What you dont know about making decisions by peterzhangonline

VIEWS: 137 PAGES: 11

									              A R T I C L E


                                 What You Don’t Know
                                 About Making
                                 by David A. Garvin and Michael A. Roberto

                                 Included with this full-text Harvard Business Review article:

                               1 Article Summary
                                 The Idea in Brief—the core idea
                                 The Idea in Practice—putting the idea to work

                               2 What You Don’t Know About Making Decisions

                              10 Further Reading
                                 A list of related materials, with annotations to guide further
                                 exploration of the article’s ideas and applications

                                 Product 5399
                                                                                                                                         What You Don’t Know About Making

                                                                                        The Idea in Brief                                The Idea in Practice
                                                                                        The quality of a leader’s decisions can make     CONSTRUCTIVE CONFLICT                                  views. Avoid indicating you’ve already
                                                                                        or break him. Yet most of us get decision        Conflict during decision making takes two              made up your mind.
                                                                                        making all wrong. Why? We take the least         forms: cognitive (relating to the substance of       • During the discussion, listen attentively.
                                                                                        productive approach: advocacy. We argue          the work) and affective (stemming from inter-          Make eye contact and show patience while
                                                                                        our position with a passion that prevents us     personal friction). The first is crucial to effec-     others explain their positions. Take notes,
                                                                                        from weighing opposing views. We down-           tive decision making; the second, destructive.         ask questions, and probe for deeper expla-
                                                                                        play our position’s weaknesses to boost our      To increase cognitive conflict while decreas-          nations.
                                                                                        chances of “winning.” And we march into          ing affective:
                                                                                        decision-making discussions armed for a                                                               • Afterward, explain the rationale behind
                                                                                        battle of wills. The consequences? Fractious     • Require vigorous debate. As a rule, ask              your decision. Detail the criteria you used
                                                                                        exchanges that discourage innovative               tough questions and expect well-framed               to select a course of action. Spell out how
                                                                                        thinking and stifle diverse, valuable view-        responses. Pose unexpected theoretical               each participant’s arguments affected the
                                                                                        points.                                            questions that stimulate productive think-           final decision.
                                                                                        Contrast advocacy with inquiry—a much
                                                                                                                                         • Prohibit language that triggers defensive-         CLOSURE
                                                                                        more productive decision-making ap-
                                                                                        proach. With inquiry, you carefully consider       ness. Preface contradictory remarks or             In addition to stimulating constructive conflict
                                                                                        a variety of options, work with others to dis-     questions with phrases that remove blame           and showing consideration, bring the deci-
                                                                                        cover the best solutions, and stimulate cre-       and fault. (“Your arguments make good              sion process to closure at the appropriate
                                                                                        ative thinking rather than suppressing dis-        sense, but let me play devil’s advocate for a      time. Watch for two problems:
                                                                                        sension. The payoff? High-quality decisions        moment.”)
                                                                                                                                                                                              • Deciding too early. Worried about being
                                                                                        that advance your company’s objectives,          • Break up natural coalitions. Assign people           dissenters, decision participants may readily
                                                                                        and that you reach in a timely manner and          to tasks without consideration of traditional        accept the first plausible option rather than
                                                                                        implement effectively.                             loyalties. Require people with different in-         thoughtfully analyzing options. Unstated
                                                                                        Inquiry isn’t easy. You must promote con-          terests to work together.                            objections surface later—preventing coop-

                                                                                        structive conflict and accept ambiguity.                                                                erative action during the crucial implemen-
                                                                                                                                         • Shift individuals out of well-worn
                                                                                        You also must balance divergence during                                                                 tation stage.
                                                                                                                                           grooves. During decision making, ask peo-
                                                                                        early discussions with unity during imple-         ple to play functional or managerial roles           Watch for latent discontent in body lan-
                                                                                        mentation.                                         different from their own; for example,               guage—furrowed brows, crossed arms, the
                                                                                        How to accomplish this feat? Master the            lower-level employees assume a CEO’s per-            curled-up posture of defiance. Call for a
                                                                                        “three C’s” of decision making: conflict,          spective.                                            break, encourage each dissenter to speak
                                                                                        consideration, and closure.                                                                             up, then reconvene. Seek input from peo-
                                                                                                                                         • Challenge stalemated participants to re-
                                                                                                                                                                                                ple known for raising hard questions and
                                                                                                                                           visit key information. Ask them to examine
                                                                                                                                                                                                offering fresh perspectives.
                                                                                                                                           underlying assumptions and gather more
                                                                                                                                           facts.                                             • Deciding too late. Warring factions face off,
                                                                                                                                                                                                restating their positions repeatedly. Or,
                                                                                                                                         CONSIDERATION                                          striving for fairness, people insist on hearing
                                                                                                                                         To gain your team’s acceptance and support             every view and resolving every question
                                                                                                                                         of a decision-making outcome—even if                   before reaching closure.
                                                                                                                                         you’ve rejected their recommendations—en-            To escape these endless loops, announce a
                                                                                                                                         sure that they perceive the decision-making          decision. Accept that the decision-making
                                                                                                                                         process as fair. How? Demonstrate consider-          process is ambiguous and that you’ll never
                                                                                                                                         ation throughout the process:                        have complete, unequivocal data.
                                                                                                                                         • At the outset, convey openness to new
                                                                                                                                           ideas and willingness to accept different

                                                                                                                                                                                                                                        page 1
                                                                                                                             Decision making is arguably the most important job of the senior
                                                                                                                             executive and one of the easiest to get wrong. It doesn’t have to be that
                                                                                                                             way—if you look at the process in a whole new light.

                                                                                                                             What You Don’t Know
                                                                                                                             About Making
                                                                                                                             by David A. Garvin and Michael A. Roberto

                                                                                                                             Leaders show their mettle in many ways—set-        read reports, mull some more, then say yea or
                                                                                                                             ting strategy and motivating people, just to       nay and send the organization off to make it
                                                                                                                             mention two—but above all else leaders are         happen. But to look at decision making that

                                                                                                                             made or broken by the quality of their deci-       way is to overlook larger social and organiza-
                                                                                                                             sions. That’s a given, right? If you answered      tional contexts, which ultimately determine
                                                                                                                             yes, then you would probably be surprised by       the success of any decision.
                                                                                                                             how many executives approach decision mak-             The fact is, decision making is not an event.
                                                                                                                             ing in a way that neither puts enough options      It’s a process, one that unfolds over weeks,
                                                                                                                             on the table nor permits sufficient evaluation     months, or even years; one that’s fraught with
                                                                                                                             to ensure that they can make the best choice.      power plays and politics and is replete with
                                                                                                                             Indeed, our research over the past several         personal nuances and institutional history; one
                                                                                                                             years strongly suggests that, simply put, most     that’s rife with discussion and debate; and one
                                                                                                                             leaders get decision making all wrong.             that requires support at all levels of the organi-
                                                                                                                                The reason: Most businesspeople treat deci-     zation when it comes time for execution. Our
                                                                                                                             sion making as an event—a discrete choice          research shows that the difference between
                                                                                                                             that takes place at a single point in time,        leaders who make good decisions and those
                                                                                                                             whether they’re sitting at a desk, moderating a    who make bad ones is striking. The former rec-
                                                                                                                             meeting, or staring at a spreadsheet. This clas-   ognize that all decisions are processes, and
                                                                                                                             sic view of decision making has a pronounce-       they explicitly design and manage them as
                                                                                                                             ment popping out of a leader’s head, based on      such. The latter persevere in the fantasy that
                                                                                                                             experience, gut, research, or all three. Say the   decisions are events they alone control.
                                                                                                                             matter at hand is whether to pull a product            In the following pages, we’ll explore how
                                                                                                                             with weak sales off the market. An “event”         leaders can design and manage a sound, effec-
                                                                                                                             leader would mull in solitude, ask for advice,     tive decision-making process—an approach we

                                                                                        harvard business review • september 2001                                                                                           page 2
                                                                                               What You Don’t Know About Making Decisions

                                          call inquiry—and outline a set of criteria for as-     tion will emerge from a test of strength among
                                          sessing the quality of the decision-making pro-        competing positions. But in fact this approach
                                          cess. First, a look at the process itself.             typically suppresses innovation and encour-
                                                                                                 ages participants to go along with the domi-
                                          Decisions as Process: Inquiry Versus                   nant view to avoid further conflict.
                                          Advocacy                                                   By contrast, an inquiry-focused group care-
                                          Not all decision-making processes are equally          fully considers a variety of options and works
                                          effective, particularly in the degree to which         together to discover the best solution. While
                                          they allow a group to identify and consider a          people naturally continue to have their own
                                          wide range of ideas. In our research, we’ve            interests, the goal is not to persuade the group
                                          seen two broad approaches. Inquiry, which we           to adopt a given point of view but instead to
                                          prefer, is a very open process designed to gen-        come to agreement on the best course of ac-
                                          erate multiple alternatives, foster the ex-            tion. People share information widely, prefera-
                                          change of ideas, and produce a well-tested so-         bly in raw form, to allow participants to draw
                                          lution. Unfortunately, this approach doesn’t           their own conclusions. Rather than suppress-
                                          come easily or naturally to most people. In-           ing dissension, an inquiry process encourages
                                          stead, groups charged with making a decision           critical thinking. All participants feel comfort-
                                          tend to default to the second mode, one we             able raising alternative solutions and asking
                                          call advocacy. The two look deceptively simi-          hard questions about the possibilities already
                                          lar on the surface: groups of people, immersed         on the table.
                                          in discussion and debate, trying to select a               People engaged in an inquiry process rigor-
                                          course of action by drawing on what they be-           ously question proposals and the assumptions
                                          lieve is the best available evidence. But de-          they rest on, so conflict may be intense—but it
                                          spite their similarities, inquiry and advocacy         is seldom personal. In fact, because disagree-
                                          produce dramatically different results.                ments revolve around ideas and interpreta-
                                             When a group takes an advocacy perspec-             tions rather than entrenched positions, con-
                                          tive, participants approach decision making as         flict is generally healthy, and team members
                                          a contest, although they don’t necessarily com-        resolve their differences by applying rules of
                                          pete openly or even consciously. Well-defined          reason. The implicit assumption is that a con-
                                          groups with special interests—dueling divi-            summate solution will emerge from a test of
                                          sions in search of budget increases, for exam-         strength among competing ideas rather than
                                          ple—advocate for particular positions. Partici-        dueling positions. Recent accounts of GE’s suc-
                                          pants are passionate about their preferred             cession process describe board members pursu-
                                          solutions and therefore stand firm in the face         ing just such an open-minded approach. All
                                          of disagreement. That level of passion makes it        members met repeatedly with the major can-
                                          nearly impossible to remain objective, limiting        didates and gathered regularly to review their
                                          people’s ability to pay attention to opposing          strengths and weaknesses—frequently without
                                          arguments. Advocates often present informa-            Jack Welch in attendance—with little or no at-
                                          tion selectively, buttressing their arguments          tempt to lobby early for a particular choice.
                                          while withholding relevant conflicting data.               A process characterized by inquiry rather
                                          Their goal, after all, is to make a compelling         than advocacy tends to produce decisions of
                                          case, not to convey an evenhanded or bal-              higher quality—decisions that not only ad-
                                          anced view. Two different plant managers               vance the company’s objectives but also are
                                          pushing their own improvement programs, for            reached in a timely manner and can be imple-
                                          example, may be wary of reporting potential            mented effectively. Therefore, we believe that
                                          weak points for fear that full disclosure will         leaders seeking to improve their organizations’
David A. Garvin is the Robert and         jeopardize their chances of winning the debate         decision-making capabilities need to begin
Jane Cizik Professor of Business Admin-   and gaining access to needed resources.                with a single goal: moving as quickly as possi-
istration at Harvard Business School in      What’s more, the disagreements that arise           ble from a process of advocacy to one of in-
Boston. His most recent HBR article is    are frequently fractious and even antagonistic.        quiry. That requires careful attention to three
“Leveraging Processes for Strategic Ad-   Personalities and egos come into play, and dif-        critical factors, the “three C’s” of effective deci-
vantage” (September–October 1995).        ferences are normally resolved through battles         sion making: conflict, consideration, and clo-
Michael A. Roberto is an assistant        of wills and behind-the-scenes maneuvering.            sure. Each entails a delicate balancing act.
professor at Harvard Business School.     The implicit assumption is that a superior solu-

harvard business review • september 2001                                                                                                     page 3
                                                                                            What You Don’t Know About Making Decisions

                                       Constructive Conflict                                  separate. People tend to take any criticism per-
                                       Critical thinking and rigorous debate invari-          sonally and react defensively. The atmosphere
                                       ably lead to conflict. The good news is that           quickly becomes charged, and even if a high-
                                       conflict brings issues into focus, allowing lead-      quality decision emerges, the emotional fall-
                                       ers to make more informed choices. The bad             out tends to linger, making it hard for team
                                       news is that the wrong kind of conflict can de-        members to work together during implemen-
                                       rail the decision-making process altogether.           tation.
                                          Indeed, conflict comes in two forms—cogni-             The challenge for leaders is to increase cog-
                                       tive and affective. Cognitive, or substantive,         nitive conflict while keeping affective conflict
                                       conflict relates to the work at hand. It involves      low—no mean feat. One technique is to estab-
                                       disagreements over ideas and assumptions and           lish norms that make vigorous debate the rule
                                       differing views on the best way to proceed.            rather than the exception. Chuck Knight, for
                                       Not only is such conflict healthy, it’s crucial to     27 years the CEO of Emerson Electric, accom-
                                       effective inquiry. When people express differ-         plished this by relentlessly grilling managers
                                       ences openly and challenge underlying as-              during planning reviews, no matter what he
                                       sumptions, they can flag real weaknesses and           actually thought of the proposal on the table,
                                       introduce new ideas. Affective, or interper-           asking tough, combative questions and expect-
                                       sonal, conflict is emotional. It involves per-         ing well-framed responses. The process—
                                       sonal friction, rivalries, and clashing personali-     which Knight called the “logic of illogic” be-
                                       ties, and it tends to diminish people’s                cause of his willingness to test even well-
                                       willingness to cooperate during implementa-            crafted arguments by raising unexpected, and
                                       tion, rendering the decision-making process            occasionally fanciful, concerns—was undoubt-
                                       less effective. Not surprisingly, it is a common       edly intimidating. But during his tenure it pro-
                                       feature of advocacy processes.                         duced a steady stream of smart investment de-
                                          On examination, the two are easy to distin-         cisions and an unbroken string of quarterly
                                       guish. When a team member recalls “tough de-           increases in net income.
                                       bates about the strategic, financial, and operat-         Bob Galvin, when he was CEO of Motorola
                                       ing merits of the three acquisition candidates,”       in the 1980s, took a slightly different approach.
                                       she is referring to cognitive conflict. When a         He habitually asked unexpected hypothetical
                                       team member comments on “heated argu-                  questions that stimulated creative thinking.
                                       ments that degenerated into personal attacks,”         Subsequently, as chairman of the board of
                                       he means affective conflict. But in practice the       overseers for the Malcolm Baldrige National
                                       two types of conflict are surprisingly hard to         Quality Program, Galvin took his colleagues by
                                                                                              surprise when, in response to pressure from
                                                                                              constituents to broaden the criteria for the
      Two Approaches to Decision Making                                                       award, he proposed narrowing them instead.
                                                                                              In the end, the board did in fact broaden the
                           Advocacy                    Inquiry                                criteria, but his seemingly out-of-the-blue sug-
                                                                                              gestion sparked a creative and highly produc-
      Concept of           a contest                   collaborative problem solving          tive debate.
      decision making                                                                            Another technique is to structure the con-
                                                                                              versation so that the process, by its very na-
      Purpose of           persuasion and lobbying     testing and evaluation                 ture, fosters debate. This can be done by divid-
                                                                                              ing people into groups with different, and
      Participants’ role   spokespeople                critical thinkers                      often competing, responsibilities. For example,
                                                                                              one group may be asked to develop a proposal
      Patterns of          strive to persuade others   present balanced arguments             while the other generates alternative recom-
      behavior                                                                                mendations. Then the groups would exchange
                           defend your position        remain open to alternatives
                                                                                              proposals and discuss the various options.
                           downplay weaknesses         accept constructive criticism          Such techniques virtually guarantee high lev-
                                                                                              els of cognitive conflict. (The exhibit “Structur-
      Minority views       discouraged or dismissed    cultivated and valued                  ing the Debate” outlines two approaches for
                                                                                              using different groups to stimulate creative
      Outcome              winners and losers          collective ownership

harvard business review • september 2001                                                                                                 page 4
                                                                                              What You Don’t Know About Making Decisions

                                             But even if you’ve structured the process          the ‘b’ word, and we don’t use the ‘f’ word. We
                                         with an eye toward encouraging cognitive con-          don’t place blame, and we don’t find fault.”
                                         flict, there’s always a risk that it will become          Second, leaders can help people step back
                                         personal. Beyond cooling the debate with               from their preestablished positions by break-
                                         “time-outs,” skilled leaders use a number of           ing up natural coalitions and assigning people
                                         creative techniques to elevate cognitive debate        to tasks on some basis other than traditional
                                         while minimizing affective conflict.                   loyalties. At a leading aerospace company, one
                                             First, adroit leaders pay careful attention to     business unit president had to deal with two
                                         the way issues are framed, as well as to the lan-      powerful coalitions within his organization
                                         guage used during discussions. They preface            during a critical decision about entering into a
                                         contradictory remarks or questions with                strategic alliance. When he set up two groups
                                         phrases that remove some of the personal               to consider alternative alliance partners, he in-
                                         sting (“Your arguments make good sense, but            terspersed the groups with members of each
                                         let me play devil’s advocate for a moment”).           coalition, forcing people with different inter-
                                         They also set ground rules about language, in-         ests to work with one another. He then asked
                                         sisting that team members avoid words and be-          both groups to evaluate the same wide range
                                         havior that trigger defensiveness. For instance,       of options using different criteria (such as tech-
                                         in the U.S. Army’s after-action reviews, con-          nological capability, manufacturing prowess,
                                         ducted immediately after missions to identify          or project management skills). The two groups
                                         mistakes so they can be avoided next time, fa-         then shared their evaluations and worked to-
                                         cilitators make a point of saying, “We don’t use       gether to select the best partner. Because no-
                                                                                                body had complete information, they were
Structuring the Debate                                                                          forced to listen closely to one another.
                                                                                                   Third, leaders can shift individuals out of
By breaking a decision-making body into two subgroups, leaders can often create                 well-grooved patterns, where vested interests
an environment in which people feel more comfortable engaging in debate. Scholars               are highest. They can, for example, ask team
recommend two techniques in particular, which we call the “point-counterpoint” and              members to research and argue for a position
“intellectual watchdog” approaches. The first three steps are the same for both techniques:      they did not endorse during initial discussions.
                                                                                                Similarly, they can assign team members to
Point-Counterpoint                           Intellectual Watchdog                              play functional or managerial roles different
                                                                                                from their own, such as asking an operations
The team divides into two subgroups.         The team divides into two subgroups.               executive to take the marketing view or asking
                                                                                                a lower-level employee to assume the CEO’s
Subgroup A develops a proposal,              Subgroup A develops a proposal,                    strategic perspective.
fleshing out the recommendation,              fleshing out the recommendation,                       Finally, leaders can ask participants locked
the key assumptions, and the                 the key assumptions, and the critical              in debate to revisit key facts and assumptions
critical supporting data.                    supporting data.                                   and gather more information. Often, people
                                                                                                become so focused on the differences between
Subgroup A presents the proposal             Subgroup A presents the proposal                   opposing positions that they reach a stalemate.
to Subgroup B in written and oral forms.     to Subgroup B in written and oral forms.           Emotional conflict soon follows. Asking people
                                                                                                to examine underlying presumptions can de-
Subgroup B generates one or more             Subgroup B develops a detailed critique            fuse the tension and set the team back on
alternative plans of action.                 of these assumptions and recommenda-               track. For instance, at Enron, when people dis-
                                             tions. It presents this critique in written        agree strongly about whether or not to apply
                                             and oral forms. Subgroup A revises its             their trading skills to a new commodity or mar-
                                             proposal based on this feedback.
                                                                                                ket, senior executives quickly refocus the dis-
                                                                                                cussion on characteristics of industry structure
The subgroups come together to debate        The subgroups continue in this revision-           and assumptions about market size and cus-
the proposals and seek agreement             critique-revision cycle until they converge
on a common set of assumptions.              on a common set of assumptions.                    tomer preferences. People quickly recognize
                                                                                                areas of agreement, discover precisely how
                                                                                                and why they disagree, and then focus their
Based on those assumptions, the              Then, the subgroups work together to
subgroups continue to debate various         develop a common set of recommendations.           debate on specific issues.
options and strive to agree on a
common set of recommendations.

harvard business review • september 2001                                                                                                   page 5
                                                                                         What You Don’t Know About Making Decisions

                                     Consideration                                         missed the data and went ahead with his
                                     Once a decision’s been made and the alterna-          plans. Schrempp may have solicited views
                                     tives dismissed, some people will have to sur-        from many parties, but he clearly failed to
                                     render the solution they preferred. At times,         give them much weight.
                                     those who are overruled resist the outcome; at           Leaders can demonstrate consideration
                                     other times, they display grudging accep-             throughout the decision-making process. At
                                     tance. What accounts for the difference? The          the outset, they need to convey openness to
                                     critical factor appears to be the perception of       new ideas and a willingness to accept views
                                     fairness—what scholars call “procedural jus-          that differ from their own. In particular, they
                                     tice.” The reality is that the leader will make       must avoid suggesting that their minds are al-
                                     the ultimate decision, but the people partici-        ready made up. They should avoid disclosing
                                     pating in the process must believe that their         their personal preferences early in the process,
                                     views were considered and that they had a gen-        or they should clearly state that any initial
                                     uine opportunity to influence the final deci-         opinions are provisional and subject to change.
                                     sion. Researchers have found that if partici-         Or they can absent themselves from early de-
                                     pants believe the process was fair, they are far      liberations.
                                     more willing to commit themselves to the re-             During the discussions, leaders must take
                                     sulting decision even if their views did not pre-     care to show that they are listening actively
                                     vail. (For a detailed discussion of this phenome-     and attentively. How? By asking questions,
                                     non, see W. Chan Kim and Renée Mauborgne,             probing for deeper explanations, echoing com-
                                     “Fair Process: Managing in the Knowledge              ments, making eye contact, and showing pa-
                                     Economy,” HBR July–August 1997).                      tience when participants explain their posi-
                                        Many managers equate fairness with                 tions. Taking notes is an especially powerful
Researchers have found               voice—with giving everyone a chance to ex-            signal, since it suggests that the leader is mak-
                                     press his or her own views. They doggedly             ing a real effort to capture, understand, and
that if participants                 work their way around the table, getting every-       evaluate people’s thoughts.
                                     one’s input. However, voice is not nearly as im-         And after they make the final choice, lead-
believe the process was              portant as consideration—people’s belief that         ers should explain their logic. They must de-
fair, they are far more              the leader actively listened to them during the       scribe the rationale for their decision, detailing
                                     discussions and weighed their views carefully         the criteria they used to select a course of ac-
willing to commit to the             before reaching a decision. In his 1999 book,         tion. Perhaps more important, they need to
resulting decision, even if          Only the Paranoid Survive, Intel’s chairman           convey how each participant’s arguments af-
                                     Andy Grove describes how he explains the dis-         fected the final decision or explain clearly why
their views did not                  tinction to his middle managers: “Your crite-         they chose to differ with those views.
                                     rion for involvement should be that you’re
                                     heard and understood…. All sides cannot pre-          Closure
                                     vail in the debate, but all opinions have value       Knowing when to end deliberations is tricky;
                                     in shaping the right answer.”                         all too often decision-making bodies rush to a
                                        In fact, voice without consideration is            conclusion or else dither endlessly and decide
                                     often damaging; it leads to resentment and            too late. Deciding too early is as damaging as
                                     frustration rather than to acceptance. When           deciding too late, and both problems can usu-
                                     the time comes to implement the decision,             ally be traced to unchecked advocacy.
                                     people are likely to drag their feet if they             Deciding Too Early. Sometimes people’s
                                     sense that the decision-making process had            desire to be considered team players overrides
                                     been a sham—an exercise in going through              their willingness to engage in critical thinking
                                     the motions designed to validate the leader’s         and thoughtful analysis, so the group readily
                                     preferred solution. This appears to have              accepts the first remotely plausible option.
                                     been true of the Daimler-Chrysler merger.             Popularly known as “groupthink,” this mind-
                                     Daimler CEO Jurgen Schrempp asked for ex-             set is prevalent in the presence of strong advo-
                                     tensive analysis and assessment of potential          cates, especially in new teams, whose mem-
                                     merger candidates but had long before set-            bers are still learning the rules and may be less
                                     tled on Chrysler as his choice. In fact, when         willing to stand out as dissenters.
                                     consultants told him that his strategy was               The danger of groupthink is not only that it
                                     unlikely to create shareholder value, he dis-         suppresses the full range of options but also

harvard business review • september 2001                                                                                              page 6
                                                                                              What You Don’t Know About Making Decisions

                               Advocacy Versus Inquiry in Action
                                     The Bay of Pigs and the Cuban Missile Crisis
       Perhaps the best demonstration of ad-       ing would in fact lead to a rapid do-         Kennedy learned that the Soviet Union
       vocacy versus inquiry comes from the        mestic uprising against Castro, and           had placed nuclear missiles on Cuban
       administration of President John F.         they failed to find out whether the ex-       soil, despite repeated assurances from
       Kennedy. During his first two years in      iles could fade into the mountains            the Soviet ambassador that this would
       office, Kennedy wrestled with two criti-    (which were 80 miles from the landing         not occur. Kennedy immediately con-
       cal foreign policy decisions: the Bay of    site) should they meet with strong re-        vened a high-level task force, which
       Pigs invasion and the Cuban Missile         sistance. The resulting invasion is gen-      contained many of the same men re-
       Crisis. Both were assigned to cabinet-      erally considered to be one of the low        sponsible for the Bay of Pigs invasion,
       level task forces, involving many of the    points of the Cold War. About 100 lives       and asked them to frame a response.
       same players, the same political inter-     were lost, and the rest of the exiles         The group met night and day for two
       ests, and extremely high stakes. But        were taken hostage. The incident was          weeks, often inviting additional mem-
       the results were extraordinarily differ-    a major embarrassment to the                  bers to join in their deliberations to
       ent, largely because the two groups op-     Kennedy administration and dealt a            broaden their perspective. Occasion-
       erated in different modes.                  blow to America’s global standing.            ally, to encourage the free flow of
          The first group, charged with decid-        After the botched invasion, Kennedy        ideas, they met without the president.
       ing whether to support an invasion of       conducted a review of the foreign pol-        Robert Kennedy played his new role
       Cuba by a small army of U.S.-trained        icy decision-making process and intro-        thoughtfully, critiquing options fre-
       Cuban exiles, worked in advocacy            duced five major changes, essentially         quently and encouraging the group to
       mode, and the outcome is widely re-         transforming the process into one of          develop additional alternatives. In par-
       garded as an example of flawed deci-        inquiry. First, people were urged to          ticular, he urged the group to move be-
       sion making. Shortly after taking of-       participate in discussions as “skeptical      yond a simple go-no-go decision on a
       fice, President Kennedy learned of the      generalists”—that is, as disinterested        military air strike.
       planned attack on Cuba developed by         critical thinkers rather than as repre-          Ultimately, subgroups developed
       the CIA during the Eisenhower admin-        sentatives of particular departments.         two positions, one favoring a blockade
       istration. Backed by the Joint Chiefs of    Second, Robert Kennedy and The-               and the other an air strike. These
       Staff, the CIA argued forcefully for the    odore Sorensen were assigned the role         groups gathered information from a
       invasion and minimized the risks, fil-      of intellectual watchdog, expected to         broad range of sources, viewed and in-
       tering the information presented to         pursue every possible point of conten-        terpreted the same intelligence pho-
       the president to reinforce the agency’s     tion, uncovering weaknesses and un-           tos, and took great care to identify and
       position. Knowledgeable individuals         tested assumptions. Third, task forces        test underlying assumptions, such as
       on the State Department’s Latin Amer-       were urged to abandon the rules of            whether the Tactical Air Command
       ica desk were excluded from delibera-       protocol, eliminating formal agendas          was indeed capable of eliminating all
       tions because of their likely opposi-       and deference to rank. Fourth, partici-       Soviet missiles in a surgical air strike.
       tion.                                       pants were expected to split occasion-        The subgroups exchanged position pa-
          Some members of Kennedy’s staff          ally into subgroups to develop a broad        pers, critiqued each other’s proposals,
       opposed the plan but held their             range of options. And finally, Presi-         and came together to debate the alter-
       tongues for fear of appearing weak in       dent Kennedy decided to absent him-           natives. They presented Kennedy with
       the face of strong advocacy by the CIA.     self from some of the early task force        both options, leaving him to make the
       As a result, there was little debate, and   meetings to avoid influencing other           final choice. The result was a carefully
       the group failed to test some critical      participants and slanting the debate.         framed response, leading to a success-
       underlying assumptions. For example,           The inquiry mode was used to great         ful blockade and a peaceful end to the
       they didn’t question whether the land-      effect when in October 1962 President         crisis.

harvard business review • september 2001                                                                                                     page 7
                                                                                           What You Don’t Know About Making Decisions

                                     that unstated objections will come to the sur-          Striving for fairness, team members insist on
                                     face at some critical moment—usually at a               hearing every view and resolving every question
                                     time when aligned, cooperative action is essen-         before reaching a conclusion. This demand for
                                     tial to implementation. The leader of a large di-       certainty—for complete arguments backed by
                                     vision of a fast-growing retailer learned this the      unassailable data—is its own peculiar form of
                                     hard way. He liked to work with a small subset          advocacy. Once again, the result is usually an
                                     of his senior team to generate options, evaluate        endless loop, replaying the same alternatives,
                                     the alternatives, and develop a plan of action,         objections, and requests for further informa-
                                     and then bring the proposal back to the full            tion. Any member of the group can unilaterally
                                     team for validation. At that point, his managers        derail the discussion by voicing doubts. Mean-
                                     would feel they had been presented with a fait          while, competitive pressures may be demanding
                                     accompli and so would be reluctant to raise             an immediate response, or participants may
                                     their concerns. As one of them put it: “Because         have tuned out long ago, as the same arguments
                                     the meeting is the wrong place to object, we            are repeated ad nauseam.
                                     don’t walk out of the room as a unified group.”            At this point, it’s the leader’s job to “call the
                                     Instead, they would reopen the debate during            question.” Jamie Houghton, the longtime CEO
                                     implementation, delaying important initiatives          of Corning, invented a vivid metaphor to de-
                                     by many months.                                         scribe this role. He spoke of wearing two hats
                                        As their first line of defense against               when working with his senior team: He figura-
                                     groupthink, leaders need to learn to recognize          tively put on his cowboy hat when he wanted
                                     latent discontent, paying special attention to          to debate with members as an equal, and he
                                     body language: furrowed brows, crossed arms,            donned a bowler when, as CEO, he called the
                                     or curled-up defiance. To bring disaffected             question and announced a decision. The former
                                     people back into the discussion, it may be best         role allowed for challenges and continued dis-
                                     to call for a break, approach dissenters one by         cussion; the latter signaled an end to the de-
                                     one, encourage them to speak up, and then re-           bate.
                                     convene. GM’s Alfred Sloan was famous for                  The message here is that leaders—and their
                                     this approach, which he would introduce with            teams—need to become more comfortable
                                     the following speech: “I take it we are all in          with ambiguity and be willing to make speedy
                                     complete agreement on the decision here.                decisions in the absence of complete, unequivo-
                                     Then I propose we postpone further discussion           cal data or support. As Dean Stanley Teele of
                                     of the matter until our next meeting to give            Harvard Business School was fond of telling
                                     ourselves time to develop disagreement and              students: “The art of management is the art of
                                     perhaps gain some understanding of what the             making meaningful generalizations out of inad-
                                     decision is all about.”                                 equate facts.”
                                        Another way to avoid early closure is to cul-
                                     tivate minority views either through norms or           A Litmus Test
                                     through explicit rules. Minority views broaden          Unfortunately, superior decision making is dis-
                                     and deepen debate; they stretch a group’s               tressingly difficult to assess in real time. Success-
                                     thinking, even though they are seldom                   ful outcomes—decisions of high quality, made
                                     adopted intact. It is for this reason that Andy         in a timely manner and implemented effec-
                                     Grove routinely seeks input from “helpful Cas-          tively—can be evaluated only after the fact. But
                                     sandras,” people who are known for raising              by the time the results are in, it’s normally too
                                     hard questions and offering fresh perspectives          late to take corrective action. Is there any way to
                                     about the dangers of proposed policies.                 find out earlier whether you’re on the right
                                        Deciding Too Late. Here, too, unchecked              track?
                                     advocacy is frequently the source of the prob-             There is indeed. The trick, we believe, is to
                                     lem, and in these instances it takes two main           periodically assess the decision-making process,
                                     forms. At times, a team hits gridlock: Warring          even as it is under way. Scholars now have con-
                                     factions refuse to yield, restating their positions     siderable evidence showing that a small set of
                                     over and over again. Without a mechanism for            process traits is closely linked with superior out-
                                     breaking the deadlock, discussions become an            comes. While they are no guarantee of success,
                                     endless loop. At other times, people bend over          their combined presence sharply improves the
                                     backward to ensure evenhanded participation.            odds that you’ll make a good decision.

harvard business review • september 2001                                                                                                  page 8
                                                                                           What You Don’t Know About Making Decisions

                                         Multiple Alternatives. When groups con-                 Some questions open up discussion; others
                                     sider many alternatives, they engage in more            narrow it and end deliberations. Contrarian hy-
                                     thoughtful analysis and usually avoid settling          pothetical questions usually trigger healthy de-
                                     too quickly on the easy, obvious answer. This is        bate. A manager who worked for former Amer-
                                     one reason techniques like point-counterpoint,          ican Express CEO Harvey Golub points to a
                                     which requires groups to generate at least two          time when the company was committed to low-
                                     alternatives, are so often associated with supe-        ering credit card fees, and Golub unexpectedly
                                     rior decision making. Usually, keeping track of         proposed raising fees instead. “I don’t think he
                                     the number of options being considered will tell        meant it seriously,” says the manager. “But he
                                     if this test has been met. But take care not to         certainly taught us how to think about fees.”
                                     double count. Go-no-go choices involve only                 The level of listening is an equally important
                                     one option and don’t qualify as two alternatives.       indicator of a healthy decision-making process.
                                         Assumption Testing. “Facts” come in two             Poor listening produces flawed analysis as well
                                     varieties: those that have been carefully tested        as personal friction. If participants routinely in-
                                     and those that have been merely asserted or as-         terrupt one another or pile on rebuttals before
                                     sumed. Effective decision-making groups do not          digesting the preceding comment, affective
                                     confuse the two. They periodically step back            conflict is likely to materialize. Civilized discus-
                                     from their arguments and try to confirm their           sions quickly become impossible, for collegial-
                                     assumptions by examining them critically. If            ity and group harmony usually disappear in the
                                     they find that some still lack hard evidence, they      absence of active listening.
                                     may elect to proceed, but they will at least know           Perceived Fairness. A real-time measure of
                                     they’re venturing into uncertain territory. Alter-      perceived fairness is the level of participation
                                     natively, the group may designate “intellectual         that’s maintained after a key midpoint or mile-
                                     watchdogs” who are assigned the task of scruti-         stone has been reached. Often, a drop in partici-
                                     nizing the process for unchecked assumptions            pation is an early warning of problems with im-
                                     and challenging them on the spot.                       plementation since some members of the group
                                         Well-Defined Criteria. Without crisp, clear         are already showing their displeasure by voting
                                     goals, it’s easy to fall into the trap of comparing     with their feet.
                                     apples with oranges. Competing arguments be-                In fact, keeping people involved in the pro-
                                     come difficult to judge, since advocates will sug-      cess is, in the end, perhaps the most crucial fac-
                                     gest using those measures (net income, return           tor in making a decision—and making it stick.
                                     on capital, market presence, share of mind, and         It’s a job that lies at the heart of leadership and
                                     so on) that favor their preferred alternative.          one that uniquely combines the leader’s nu-
                                     Fuzzy thinking and long delays are the likely re-       merous talents. It requires the fortitude to pro-
                                     sult.                                                   mote conflict while accepting ambiguity, the
                                         To avoid the problem, the team should spec-         wisdom to know when to bring conversations
                                     ify goals up front and revisit them repeatedly          to a close, the patience to help others under-
                                     during the decision-making process. These               stand the reasoning behind your choice, and,
                                     goals can be complex and multifaceted, quanti-          not least, a genius for balance—the ability to
                                     tative and qualitative, but whatever form they          embrace both the divergence that may charac-
                                     take, they must remain at the fore. Studies of          terize early discussions and the unity needed
                                     merger decisions have found that as the process         for effective implementation. Cyrus the Great,
                                     reaches its final stages and managers feel the          the founder of the Persian Empire and a re-
                                     pressure of deadlines and the rush to close, they       nowned military leader, understood the true
                                     often compromise or adjust the criteria they            hallmark of leadership in the sixth century BC,
                                     originally created for judging the appropriate-         when he attributed his success to “diversity in
                                     ness of the deal.                                       counsel, unity in command.”
                                         Dissent and Debate. David Hume, the great
                                     Scottish philosopher, argued persuasively for           Reprint R0108G; Harvard Business Review
                                     the merits of debate when he observed that the          OnPoint 5399
                                                                                             To order, see the next page
                                     “truth springs from arguments amongst
                                                                                             or call 800-988-0886 or 617-783-7500
                                     friends.” There are two ways to measure the
                                                                                             or go to www.hbr.org
                                     health of a debate: the kinds of questions being
                                     asked and the level of listening.

harvard business review • september 2001                                                                                                 page 9
                                        What You Don’t Know About Making

                                        Further Reading
                                        How Management Teams Can Have a                     imize political jockeying and favoritism, en-
                                        Good Fight                                          abling people to focus on the decision at
                                        by Kathleen M. Eisenhardt, Jean L.                  hand.
                                        Kahwajy, and L.J. Bourgeois III
                                                                                            Conquering a Culture of Indecision
                                        Harvard Business Review
                                                                                            by Ram Charan
                                        July–August 1997
                                                                                            Harvard Business Review
                                        Product no. 536X
                                                                                            April 2001
                                        This article provides additional suggestions for    Product no. 9373
                                        cultivating cognitive conflict while mitigating
                                                                                            Charan focuses on closure, describing tech-
                                        affective conflict. For example: 1) Arm yourself
                                                                                            niques for mitigating the collective indecision
                                        with a wealth of data about the decision at
                                                                                            that plagues many teams and companies. A
                                        hand. This encourages decision makers to de-
                                                                                            key to decisive decision making? Enabling
                                        bate issues, not argue out of ignorance. 2)
                                                                                            people to express their real views during
                                        Consider multiple alternatives while weighing
                                                                                            meetings. How? 1) Ask questions that prompt
                                        decisions—even some you don’t support. This
                                                                                            team members to test their assumptions and
                                        prevents teams from polarizing around just
                                                                                            generate new insights—“How will you make
                                        two possibilities. 3) Define common goals to
                                                                                            those gains?” and “Where are the gaps in your
                                        unite your team. You’ll make it in everyone’s in-
                                                                                            competitor’s product line?” 2) Encourage peo-
                                        terest to achieve the best solution. 4) Within
                                                                                            ple to air conflicts underlying apparent con-
                                        the team, use humor to relieve tension and
                                                                                            sensus. 3) Ensure that people leave meetings
                                        promote collaborative esprit, and increase
                                                                                            knowing what they’re expected to do. 4) Use
                                        tactfulness, effective listening, and creativity.
                                                                                            follow-through and feedback to sustain im-
                                        Fair Process: Managing in the Knowledge             plementation of the decision.
                                        by W. Cham Kim and Renée Mauborgne
                                        Harvard Business Review
                                        July–August 1997
                                        Product no. 407X

                                        Kim and Mauborgne shine the spotlight on
To Order                                consideration, offering additional advice for
                                        ensuring that your team members perceive
For reprints, Harvard Business Review   the decision-making process as fair. The au-
OnPoint orders, and subscriptions       thors recommend applying three principles:
to Harvard Business Review:             1) Engagement: Invite team members’ input
Call 800-988-0886 or 617-783-7500.      and encourage them to challenge one an-
Go to www.hbr.org                       other’s ideas. You’ll communicate your re-
                                        spect and build collective wisdom. 2) Explana-
For customized and quantity orders      tion: Clarify the thinking behind a final
of reprints and Harvard Business        decision to reassure team members that
Review OnPoint products:                you’ve considered their opinions and made
Call Frank Tamoshunas at                the decision with the company’s overall inter-
617-783-7626,                           ests at heart. 3) Expectation clarity: State the
or e-mail him at                        rules of the game, including performance
ftamoshunas@hbsp.harvard.edu            standards and new responsibilities. You’ll min-

                                                                                                                                   page 10

To top