One useful way to consider alternative approaches to decision-making is through the intersection of two dimensions: the degree of autonomy for the one proposing the decision and the speed of the process to get the decision accepted by those most impacted. That’s where this essay is headed, but first, I’ll lay out the story of how I learned what I now know about decision-making.

Subscribe now

I’d really like readers to sign up for a paid annual subscription, so for the present time, I have dropped the annual subscription to $30. Note that I’ve also raised the monthly subscription to $10 from $6. Give annual a try. The biggest value is years of posts behind the paywall, and of course, seeing new posts in their entirety. However, for the Memorial Day holiday, this post is available for all.

My Story

When I first started in the world of business decades ago, I had formal training only in my discipline -- computer science -- and no management training whatsoever1. And the first few companies I worked for had well-defined decision-making approaches but did not lay them out very clearly. Apparently, I was supposed to learn them by osmosis, on the job.

The thinking then was that there were basically two sorts of decision-making. Senior executives made big decisions autocratically as the head of any organization or sub-organization. That might come after discussions with other executives (and perhaps the board) and subordinates. Alternatively, small decisions were made by individuals—if truly small—or, more generally, by a group or team through consensus.

A few years later, when I became a manager and then president of a software company, Meridian, I still had no formal training in management. As a young leader of a software startup with a background in computer science (and only a few years of experience leading small technical groups), I managed a 25-person company as a few overlapping teams, and largely let the heads of those teams determine how decisions were made in their own domains.

Personally, I was constitutionally disinclined from the most primal of decision-making approaches: autocracy. I didn’t want to make major decisions unilaterally, especially since several team leaders owned more stock than I did. I relied on the team leaders — the head of sales, programming, and operations — to work with me in a model of consensus-oriented decision-making. One of us would propose some company-critical decision — like the merger that we eventually executed with our largest (and publicly-traded) competitor, or taking on a major contract — and if an initial straw poll vote failed to pass unanimously, we’d talk the issues through until we all agreed on the decision and its ramifications. A slow but thorough process. But we, as a group, were confronted only with existential problems; simpler issues were handled by other teams.

The programmers adopted what I would now characterize as a consent approach to decision-making. Consent is an approach to group decision-making that short-circuits the hard work of consensus by limiting objections: members of the group can withhold consent, but only for a well-defined reason. The proposer can counter an objection by amending the proposal and at some point the objector assents. Once all objections are resolved — when all can live with the proposed decision — consent is achieved.

As Ted Rau characterizes it:

Consent is defined as ‘no objection’. Not having an objection is slightly different from agreeing. We refer to that extra space as the range of tolerance. We don’t have to find the overlap of our preferences in order to make a decision. Instead, we seek the overlap of our ranges of tolerances which gives us much more to work with.

It’s strange that I haven’t really learned much new about group decision-making in the decades since running that software company. (If you were waiting for the end of the story, yes, we merged with our largest competitor, and I become head of engineering for a 175-person software company, Verdix, that merged with another firm to form Rational Software, later acquired by IBM.)

However, a few years ago I learned about an approach I hadn’t encountered before. The advice process involves a great deal of autonomy for the decision proposer while retaining a means for collective intelligence to influence the decision. I read about it in 2017, from Richard Bartlett:

Anyone can make any decision, so long as they are willing to take responsibility for the outcome, and they have first listened to input from anyone who will be affected, or who has relevant expertise.

Notice it says listened to, not agreed with. If your relationships are good, this gives you most of the benefits of consensus, at a fraction of the cost.

Radical autonomy! It seems like a short-cut consent approach, with the obligation on the proposer to overcome objections removed. An element of autocracy for everyone.

Bartlett makes the point that the advice process needs to operate in the long run in the context of other organizational feedback cycles, so the responsibility element is analyzed and learned from. I confess I have never worked in an organization using the advice process as a regular, general approach to decision-making, but I am sold on the principles.

Many ideologies have arisen in the intervening years about the relationships between decision-making and organization design — such as Sociocracy, Holacracy, Frederic Laloux’s Teal, and a number of others. For today, though, I am limiting my discussion to the nuts-and-bolts of the decision-making approaches, and their differences.

I think of decision-making at this scale as the mitochondrion of the organization, the organelle in our cells that creates energy. Teal, Holacracy, and other theories of organization are aspirational rather than focused on the nuts-and-bolts, bottoms-up operations of small-group decision-making. I’ll leave larger-scale organizational forms as a topic for a later day.

Share

Comparing The Approaches

At the outset, I offered a simple approach to contrast various approaches to decision-making: the degree of autonomy for the proposer, and the rate of speed of the decision process.

Here’s a chart arraying the approaches in a 2-D matrix:

As I said earlier, autocracy is often considered the original form of decision-making, but I doubt that. People are much more social than we generally admit. However, there is no doubt that many top-down, command-and-control structures in the past were quite autocratic. Meanwhile, many theoretical approaches to organizational design — I am thinking specifically about Laloux’s Teal model — have a socio-evolutionary aspect that associates forms of decision-making with levels of organizational advance. For example, a Teal-level organization theoretically relies on the advice model. But I will leave the evolution of organizations to another post.

Autocracy can be a very fast decision-making process, perhaps to its detriment. Daniel Kahneman in Thinking, Fast and Slow offers a great deal of research that demonstrates that slower decision-making is more likely to lead to good decisions.

The advice model is a high-autonomy and medium-speed approach. Decision makers in this approach can ignore advice if they want, but Kahneman’s warning applies to them as well. The larger the number of people interviewed the slower the process, but the likelihood of satisfactory outcomes rises.

The conceptually democratic majority voting approach, like the straw polling I used with my direct reports — can lead to faster or slower decision-making depending on the result. At Meridian, I was looking for unanimity, but a majority approach could be based on a simple majority, or some specific level of agreement, like a supermajority of 66% or 75%. Obviously, this has to be established at the outset, and is perhaps most useful for large groups and decisions of low impact, like ‘should we order Chinese or pizza for this week’s company meeting?’.

Consensus is the slowest decision-making approach, and that might be justified if it uniformly leads to better decisions, but it does not seem to. It tends to smooth off all the potentially revolutionary or highly innovative aspects of proposed plans because many people are risk averse. In my estimation, consensus should be reserved for the narrow set of decisions that directly impact all the members of the group and require all to actively support the decision going forward.

My recommendation is that when considering a decision-making approach, look toward the center of the table. Is there a reason consent should not be used? It is relatively fast, relatively democratic (all can object if justified), and relatively autonomous. If greater autonomy is needed or greater speed is necessary because of external forces, the advice approach might be better.

Autocracy, consensus, and voting approaches should be reserved only for edge cases, which are either very low stakes (pizza versus Chinese food), for very small teams, or individuals autocratically making decisions about their own solitary work.

Individuals make their own small-scale decisions — like ‘which of the things on today’s to-do list should I do first?’ — as autocrats, or in an advice process, perhaps asking a colleague’s thoughts on solving a programming bug.

Come to think of it, perhaps that’s why we think autocracy is primal.

Share workfutures.io

Originally published at Sunsama.

…

The human condition can almost be summed up in the observation that, whereas all experiences are of the past, all decisions are about the future. It is the great task of human knowledge to bridge this gap and to find those patterns in the past which can be projected into the future as realistic images.

| Kenneth Boulding

Decision-making is critical to business performance, so much more time should spent on systematically and rigorously examining the factors that can impede it. And many of those factors are in our wiring, both as individuals and as organizations.

If Decision-Making Is So Important, Why Do We Take Shortcuts?

Decision-making is critical to business performance. To go even further, it is an existential requirement. Yet so often things go wrong. Why?

Some problems in strategic decision-making seem obvious when you step back far enough. Consider the ‘loss aversion’ problem. Operating unit managers are focused on short-term timeframes and therefore tend to take on only small risks, instead of larger ones that may contribute more to long-term corporate growth. They are more concerned about the possibility of a status decline if things go wrong: the potential loss for the manager may be perceived as worse than any potential upside for the company and the manager.

Senior managers, however, must take a portfolio view and therefore must anticipate a portion of projects to fail in the search for a few that will yield big returns. The incentives — and fear of downside risk — between corporate goals and the individual manager’s goals are mismatched.

The important takeaway is this: the economic downside of loss aversion is only one instance of a much larger case of cognitive biases that distort human behavior whenever decision-making is called for. And like so many other aspects of life, avoiding examination of these influences on our behavior leads to big problems.

The basic assumption is wrong.

The basic assumption in business is that strategic decision-making requires three basic elements:

1. fact-gathering and analysis,

2. the insights and judgment of a defined group of people (stakeholders or advisors), and

3. some process -- ranging between very formal to very informal -- for that group to make a decision, reflecting that analysis and judgment.

However, as Dan Lovallo and Olivier Sibony reveal,

Our research indicates that, contrary to what one might assume, good analysis in the hands of managers who have good judgment won’t naturally yield good decisions.

And why is that? The authors go on to draw our attention to the third part: the process. After extensive analysis of 1,048 major decisions over a five-year period, including investments in new products, M&A decisions, and large capital expenditures they determined that

process mattered more than analysis—by a factor of six.

Process, by extension, is also more valuable than the judgment of those involved in the decision-making, too, since flaws in both analysis and judgment are countered by processes designed to do exactly that.

Process, process, process.

Lovallo and Sibony took a hard look at decision-making effectives, and broke things down:

And a major value of good decision-making processes are to overcome — or at least moderate — cognitive biases that lead to poor decisions. As Lovallo and Sibony put it:

The prevalence of biases in corporate decisions is partly a function of habit, training, executive selection, and corporate culture. But most fundamentally, biases are pervasive because they are a product of human nature—hardwired and highly resistant to feedback, however brutal. For example, drivers laid up in hospitals for traffic accidents they themselves caused overestimate their driving abilities just as much as the rest of us do.

Improving strategic decision making therefore requires not only trying to limit our own (and others’) biases but also orchestrating a decision-making process that will confront different biases and limit their impact. To use a judicial analogy, we cannot trust the judges or the jurors to be infallible; they are, after all, human. But as citizens, we can expect verdicts to be rendered by juries and trials to follow the rules of due process. It is through teamwork, and the process that organizes it, that we seek a high-quality outcome.

Structuring decision-making processes helps, but we need to be vigilant about biases.

Perhaps the core message to be gained from Lovallo and Sibony’s research is this: while awareness of our cognitive blinders is helpful, the best approach to debiasing is to create processes that structure decision-making to intentionally counter biases to the greatest extent possible.

Here are five of the most common groups of biases relevant to business decision-making.

Pattern-recognition biases — We are pattern recognition machines: it is a defining characteristic of our species. But there are some side effects of our pattern-matching prowess that skew our reasoning. For example, confirmation bias leads us to discount new information when it doesn’t support an initial hypothesis. As Sydney Finkelstein, Jo Whitehead, and Andrew Campbell detailed in Think Again: Why Good Leaders Make Bad Decisions and How to Keep It from Happening to You, pattern-matching can lead to huge mistakes [emphasis mine]:

Most of the time, pattern recognition works remarkably well. But there are important exceptions. If we are faced with unfamiliar inputs— especially if the unfamiliar inputs appear familiar-we can think we recognize something when we do not. We refer to this as the problem of misleading experiences. Our brains may contain memories of past experiences that connect with inputs we are receiving. Unfortunately, the past experiences are not a good match with the current situation and hence mislead us.

Another exception is when our thinking has been primed before we receive the inputs, by, for example, previous judgments or decisions we have made that connect with the current situation. If these judgments are inappropriate for the current situation, they disrupt our pattern recognition processes, causing us to misjudge the information we are receiving. We refer to these as misleading prejudgments.

In other words, our pattern recognition process is fallible. We have all experienced the embarrassment of accosting a complete stranger we thought we recognized. We have also experienced some degree of terror because we have misjudged the sharpness of a bend in the road or the speed of an oncoming car. What is less obvious is that our pattern recognition processes can also let us down when we are trying to judge the severity of a financial crisis, the value of an acquisition target, or the threat from an incoming hurricane.

Counter: One technique is to undermine the presuppositions that lead to the pattern match in the first place. Why are the team members so certain about a particular course of action? Explicitly draw our what is the underlying analogy between this circumstance and others from the past. One means is to dramatically enlarge the set of comparison scenarios, and supply known facts for each example.

Action biases — There is a well-known bias toward action among managers which can lead to accepting overly optimistic analysis and yielding to cultural pressures to make decisions quickly without accurately assessing uncertainties.

Counter: Promote the recognition of uncertainty rather than trying to squeeze it out of the conversation. Tools like scenario planning, decision trees, and Gary Klein’s ‘premortems’ explore the many dimensions of uncertainty before pressing ahead with a decision.

Read more

RACI is vague, hard to use, and reinforces the "what the hell is happening here" status quo. DICE is specific, easy to use, and shines a bright light on dysfunction.

A…

Read more

How wonderful it is that nobody need wait a single moment before starting to improve the world.

Anne Frank, The Diary of a Young Girl

Read more

Read more