Humans are a social species, and humans organise spontaneously into groups. Group behaviour is something we share with other non-human primate species. Certain principles are starting to emerge from studies of group structure, formation and behaviour. Group meetings of every type and stripe are a ubiquitous feature of organisational life. Peter Drucker, the renowned management theorist, famously stated ‘Meetings are a symptom of bad organization. The fewer meetings the better’. He had it right and wrong. Coming together in meetings and in other, less formal, places (such as ‘the water cooler’, coffee bar or wherever) is an essential part of maintaining the social disposition of organisations. I use the phrase ‘social disposition’ deliberately. Organisations, at their core, are social: they rely on the patterns of interactions among the individuals that work in those organisations to ensure that the organisation achieves its goal, whatever this goal may be. Humans are a profoundly social species. This is a theme reiterated time and again throughout this book. Social learning and social transmission of knowledge, on a one-to-one or small group basis, form an essential and core part of how it is that we learn during the course of both childhood and adolescence, but also how we learn in organisations. So all is well then, isn’t it? Unfortunately not. Survey after survey of employees indicates that meetings can be among the least productive and most complained-about activities that employees complain about. Complaints take myriad forms, from meetings with no apparent agenda, to meetings that have no apparent conclusion, to meetings whose minutes do not meaningfully reflect the discussion that has just taken part, to meetings where the boss spends their time bullying and harassing individual employees, in the mistaken belief that this will in some way encourage better performance in others. Steve Jobs is the archetypal example of this kind of boss, so much so that his employees referred to his ‘reality distortion field’, which surrounded many of the things that he would do and say. And of course, when we look back on his amazing career, we tend now to just focus on his successes and forget his many failures.

There are at least two ways to think of how it is that meetings fail. One is in terms of the economic cost and opportunity cost of the meeting. Let’s look at the economic cost and opportunity cost first. Think of a professional services firm such as a law firm or a management consultancy. It may have a number of so-called fee-earners (the professionals whose time can be directly charged on a project basis or an hourly basis to some piece of client-originating work). In addition, these individuals may have support staff, personal assistants, junior lawyers or the like; then, they have an attributable part of the cost of running the business as well (things like the electricity bills, the cost of their building, professional indemnity and other similar kinds of costs). Let’s say, for the sake of example, these individuals are paid on average gross salary of €200 per hour. That means a meeting of one hour’s duration with ten such individuals involved costs €2,000 of their time. Furthermore, their time might be charged out at, say, €350 per hour. Time spent in a routine office meeting can’t be charged to a client, so the gross cost is in fact €2,000 plus €3,500—in other words, €5,500. It gets worse. Taking people away from their work also breaks the pattern of their working day, so there is an opportunity cost also deriving from taking them away from the work that they are engaged in. This might be difficult to capture directly, but it could manifest itself in things such as ideas that were not had for the court case that was about to be fought, and therefore lost (a hidden though palpable opportunity cost), or, in the case of a knowledge worker such as a computer coder, the time away from what might have been a breakthrough moment that would have led to a faster, better, cheaper, more efficient method of delivering some service. These are many costs for a business to bear. My example, if anything, understates the matter. Michael Mankins and his colleagues (2014), for example, found that, in one large organisation they examined, the total cost in terms of time of the weekly executive meeting was an astonishing 300,000 hours per annum. This happens because of a ripple or cascade effect through the organisation of staff needing to service, in an upward fashion, the informational and other demands of the top-tier of management. Therefore, meetings, when they are held, should be held sparingly; they should be useful, productive and carefully structured so that the meeting achieves some meaningful goal or aim. Businesses and organisations should attempt to ensure when it decides to hold meetings that these meetings are capable of meeting these types of exacting criteria.

Meetings Going Awry

The example I examine here is a straightforward one. Consider mid-level, interregional management meetings in a large pharmaceutical company, employing several tens of thousands of workers. Managers who are requested to attend those meetings, in addition to losing time in the office, will also spend time in the air, time in the hotel and time engaged in preparation for the meeting, which may be entirely wasted or pointless. Additionally, there is transit time to and from the meeting, settling-in time in the meeting and perhaps other time costs arising from the meeting also. Flights can be delayed or cancelled; taxis may not be available; trains may breakdown, or whatever. There are also the lost hours of productivity resulting from a lack of clearly defined goals as a result of the meeting or from bullying or harassment that can occur during the meeting. Again, to emphasise the point, meetings should be held sparingly, they should be useful, and they should be productive. They should leave participants feeling that they have gained more from attending the meeting than they lost from being at the meeting, and they should leave individuals with at least some sense of a common or shared purpose in the greater organisation itself.

We have all been there: having a difficult and ill-focused discussion where it is not clear what the source of division is and from where it is difficult to achieve a shared outcome. The important lesson from this section is simple: when a meeting is drifting toward conflict, and it is obvious that there is a problem with communication, not necessarily with the topics at hand, stop the conversation or discussion. State clearly, but politely that what is being said is not necessarily what is being heard. Ask the other party to state what it is that they think you are saying. Try to achieve a shared statement that reflects the underlying content and intent of all parties present. The effect of these strategies will be in effect to re-boot the brains of all present: the language areas of the brain will be re-engaged, but with a new task.

Why do meetings fail? Meetings fail because there are ubiquitous tendencies in both our thinking and our behaviour, arising from cognitive biases on the one hand, and status biases on the other, that can serve to derail a meeting quite badly. Chairing and running a meeting is a skill, one that requires some degree of learning and the ability to take the perspective of others. If you think the meeting was a great success because you chaired it and spoke a lot, then you’re wrong. During the meeting, did you notice that after a few minutes you were just zoned out, because the people there are hearing the same old same old, and they’re contemplating their next game of Pokémon Go? Did you fail to notice that the department was just fine that time you were out for a few months because of major back surgery? All of these kinds of things suggest that a crash course in learning to take the perspective of others, and engaging in active listening during a meeting, might pay dividends. Happily, we now have a substantial body of empirical work showing how it is that meetings can become derailed and how we can run meetings so that they are much more successful in terms of their outcomes. We will focus first on some of the ubiquitous types of cognitive bias that are apparent when people deliberate and discuss things together in a group context. Then, we will focus on how status in groups affects how we behave in those groups. Along the way we will learn that the skill of ‘active listening’ and perspective taking pays great dividends for both the chairing of successful meetings and also successful leadership.

How Groups Make Poor and Disastrous Decisions

When thinking about the deliberative processes that group meetings are called for, the best practice might be to think of a group meeting, at least in its early stages, as being analogous to the resting state network, or default mode network of the brain, whose job, in part at least, is to push or surface new ideas into consciousness; the consciousness here being a collective and trans-active one, with the social functions of consciousness being key. One answer to the question of why we are conscious arises from thinking about social life and group life. Consciousness is fundamental as far as our social lives are concerned—the best parties do not happen while we are asleep! Consciousness therefore has an important social transaction function—one that extends far beyond what happens within the hidden recesses of a single individual’s brain. It is reasonable to argue that without conscious experience, and the capacity to represent and comprehend the internal mental worlds of others, that our rich social world would not exist.

Poor Decisions by Groups

History is replete with examples of poor decisions taken by groups. These can run from bad decisions over acts of war that sink whole countries to product launches that sink companies. Poor decision-making can, of course, be of a lesser level, as Adam Smith famously put it: ‘There’s a great deal of ruin in a nation’, meaning that nations can continue to function below par as a result of poor decision-making, for very considerable periods of time, while they quietly squander their accumulated capital of the ages. In Western Europe, for example, Italy provides a sad example of this: a country that, because of its idiosyncratic decision-making processes, is unable to reform itself, and at the same time has generated near-zero levels of economic growth over a 30-year period, while simultaneously being capable of producing world-class organisations that dominate the fashion, motor and food industries. That there is a lot of ruin in nations is equally true of organisations. Organisations can run sub-optimally because of their poor procedures and processes, with meetings providing a key pinch-point at which these poor procedures and processes become visible. Organisational and institutional inertia, however, may strongly militate against badly needed organisational change. Cass Sunstein and Reid Hastie, in their remarkable book ‘Wiser: getting beyond groupthink to make groups smarter’ (2015), analyse in detail the cognitive processes and behavioural problems that can cause organisations trouble with regard to the deliberative processes that the organisations use in order to develop some product or process, or indeed simply to take a decision. The key problem that bedevils deliberative processes at all levels of organisations is one of accessing, integrating and acting upon critical hidden information that may be held by members of the deliberative group, or indeed held by people who should be members of that group, but for whatever reason are not. Groups lacking cognitive diversity fail at vital problem-solving tasks and are subject to a whole variety of processes that amplify faulty decision-making. The consequences of this can be disastrous. These can be straightforward, from product launches failing terribly, which sink a company; from product rebranding, which can be undertaken for little better than vanity reasons, which in turn can cause catastrophic decline in sales. Kellogg’s, the cereal company, discovered this in 1998, when they renamed their very successful cereal product ‘Coco Pops’ as ‘Choco Krispies’ in the UK. Sales were affected, consumers complained, and Kellogg’s had to quickly rename ‘Choco Krisps’, in the guise of a consumer preference competition, as Coco Pops again in 1999.

Sunstein and Hastie (pp. 23–24) suggest groups may encounter four problems when they are engaged in deliberative processes. First, they might not correct errors that particular individuals (who may be central, cognitively or in status terms) to the group’s functioning, leading to the amplification of the error by a group. Second, there can be cascade effects, where the initial speaker or speakers create a follower effect, where those that follow afterwards take their lead and do not deviate from it. Third, groups can become substantially more polarised than even the most extreme member of their groups, leading them to take risks or to have beliefs that are substantially at variance with reality. Finally, group discussions or deliberative discussions have a very strong tendency to engage in discussing information that is already known and shared by most or all of the members of the group, information that Sunstein and Hastie characterise as knowledge that ‘everyone knows already’. The error here, of course, is to engage in a discussion that supports or reinforces a pre-existing confirmation bias, because, of course, this is more intrinsically rewarding than engaging in a probing discussion that seeks the limits of what is known or that focuses on areas of error. Additional variables may also come into play that support these four fundamental problems. People in groups may engage in acts of self-silencing, whereby they do not volunteer critical information because of problems revolving around social incentives or informational signals. Social incentives can be quickly and clearly understood: this is where, for reasons of self-preservation, self-interest or fear of the group chair, the possessor of important information judges that the cost or consequence of speaking up outweighs the benefit. One solution here, of course, is to solicit opinions in advance and circulate these opinions anonymously ahead of time, in other words, having an information-gathering phase distinct from a deliberation phase. Information signals are somewhat more complex. These revolve around the individual in a group context coming to judge that their knowledge is in some way defective, even if in fact their knowledge is correct.

Social Conformity in Groups

The classic example arises from the Solomon Asch studies of social conformity originally conducted in the 1950s. Asch was interested to understand the dynamics of what it is makes people conform to judgements in a group context, which are clearly at variance with reality. His experimental paradigm was deceptively simple and is easy to replicate. A group of individuals are brought to a conference table. One is a stooge; the other are confederates. The stooge is led to believe that this is the first time the confederates and the stooge have met each other. They are told by the experimenter that they are engaged in a simple task to do with visual perception. A series of cards are then displayed to the room. These cards have lines drawn on them, two of which are the same, the third of which may or may not be different. Each person presents their judgement in turn, with the stooge usually being second last or last. On the first couple of passes, the clearly different line (the shorter or longer line) is adjudged to be different from the two comparison lines. And then, after these few passes, the judgements start to change and the confederates give a line length judgement, clearly at variance with reality. They do this slowly and carefully and deliberately in sequence, and the stooge is left in the uncomfortable position of possibly having to state something in public that he knows to be incorrect, namely that the shorter or longer line is identical in length to the other two lines. Most people, most of the time, find this situation quite stressful, and they start to doubt their own judgement and defer to the judgement of the group.

So what happens with these individuals? Are they simply overtly denying the covert and correct judgement of their senses, or is it that they are coming to see the incorrect line as the same length as the result of the group or peer pressure that they have been exposed to? The answer, sadly, appears to be the latter. In people who are susceptible to the Asch effect, brain-imaging experiments conducted by Gregory Berns and his colleagues (2005) have found that there is a modification in the activity of the visual perceptual areas of the brain during these socially mediated perceptual judgement tasks. In other words, people come quickly to disbelieve the evidence of their own senses, and their senses in turn start to conform to what it is that the group sees, not what they individually see. This effect also turns out to be true of memory, as has been demonstrated by Ray Dolan and his colleagues. They have found that when people are misled during recall by the statements of group members who say they recall information that was not, in fact, present during the original learning experience, those who behaviourally show false recall also show modification of the brain circuits that support memory, relative to those who have not had false recall.

These experiments demonstrate that there are powerful psychological and neural mechanisms at work here. It may in fact be reasonable for an individual to doubt the evidence of his or her own eyes, or indeed the lack of evidence of his or her own eyes. Being a member of a group that might be stalked by a predator doesn’t leave you with much time to make a judgement: if someone shouts ‘There’s a tiger’ and you think it’s a bush, then you may end up as a tasty meal, whereas your group members, which may include those who decided that, actually, they don’t care whether it’s a bush or a tiger, they’re going to run because they don’t have the time to make such a judgement, will have survived. The problem, of course, is that these socially mediated transactive processes exert powerful effects in domains for which they never evolved and in which we routinely find ourselves in modern life.

Is the news all bad with regard to group deliberation or group discussions? The answer is, of course, no. Sunstein and Hastie (pp. 25–26) emphasise that groups can and do work exceptionally well together, and when they do work very well together, some remarkable effects start to appear in the group. They emphasise three in particular. The first is that a group that works well, which surfaces the relevant and necessary information and is capable of integrating it and acting upon it, work at the level of their best members, not at the level of the average of their members of or their worst members. Such groups are capable of solving difficult problems with what may be fuzzy or ill-defined solutions, but for which solutions are required quickly. A key example, of course, is the rapid response of the healthcare company Johnson & Johnson to the Tylenol controversy that engulfed it several decades ago. In this case, Johnson & Johnson quickly and effectively took ownership of the problem, removed the product that had been tampered with from the shelves, gave refunds and put in place tamper-proof bottles to replace the old-style bottles. As a result, Tylenol continues to be a very effective and prominent brand name in over-the-counter medication. A dark version of effective group deliberation revolves around how well the tobacco industry fought over a period of decades against the evidence provided by public health doctors and others, showing that smoking has catastrophic effects on human health and exacerbates disease. To this day, the tobacco industry has fought very effectively to forestall regulatory change that would damage what is for them a very powerfully incentivised product: one that causes addiction and that has been a public health catastrophe, with all that that entails.

Sunstein and Hastie also emphasise that groups, if they are working correctly and properly, can aggregate information in new and creative ways as a result of an iterative process of discussion and idea testing within groups. Finally, they also suggest that very effective groups generate powerful synergies where new and non-obvious solutions to difficult problems are devised and pursued vigorously. A good example of this kind of thinking has occurred in the aviation industry. A very successful Irish aviation company, Ryanair, originally started life as a full service airline, competing with the historic flag-carriers of Aer Lingus and British Airways, on a very limited number of routes between Ireland and the UK. The company was losing tremendous amounts of money attempting to compete with what were country flagship airlines. Unable to compete, and unwilling to simply go bankrupt, Ryanair decided to invert their business model and become a low-cost airline, with all of the frills removed. Three decades, 100 million passengers per year, and hundreds of millions in euros of profit every year, the rest is history. Other low cost airlines have also come into the market, and the air transport industry was decisively changed as a result. The decision to make this change, and to change the internal policies to make it happen, resulted from two processes: an information gathering one, conducted in the USA, where Southwest Airlines (the original low-cost airline) was studied closely as a model. The second was a deliberative process in Ryanair itself, which considered a number of different possibilities, which included closing the airline. In the end, the low-cost model was chosen, and air travel in Europe was altered dramatically.

Curing the Problems of Deliberative Groups

Sunstein and Hastie make a number of suggestions for curing some of the problems of deliberative groups. There are others available from a consideration of the relevant literature as well. Here are a couple of suggestions. First, be very clear about the purpose of the group. Is it merely an information transmission group, or is it an information-gathering group as well? In other words is the group simply being provided information, but substantive discussion is not required of it? It is very important to ensure that the purpose of the group and the economic cost of assembling that group are considered very carefully. Having figured this absolutely central and elemental problem out, it then may be appropriate to consider the processes and procedures that are to be adopted, depending on the type of problem that the group has to solve. If group problems revolve around large strategic issues that will set the tone and direction for the organisation over a coming period of time, then an appropriate deliberative checklist needs to be established for the group. In essence, the purpose of the checklist is to ensure that during the course of deliberation, the group does not rely simply on the hope that individuals will have sudden insight, that individuals do not suffer from cognitive fatigue, variations in cognitive load or whatever. Next, especially for larger strategic issues, the group should separate information-gathering and solution-generating as much as is possible from decision-making processes. In fact, Sunstein and Hastie suggest that the second stage should be one where senior individuals not normally involved in the group’s work should be brought in to test and provide feedback on the various ideas provided by the group from its solution-generation and information-gathering phase.

Central to optimal group functioning is the idea that cognitive diversity is important and necessary for the group to work at its most effective. Diversity in what each individual brings cognitively to the table allows the group to use individual expertise that each individual may possess and to use it in a way that is most effective for generating the best solution for the problem at hand. For difficult, contentious or simply very important strategic decisions, there must be a formal process that allows all information possessed by all of the members to be brought to the table. There are many ways of ensuring that this can happen, from simple emails to a nominated individual who provides all the information suggested by the group, back to the group, in an anonymous fashion, or whatever. The key thing here, though, is that the chair of the group, the person who will be running the meeting, in turn is open to generating the best solution, which may differ from their preferred solution. Ego might be a problem here. Finally, in the information-gathering stage, having empirical information in the form of data that allow you to benchmark the quality of the decision or decisions that need to be made, should also be made. In other words, the metrics by which you can measure outcomes need to be stated, and these in turn can be used to anchor the discussion. These need to be reliable; they need to be realistic, and in certain cases, may not exist at all. For the launch of an entirely new product or category of product, no such information may be available, but the necessity for the product might be available through other means. A simple example is to go back to the early days of the World Wide Web. It was possible in the early 1990s, over the course of a day, to visit all of the websites that existed on the then internet, using a simple browser such as Mosaic. As the internet grew and evolved, remembering which website to go to was ever more of a problem, and finding which website to go to was also a problem. Lists or spread sheets of these websites started to become available, often in an online fashion, and thus the problem of search was born: an entirely new problem, driven by cognitive overload on the internet, and which required a rapid solution. The solution that most people now use ultimately derives from a slightly obscure application of set theory in mathematics, but which gave us, by certain counts, the world’s most valuable company, namely Google (or Alphabet, as the overall holding company is known).

One final process might need to be added, and this is the process known as a ‘premortem’. This is where, having arrived at the solution or strategy, a fresh team or a reconstituted version of the current team is asked to figure out why, if this product fails, why it has failed. And working through such a counterintuitive solution space might be very valuable indeed, because it forces you to ask questions that otherwise one might prefer to leave unasked. Is our bank sufficiently capitalised? What can we do to prevent a run on our capital? Are we overleveraged? Should we, as a bank (in 2006), jump wholeheartedly into the mortgage market—a market that blew up spectacularly two years later and the fallout from which we are still attempting to adjust? Should we, as a car manufacturer, deliberately massage the safety of our vehicles to sell more products? What will we do if we are found out? Should we as a pharmaceutical manufacturer close arms of clinical trials that are costing huge sums of money, but which are ones in which we as a corporation are heavily invested? These kinds of latter decisions involve recognising that sunk costs are just that—they are gone, and the decision that needs to be made now, needs to be made, as the economists put it, at the margin. The past is just that, past. The money spent is just that, spent. A vital question then is: what should we do now? This is a particularly difficult problem for groups to get beyond—the idea that sunk costs can be recovered (the ‘sunk cost’ fallacy). Someone needs to be empowered to say ‘don’t throw good money after bad’. And meetings need to have stop mechanisms for decisions—just as much they need go decisions. Our brains are biased for action—and, paradoxically, sometimes inaction might just be the best course of action.

Optimally Designing Groups and High-Performance Teams

There has been much debate recently about the related issues of designing groups and high-performance teams to ensure that they function optimally, and avoid the dangers of groupthink, and the myriad other cognitive biases that afflict individual and group decision-making. The design of groups and feedback to individuals within groups can dramatically impair or enhance individual and group performance. One important study (Kishida and colleagues 2012) shows that feedback regarding IQ in a group context can dramatically diminish IQ and individual performance on tasks performed in the group. Furthermore, activity changes are seen in brain areas that support memory and executive function. In other words, the how, where, what and why of feedback provided to individuals while in a group can dramatically reduce their performance. Managers and leaders need to consider very carefully indeed how to give feedback. These data might explain why ‘stack-ranking’ regimes are profoundly demotivating and destructive. A complementary study shows that well-designed groups have a collective intelligence that rises above what you predict from the sum of individual intelligences (Woolley et al. 2010). The key variables are for group designs are: the average social sensitivity of group members, the equality in distribution of conversational turn taking and the proportion of females in the group.

Exercise

  1. 1.

    Think of the worst meeting or series of meeting that you have ever attended. Write down, in short words and phrases, why it was so bad. How could it have been changed?

  2. 2.

    Think of the best meeting or series of meeting that you have ever attended. Write down, in short words and phrases, why it was so good. How could it have been changed?

  3. 3.

    Take the case study that was used to introduce this book. How, if you were Tom Spengler, and knowing what you know happened, would you have set up the group deliberation processes to ensure a different, and perhaps, better outcome. How would you define better—in measurable ways?

  4. 4.

    You have a complex decision to make: you have what looks like a good outcome from a small-scale clinical trial involving about 80 patients who were trialled on a new drug for Alzheimer’s disease. You know that at least 100 candidate drugs have failed in the past and the likely cost of a full-scale clinical trial will be at least in the £100–200 million range. What deliberative processes do you put in place to help the decision about what to do next?

  5. 5.

    You are the new CEO of a failing tech company with a high-recognition product, poor innovation and company culture that is too indecisive and riven with in-fighting. What do you do to turn things around?