighteen months with the company.