60. On Decision-Making

A decision is the cognitive process of choosing between possible actions in a situation of uncertainty. By definition, the steps entailed lead to a final choice, that is, the selection of a sequence of activities among several alternative scenarios, based on values and preferences, purportedly resulting in a more optimal outcome.

In view of the resources organizations pool, decision-making permeates all dimensions of corporate life, be they (i) strategic—related to the design of a long-term plan of action to achieve a particular goal, (ii) organizational—related to the way different parts and aspects of a group are arranged to deliver the goal, or (iii) operational—related to the way individuals and groups work on a daily basis to accomplish specific results toward the goal.¹ It follows that decision-making is a conditioning ingredient of success in any venture. The stakes are high: organizations that make better, faster, and more effective decisions—both small, routine, and big, one-off decisions—will outrun competitors and outshine peers. Therefore, one might expect that organizations would put copious options on the table and invite sufficient evaluation to make certain the best choice emerges.

Inevitably, given the pressing omnipresence of decision situations, the world of organizations is not short of techniques. The main clusters articulate decision-making models, help choose between options, make financial decisions, improve decision-making, organize group decision-making, surface values and preferences, and decide whether to go ahead. Then again, whether the tools at hand are leveraged depends on styles that—born of the typology of the organization and associated configuration²—range from autocratic to unanimity-based decision-making, each with its raison d’être and related pros and cons (Figs. 60.1 and 60.2).

Ten primary criteria shape decision-making. These comprise the decision environment ³ that may influence the decision style, the complexity of the decision being made, the value of the decision’s desired outcome, alternative scenarios that have the potential to lead to the desired outcome, the information available to support the decision-making process and cognitive biases to its selection and interpretation, the quality requirements of the decision, the personalities of those involved in decision-making, the time available to conduct the decision-making process, the necessary level of commitment to or acceptance of the decision, and the impact on valued relationships that the choice of decision style may have.

Sure enough, several of these criteria can hold at the same time and amplify one another. Assuming organizations do not eschew problem analysis to rush decision-making—a big, hairy, and audacious hypothesis,⁴ that, four recurring themes regularly conspire to warp decisions. They have to do with bounded rationality, cognitive bias, personality, and free will. First, the information at hand, the information-processing ability of the mind, and what time is available bear strongly on decision-making. (Bounded rationality does not often conduce optimal decisions by “maximizers;” again and again, “satisficers” reach for what solution is good enough.) Second, cognitive biases creep into decision-making processes. (A select list includes anchoring and adjustment, attribution asymmetry, choice-supportive bias, framing bias, groupthink, incremental decision-making and escalating commitment, optimism or wishful thinking, premature termination of search for evidence, inertia, recency, repetition bias, role fulfillment, selective perception, selective search for evidence, source credibility bias, and underestimates of uncertainty and the illusion of control.) Third, personality profiles color cognitive styles. (Psychological traits revealed by the Myers–Briggs Type Indicator along four bipolar dimensions—extroversion and introversion, sensing and intuiting, thinking and feeling, and judging and perceiving—correlate with decision-making styles. In any organization, the predominance of one psychological type will sway approaches to decision-making. What is more, national or cross-cultural peculiarities exist across entire societies.) Fourth, advances in social neuroscience increasingly question whether and in what sense rational agents exercise control over their actions or decisions, thereby testing the easy presumption of free will. What hopes, after that, are there for better decision-making?

“It is hard to imagine a more stupid or more dangerous way of making decisions than by putting those decisions in the hands of people who pay no price for being wrong,” Thomas Sowell reasons. Indeed, many organizations treat decision-making as an event, the performance of which is more often than not the prerogative of a few—not necessarily best equipped—where there is obdurate proclivity for formal authority.⁵

To enrich decisions in the majority of decision-making styles, two related concerns should be concurrently raised and addressed:

Decision-making is where thinking and doing overlap. For that to happen profitably in an organization, a decision must be logically consistent with what the parties to it know, want, and agree they can do. Nothing, then, could do decision-making a greater disservice than to treat it as a single, isolated event, not the clearly defined process it inherently is or rather should be.

Helpfully, David Garvin and Michael Roberto compare advocacy and inquiry approaches to decision-making. Advocacy tends to push a single solution. To make a compelling case for the proposal they hope to sell, proponents assert positives and downplay negatives; they offer no alternatives—instead, a go-no-go decision on the option is forced. The pitfalls of advocacy are many: reliance on one solution precludes the chance to explore alternatives; personalities come into play and disagreements grow fractious, probably antagonistic; behind-the-scenes maneuvering comes into play; the solution inevitably produces winners and losers—losers, to the extent they can, continue to fight the decision in the execution phase, thereby stretching decision cycle time.⁷ In contrast, the goal of inquiry is to reach agreement on a course of action. Because people hold diverse interests, inquiry makes convictions visible for testing; generates multiple alternatives; evaluates feasibility according to well-defined criteria using a range of techniques; fosters collaboration to work through differences of ideas, concepts, and assumptions; and helps arrive at an agreeable solution. Rather than suppress dissent and debate, inquiry encourages constructive conflict, consideration, and closure with perceived fairness; patently, it produces decisions of higher quality—decisions that not only advance an organization’s objectives but are also reached in a timely manner and can be implemented effectively.

Marcia Blenko, Michael Mankins, and Paul Rogers make out six steps in which to (re)organize around decisions. First, organizations should know which decisions have a disproportionate impact on organizational performance—a decision inventory is a prerequisite to that.⁸ Second, they should determine where those decisions should happen. Third, they should organize the structure of decision nodes around sources of value. Fourth, they should figure out what level of authority decision-makers need, regardless of status, and give it to them. Fifth, they should align other parts of the organizational system, such as processes, data, and information—including their flows, measures, and incentives—to support decision-making and execution. Sixth, they should help managers develop the skills and behaviors necessary to make decisions and translate them into action quickly and well.

Decision rights are the coin of the realm. In a small organization, an entrepreneur might know all about his or her business and make every decision with minimal supplementary data and information. However, as the scale and scope of operations grow, an entrepreneur will find it more difficult to decide. In a world of large organizations, one solution to this problem is to convey data and information to whoever possesses decision rights; another is to grant decision rights to whoever holds data and information.⁹ To strike a balance, modern organizations turn to both solutions¹⁰: the falling prices of information and communications technology have cut the costs of transmission and the growingly intense use of these technologies in many organizations confirms they convey more data and information to those with decision rights; at the same time, the common reliance on teams and other collaboration mechanisms implies that organizations are decentralizing decision rights. Naturally, the mix of solutions—and the centralization or decentralization of relationships it implies—is unique to the organization: yet, it is still a rare organization—especially in the public sector—that actually studies the optimal allocation of its stock of decision rights and acts on that knowledge to reap the fullest advantage.

There is more: decentralization is one thing; a more differentiated chain of deliverables for decision-making is another. Establishing what helps pinpoint who and demarcate how. Rogers and Blenko (2006) offered early guidance before the last article cited: advertising a tool of Bain & Company, Inc., they further untangle the decision-making process by identifying various activities that must occur for a decision to be made well. The name of the tool is RAPID: each letter in the acronym stands for an activity associated with decision-making. To begin, someone must “recommend” that a decision be made. Next, “input” will likely be required to inform the decision. Down the road, depending on corporate governance arrangements, one or several persons will formally “agree” to a recommendation before one or several persons wield the authority to “decide”. Subsequently, someone must, of course, “perform” the decision, meaning, execute it. The acronym does not suggest a strict sequence in which the five activities must occur, certainly in the preparatory stages: reality is iterative and RAPID merely happens to be a handy mnemonic. (This writer, for instance, believes that inputs should precede any recommendation, not just follow it, and may actually be required throughout the process.)¹¹ In addition, agents may be assigned more than one activity. (The person recommending the decision may eventually be tasked with carrying it out.)

Paraphrasing Albert Camus, an organization’s value is the sum of the decisions it makes and executes. For sure, even good decisions such as may have been reached with RAPID occasionally engender bad outcomes. But redistributing decision rights because of a bad outcome—even though they were well allocated in the first place—will not help and could make matters worse. One should not confuse a particular outcome with the process itself.