Nutshell: Using our judgement to make a (conscious) decision

Complex decision-making requires us to move from divergent to convergent thinking to reach a decision – while avoiding groupthink, bias and information overload.

Let’s imagine that we have a complex decision to make in the workplace. Perhaps, for example, we are deciding to what extent we can support hybrid working in the post-COVID-19 world. Or we have to do some hard thinking about where to focus our marketing spend in the next financial year.

We have already defined our problem clearly with the help of root cause analysis and have used divergent thinking to come up with a plethora of innovative solutions.

So far, so good.

We are now at the stage of convergent thinking, evaluating solutions in an objective and systematic manner. This step is all about prioritisation and choosing – using our judgement to make a decision in a way that is both informed and timely. So, how can we move effectively from convergence to decision, dodging the pitfalls that could throw us off course so that we can make the best choice possible?

We can begin by taking a cautionary lesson from JF Kennedy.

In 1961, shortly after his inauguration as US president, Kennedy approved a plan set in motion by his predecessor, Dwight Eisenhower, for a covert military operation in Cuba. Aiming to remove Fidel Castro, the US government directed and financed an invasion of the Bay of Pigs by some 1,400 Cuban exiles.

Kennedy spent days discussing whether to go ahead with the operation, involving a large team of brilliant advisors, including the then US secretary of defense Robert McNamara, renowned historian and intellectual Arthur Schlesinger Jr, and Allen Dulles, head of the CIA.

Despite their combined expertise, the decision turned out to be flawed. The invasion force was unequal to the strength of Castro’s troops, the environment was hostile, and the operation was over within a matter of days, leading to the capture of more than 1,100 men. The move only served to strengthen the position of Castro’s administration and its links with the Soviet Union.

“There were 50 or so of us, presumably the most experienced and smartest people we could get,” Kennedy later recalled. “But five minutes after it began to fall in, we all looked at each other and asked, ‘how could we have been so stupid?’.”

The answer, according to social psychologist Irving Janis, who researched the chain of events, is that Kennedy and his colleagues fell victim to ‘groupthink’, a term Janis coined in his 1972 paper Victims of Groupthink: A Psychological Study of Foreign-Policy Decisions and Fiascos.

For him, “groupthink” refers to a deterioration of mental efficiency, reality testing and moral judgement that results from in-group pressures. It happens when a team feels the need to conform, looks at a situation through a biased, narrow lens and reaches premature conclusions. Individuals have a tendency to push aside their personal opinions or beliefs in order to reach a shared consensus with the group – often to the detriment of the group’s goals.

With Kennedy’s team, Janis describes the dynamic as a “runaway norm” – with the norm being to exceed other members of the group in taking more extreme and unrestrained actions against an ‘enemy’. Not an environment conducive to good decision-making.

Group dynamics

The lesson in this is that people don’t always make rational choices – even with access to all the right information – and that groupthink can compound individual biases.

It underlines how carefully we need to think about who to involve in decision-making – and how important it is to consider the interpersonal dynamics of the group. It may be harder to reach a consensus with a (cognitively and demographically) diverse mix, but the more relevant perspectives we gain, the greater the collective wisdom and expertise – and the less likely we are to default to groupthink and other herd behaviour.

But involving too many people can also be a problem: research suggests that once we have seven people in a decision-making group, each additional member reduces decision effectiveness by 10%.

Once we have the perfect decision-making team assembled, we need to stay alert to a whole range of biases. These include ‘bandwagon effect’, a form of groupthink where we become ‘sheep-like’, doing or believing something because a large enough group of other people do or believe the same.

Confirmation bias kicks in when we listen to and remember only the information and opinions that confirm our own preconceptions. And ‘ostrich effect’ happens when we put our heads in the sand and ignore an obviously negative situation, simply ‘hoping for the best’ instead.

There is also plain, old-fashioned overconfidence – highlighted by psychologist Daniel Kahneman as the bias he’d eliminate first if he had a magic wand.

As a leader, we should initially keep our ideas and opinions to ourselves to avoid steering opinion in a particular direction, asking other senior participants to do the same. People may not feel comfortable proposing alternative solutions or disagreeing with someone more senior than themselves. They may even compete to agree with and flatter us in order to curry favour.

Establishing an atmosphere of psychological safety, so that everyone feels able to contribute to the discussion openly and honestly, must therefore be a priority. A measure of healthy conflict should be encouraged in a discussion of this kind to avoid stifling genuine insights.

A survey by global management consultants McKinsey and Company found that “the presence of high-quality interactions and debate was the factor most predictive of whether a respondent said their company made good, fast big-bet decisions”.

In their report, the authors suggest simple behaviour changes to overcome “the conspiracy of approval” approach to group discussion. For example:

• reminding participants at the outset of overall organisational goals to reframe subsequent discussions.

• assigning someone to argue the case for and against a potential decision or the various options under consideration.

• asking the leaders of business units, regions or functions to examine the decision from outside their own point of view.

• appointing a rotating devil’s advocate role to bolster critical thinking.

• using pre-mortem exercises (where we start by assuming the initiative in question turned out to be a failure and work back for likely explanations) to pressure test for weak spots in an argument or plan.

All these can support us to explore assumptions and alternatives beyond what is being presented and actively seek information that might disconfirm the group’s initial hypotheses. 

Deciding how to decide

The precise format of our decision-making – and leadership style within this – will, of course, vary according to the decision to be made. There is no one-size-fits-all method that suits every scenario: not all decisions are created equal, and not everyone needs to be involved in every choice that is taken.

With strategic decisions impacting a broad set of stakeholders, participants’ roles and level of decision accountability can be clarified with the help of Bain & Company’s RAPID decision-making model. This spells out five roles or actions of which ‘Decide’ is the key role, with the others supporting:

• Recommend. These participants create the initial proposals and recommendations. They should back up their recommendations with facts, figures and research.

• Agree. These people must agree the proposals from the Recommend group before it can move forward – and have the power to veto the decision.

• Perform. This group will implement the decision once it is made.

• Input. These people provide information and facts to the Recommend group (although this group doesn’t have to consider these).

• Decide. This person has the authority to make the decision once all the options have been laid out. They will then delegate implementation to the Perform role, though they retain responsibility for its execution.

The value of the model lies in giving all the key people a chance to be involved in decision-making, which enhances stakeholder buy-in. The downside is that it can slow things down (particularly where there are several people in the ‘agree’ group, with the power of veto).

Where speed and decisiveness are of the essence, we may therefore need a more autocratic style, with decisions taken independently by the group leader.

The Vroom-Yetton Decision Model provides a framework to help us choose the most appropriate decision-making process for our particular situation. It poses seven ‘yes/no’ questions which relate to:

• The quality of the decision. How much impact will the decision have and how important is it to find the right solution? The higher the decision’s quality, the more people must be involved in the process.
  
  
• Involvement and collaboration. How important is it that everyone in the group agrees to the decision? The degree of participation must be raised or lowered accordingly.
  
  
• Time constraints. How much time is there to take a decision? If there’s no time to lose, a decisive, autocratic approach will be quickest; if the timescale is less tight, we will have scope to involve more people in the process.

In line with these themes, the model also sets out five leadership styles/decision-making processes:

Autocratic I (A1): We use available information to make a decision independently, without gaining the input or opinions of team members or external parties.

Autocratic II (A2): We gather specific information from group members or other stakeholders – but make our decision independently.

Consultative I (C1): We consult with group members individually and collect information, but make the final decision without group discussion.

Consultative II (C2): We bring the group together to discuss the situation, sharing the problem and seeking suggestions. However, we still make the ultimate decision ourselves and are under no obligation to incorporate collective input into it.

Group II (G2): Acting purely as a facilitator, we present the situation and problem to the group, helping to identify alternatives and make a consensus decision with members.

An autocratic style is most appropriate when:

• as leader, we have more expertise on the subject than others.

• we are confident about acting alone.

• the team will accept our decision.

• we have a tight timescale.

A consultative or collaborative style is most appropriate when:

• we need information from others to solve a problem.

• the problem cannot be defined easily.

• team members’ buy-in is important.

• we have sufficient time to co-ordinate a group decision.

Collaborative decision-making is clearly the most democratic approach, but it is also time-intensive, so is appropriate for important decisions that do not need to be made immediately.

However, even when the process is collaborative, we should generally aim to gain commitment rather than unanimous agreement. In his 2017 letter to Amazon shareholders, its founder, Jeff Bezos, introduced the concept of “disagree and commit” with respect to decision-making. 

“Any agreement voiced in the absence of a strong sense of collective responsibility can prove ephemeral,” warns McKinsey.

Creating our decision tree

To determine the right leadership style and method for our situation, we answer seven questions (in the given order) with a yes/no response, and plot the accompanying abbreviations on a matrix (or ‘decision tree’), as shown below. Not all the questions will be relevant to every scenario, but they help us to make a judgement call about the style most likely to lead to the best possible decision depending on the circumstances we find ourselves in.

We should also bear in mind that this model fails to take account of subtleties such as team dynamics and the task’s complexity, so we will need to factor these elements in ourselves.

Organising information

With complex problems, we will often be factoring in lots of loose ideas and unstructured information, particularly post-brainstorming. Here, argument mapping can help us to translate this into a format that’s easier to manage and understand – eliminating unnecessary discussion and distractions from the goal. As a rule, we can skip this where fewer than 15 items of information have been identified.

For example, the six-step KJ-Method enables us to unify large quantities of data by finding relationships between concepts, producing affinity diagrams that organise facts, opinions and issues into taxonomies so that we can diagnose complex problems and identify common issues.

Since we have already moved past the divergent phase of the problem-solving process – which generates much of this data – we can start at step three.

1. Identify the problem, writing out the problem clearly so that everyone can see and understand it.

2. Generate ideas, using brainstorming or a similar idea-generation technique to determine all aspects of the problem. We record the ideas using index cards or sticky notes.

3. Cluster all ideas that are similar to each other into groups; some might be clearly related, others connected more loosely.

4. Create an affinity header card for each group with a short statement that describes the entire group of ideas. This is an idea that captures the essential link between the ideas contained in a group of cards.

5. Group the affinity cards into ever-broader clusters until the definition of an affinity cluster becomes too broad to have any meaning.

6. Create an affinity diagram, placing the affinity cards on a single large sheet of paper or whiteboard. Draw borders around each affinity cluster, producing a diagram that provides insights into the problem.

Learning from past experience

Affinity diagrams enable us to sift through qualitative data, synthesise ideas and findings, establish new patterns of thinking and group data in an organised manner, providing us with a highly visual tool.

Meanwhile, checklists also help us to manage information, guiding us to take the right decisions when under pressure and to avoid repeating avoidable mistakes.

In The Checklist Manifesto: How to Get Things Right, US surgeon Atul Gawande explains that while we sometimes make mistakes because we don’t know enough about something, we often do so because we fail to make proper use of what we do know. Experts, he argues, (particularly in fields such as medicine and aviation), need written guides that walk them through the key steps in any complex procedure.

In decision-making, checklists provide us with guidelines drawn from relevant previous decisions made by ourselves or others. They can be developed, for example, from the decision journals we keep in order to evaluate and learn from our own past decisions, as long as we factor ‘hindsight bias’ into our learnings: even when an outcome is poor, it doesn’t automatically mean that the decision-making was flawed – and vice versa.

These insights can help us to avoid common pitfalls, negotiate biases and overcome blind spots, enabling us to apply certain criteria consistently across decisions, maintain alignment with our organisational goals and priorities, and eliminate options faster.

Of course, checklists shouldn’t be treated as a crutch – they need to be used mindfully – but they can help to prevent us making similar mistakes in our decision-making over and again.

Using our judgement

Tools such as these can support us in making conscious decisions – with our eyes open – helping us to make the best use of data and analytics, involve the right people in the right ways, and to mitigate against cognitive biases.

With effective processes, we can avoid many of the pitfalls that often undermine corporate decision-making – and where our choices prove wrong, we should draw on our learnings, just as JFK learned profoundly from the Bay of Pigs.

In his book Collaboration, Morten T Hansen explains that Kennedy subsequently instituted four key changes to how his top team would make critical decisions:

• Each participant would function as a “skeptical generalist”, focusing on the problem as a whole, rather than approaching it from their department’s standpoint.

• To avoid the status-laden meetings at the White House, and encourage freewheeling discussions, no formal agenda or protocol would be used.

• The team would be broken into sub-groups to work on alternative solutions and then reconvene.

• The team should sometimes meet without Kennedy present, to avoid people simply following his views.

The aim was to “solicit diverse viewpoints, stimulate debate, explore options, probe assumptions, and let the best plan win on its merits”.

Though hard-won, Kennedy’s insights changed the way he approached decision-making and had a huge and positive influence on today's management thinking Similarly, the lessons we learn from our own decision-making can inform and improve the processes within our own teams and organisations, helping us to hone our judgement and make better – and more conscious – decisions.

Test your understanding

• Describe two practical ways in which we can address a ‘conspiracy of approval’ to improve group decision-making.

• Explain when we should opt for an autocratic decision-making style and when a consultative or collaborative style is more effective.

What does it mean for you?

• Try using the Vroom-Yetton Decision Model to help you select the most appropriate decision-making process.

• Tackle a complex decision which involves lots of unstructured information by using the KJ-Method to find relationships between concepts or ideas to produce an affinity diagram.