Getting things doneGetting things doneby Matthew Leitch, 17 May 2002
No theory of how to get things done has had more impact on personal and business life in the last fifty years than a particular mix of control theory and pseudopsychology that I know you will recognise immediately. The theory goes that the best way to get things done is to set clear and specific goals up front, backed up by clear milestones, monitor your progress against them, and take corrective action when actual achievement deviates from your plan. It is usually linked with the belief that setting aspirational goals is good, and that positive thinking is better than negative thinking. "Focus on success" and keep using feedback until you get it.
This appears in various forms including budgetary control, management and control of projects, and advice on how to achieve success in life, boost your motivation, win at sport, and so on. There's even been research that seems to show that optimists are happier and healthier than pessimists, even though the pessimists are more often right. (Optimists are people who tend to think things will turn out well, while pessimists tend to expect things to turn out badly.)
And yet, despite the enormous popularity of these ideas and their status as "good management common sense", I will show that they are associated with a series of damaging errors in reasoning about uncertainty. Some of these are so serious, and so common, that they amount to a strong case against this whole approach.
As someone with an education in science, mathematics, and logic I sometimes forget that there are many people who deliberately cultivate beliefs they also know to be distortions of the truth. Here are some of the errors of that type:
Pessimism is the tendency to expect bad outcomes and to think about them disproportionately. It may be adopted as a way to avoid disappointment, but is usually more likely to predict the future correctly than its opposite, Optimism. We have already seen how people tend to underestimate the variability of outcomes, such as the value of the FTSE on a particular day. The Pessimist is someone who does this and also tends to expect bad outcomes. Faced with a proposed course of action the Pessimist sees only problems and downside risks and expects failure or at least a poor result. The Pessimist, therefore, is reluctant to act since nothing will be achieved anyway, and fails to anticipate, encourage, and exploit opportunities.
Optimism is the tendency to expect good outcomes and to think about them disproportionately. It may be adopted as a way to avoid unpleasant thoughts, or more often out of a belief that Optimism is good, creating enthusiasm, "self confidence", or mobilising the mind, perhaps unconsciously, to move in the direction of success. Think about good things and they are more likely to happen, goes the theory of positive thinking. Some believe that visualising success is a good thing to do, while visualising failure should be avoided. Like the Pessimist, the Optimist has too narrow a view of future outcomes, but also tends to expect good outcomes. Faced with a proposed course of action the Optimist sees only advantages and opportunities and expects success. The Optimist, therefore, is reluctant to think about risks that might jeopardise success or do anything about them. He/she will continue with a venture long after a rational person would have abandoned it as a bad bet. Optimists may be happier and healthier than others but not necessarily as successful, and the consequences of their faulty decisions may well contribute to the unhappiness and ill health of others.
Optimism in groups extends the philosophy of positive thinking to group behaviour. In a group dominated by this theory anyone who raises downside risks or otherwise appears to be thinking "negatively" will be disapproved of and may be regarded as not a "team player". A group thinking this way may suppress uncertainty, fail to manage downside risks, and resist admitting failure until it is long since obvious to others.
The alternative to these views might be called Rational Uncertainty Management. Rational Uncertainty Managers consider the full range of outcomes and have reasonable expectations about the likelihood of their occurrence. They take action to make good outcomes more likely and prepare to exploit them fully if they occur. They also take action to make bad outcomes less likely and prepare to minimise the damage if they should occur. If the odds really don't justify continuing with something, they stop - after the Pessimist, but before the Optimist.
The Rational Uncertainty Manager benefits from better decisions flowing from a much more realistic view of the future without the lethargy of the Pessimist. This approach natually promotes taking action rather than being passive.
Example: Imagine a long distance running race. You are in fourth place and some way behind the person at number three. You're in pain and tired. Someone who believes in positive thinking would need to convince themselves that they were definitely going to get into the top three in order to justify making that effort. With an optimistic attitude they might succeed in convincing themselves of this and pick up speed to catch up the runner ahead - at the risk of exhausting too early and dropping places, which is a risk they prefer to ignore. The pessimist would hang back and concentrate on keeping ahead of the person behind. The Rational Uncertainty Manager would recognise that the outcome is uncertain. Runners ahead could be about to reach exhaustion or suffer injury. Your own resources may be greater than expected - or lower. The most likely outcome may be fourth place but there is a reasonable chance of doing better. Despite the pain and fatigue you know you are not in imminent danger of injury and, since you have nothing to do that requires great physical effort over the next few days, you could exert yourself to complete exhaustion without penalty other than the discomfort at the time. If the chance of a better position justifies the discomfort you press on, adjusting speed based on the risk of exhausting too early or not fully exerting.
It helps to be able to distinguish between better and worse outcomes. However, this idea is often over-extended to a belief in the special value of "specific" objectives/goals/targets/budgets set at the outset of a venture, leading to a number of errors:
Distorted relationships between outcomes and utility: There are many ways that the value (aka utility) of different outcomes varies. Sometimes only a very specific outcome is useful, and anything else is not. In other situations utility varies more gently. For good decision making it helps to try to get the most accurate view possible of the value of different outcomes. Mistakes can be made when the relationship between outcomes and utility is distorted.
Many people believe that "specific" objectives are important, even essential. For example, saying "We'd like the company to grow this year" is not considered as good as "We aim to grow the company by 5.5% by 31 March 2003." This tends to create the view that growth of 5.5% is success and highly valuable, growth of just 5.45% is failure, and not nearly so valuable, while growth of 7% is not much better than 5.5%. This is reinforced if the target is used as the trigger for payments of money or other rewards to employees. However, in a situation like this it is unlikely that this variation in utility with outcomes is realistic and such targets can easily lead to poor decisions. For example, "We've no chance of reaching budget this month so let's forget it and go to a conference." or "We've already reached our target for the year so let's relax for this month." More subtly, a plan might be drawn up specifically to reach a particular specific outcome but in such a way that the plan is brittle, achieving the target or collapsing and achieving very little.
Overstating the certainty behind goals: It is often very difficult to say in advance precisely how you will value a particular future outcome. Fortunately it is rarely necessary and looser statements about objectives are quite adequate for many purposes. A mistake is to make a very specific objective when there is little certainty about what the objective should be. It is better to reflect the uncertainty in the objective and encourage reconsideration as more information becomes available and more thinking is done.
Premature decisions: Setting a specific goal early on can lead to premature decision making. With most decisions one of our options is to defer a decision and wait for more information, or for conditions to change. (In "real options" theory this is called a "wait option".) There are many ways we can benefit from deferring decisions, or doing things speculatively so that if events unfold in different ways we will be well positioned. A premature decision is like throwing away the value of that option.
Confusion between plans and forecasts: The word "budget" is a source of much confusion. Is a budget a plan or a forecast? Many people appear to think they are both, and that may be the case at the moment the budget is made. However, as soon as time has passed and uncertainty has done its work the budget should be superceded by a more up to date forecast, and now the confusion often begins. It is important to be clear about the difference between plans and forecasts, and to recognise that forecasts are needed.
Reasoning from outdated goals: Goals are not sacred. Setting a goal is a technique for making and communicating plans. One hypothesises that achieving some outcome (not necessarily very precisely specified) would be beneficial, and tries to think of a way to bring about that outcome. As time goes by, conditions change, and more is learned, it may emerge that the benefits of the outcome are not what was expected initially, in which case it is time to reconsider.
Unfortunately, many people do not accept that goals should change and continue to reason from goals set some time ago that have since been outdated and should have been revised. Examples are:
Judging the success/failure of a venture by comparison of actual results with the original targets or expectations. This is almost certain to be inappropriate as few worthwhile ventures are so simple and so brief that the original targets are still relevant.
Judging a person's performance by comparison of actual results with the original "agreed objectives". This is a more complex example since there is also a difference between the conditions assumed at the beginning of the year and those that actually applied. Someone may have performed brilliantly against unexpectedly difficult circumstances and achieved less than the original target.
Deciding to rethink a plan or reallocate resources because actual results have diverged from original expectations. It is more rational to rethink or reallocate when conditions change.
Tedious comparisons of actual financial results with original budgets, often long after the budgets have become meaningless, and often despite the original budgets being little better than guesswork or wishful thinking.
It makes more sense to revise one's ideas about what is valuable and use these as the basis for judging success/failure. We can then compare these with the original target to learn about our abilities to forecast.
Committing prematurely to the wrong goals: In some situations it is not sensible to have a specific goal to which you are committed. Instead, it can be more effective to have identified a number of things that would be desirable if the opportunity arose. Quite often we need to go into a situation and see what might be in it for us. For example, Neil Rackham's research on negotiation and sales showed that more effective negotiators and sales people prepare for a variety of outcomes. The best sales people have more than one objective from a sales meeting and adjust their expectations and goals during the meeting depending on how things turn out. The best negotiators think of lots of things to ask for so they can be flexible in the face of resistance to some of their requests. Insisting on a specific goal in these situations is ignoring the true uncertainty of the situation and detailed empirical studies show it is not effective.
The foregoing points have given a number of reasons why monitoring progress against a plan may not be a rational way to think about uncertainty. However, there are yet more.
False assumption that a corrective action exists and can be found: The theory of control against a target or plan assumes that a "corrective action" exists and can be found to bring actual results once more into line with the plan. However, in some situations this assumption is not correct. Many projects are constrained by a deadline, resource constraints, and quality requirements. If there is a departure from the plan there is nothing to trade to get back on track. For example, if you are falling behind you can't put more people on the task because that would cost more money so putting the costs over budget. Furthermore, in some real life situations the first departure from plan is a foretaste of what is to come since it indicates that, generally, conditions are proving more or less favourable for the project than was expected.
Delayed action: A simple but fundamental problem with feedback from actual results is that it is often too late to be useful. The time to rethink is the moment news is received of some change in conditions or risks, not the moment when that change eventually feeds through into measurable problems. Overconfidence in the power of negative feedback loops (as they are known in control theory) can lead to failure to search for and monitor news of events relevant to the project or operation and to act quickly on them. It is like driving a car with your eyes fixed on the mileometer. It is much safer to drive with your eyes looking out of the front windscreen at the road ahead. Business leaders should spend a lot of time looking at the road ahead and forecasting in detail the implications of what they see in order to detect conditions which are outside the capabilities of their companies. For example, to meet a raised level of demand with the existing number of packers, or drivers, or pallets, or storage space, and so on.
It makes more sense to evaluate progress against your current view of what outcomes are valuable and what the goals should be.
© 2002 Matthew Leitch