“Those who prevent a problem have done their duty; those who tackle it once it has erupted may become heroes.”
Introduction
In many organisations, judgements about work end up focusing on the final figures: how much was spent, how much was saved, how many contracts were completed, how many problems were solved. Numbers are indispensable, but they can become dangerous when separated from the story that produced them. Two apparently similar results can conceal opposite performances; two variances in different directions can have completely different economic consequences; a project completed after a prolonged crisis may receive more attention than one completed on time and without fanfare.
Three episodes from experience in project work illustrate this risk well. All three reveal the same fundamental mistake: confusing the accounting representation of a result with the actual result. The manager looks at the figures, makes them comparable, balances them out or celebrates them, but in doing so loses the most important information: who forecast well, who met the deadlines, who generated additional costs, who prevented problems and who simply solved them after suffering or causing them.
The point is not to argue that managers should automatically reward those who spend less or punish those who spend more. The problem is subtler: forecasts, execution, final results and the causes of variances need to be distinguished. Without this distinction, the appraisal system becomes unfair and, above all, encourages the wrong behaviour.
Budgets: coming in below and above forecast are not always the same thing
Imagine two managers. The first forecast a cost of 100 and finds, at a certain point, that it has reached 120. The second forecast 100 and completes the work having spent 85. A purely statistical reading might say that both got their forecasts wrong: the first by 20%, the second by 15%. From this perspective, there really is a symmetry. The symmetry lies in the fact that both depart from the forecast, not in the size of that departure: 20% and 15% remain different variances. A forecast is good when it comes close to the actual figure, so both overshooting and undershooting represent a variance.
Yet this observation becomes inadequate as soon as we move from the accuracy of the forecast to the assessment of the project. If the manager who spent 85 has finished the work while meeting quality standards and objectives, that manager has delivered a real saving. If the manager who has already reached 120 has not yet finished, the variance is not even final: the cost will probably continue to rise. Moreover, the delay may lead to travel, additional acceptance testing, penalties, prolonged use of staff, or the deferral of revenue or expected benefits.
It is therefore wrong to put the two cases on the same footing with the simple statement: “Underestimating and overestimating are the same thing.” They are the same thing only if what is being measured is the initial accuracy of the estimate. They are not the same thing if what is being measured is overall performance. A project completed under budget and a delayed project already over budget are not two mirror-image mistakes: they are two different economic stories.
The danger of this simplification is that it turns the budget from a management tool into a numerical target. If the manager's aim becomes merely to “come close to the number”, the meaning of the number itself is lost. The budget exists to forecast and allocate resources, but the project exists to deliver a result. Good management must keep both in view.
| Case | Budget and recorded cost | Status | Management significance |
|---|---|---|---|
| A | 100 -> at least 120 | Incomplete | Rising cost, delay and possible loss of benefits |
| B | 100 -> 85 | Complete | Real saving, provided quality standards and objectives are met |
Moving costs does not mean improving performance
A second episode makes the problem even clearer. Two departments, A and B, belong to the same division. A works well; B works poorly. To bring their accounting results closer together, the manager asks some people from A to work for B, but continues to charge their cost to A's contracts.
In accounting terms, the effect is immediate: A gets worse, B gets better and the two departments look more alike. Operationally, however, none of what the figures suggest has happened. B still needs additional resources; A is still the department providing those resources. The cost of the help has simply been assigned to the department giving it rather than the one receiving it.
Before it is even a question of fairness, this is a distortion of information. A management control system should make the causes of differences visible, not erase them. If a department consistently needs staff borrowed from another department, that information is valuable: it may point to organisational shortcomings, ineffective processes, a poor distribution of workloads, gaps in skills or an incorrect estimate of the resources required. Hiding it in A's costs means giving up the opportunity to put B right.
There is also an effect on incentives. Those who work well discover that their efficiency is being used to compensate for others' inefficiency and that, paradoxically, this makes their own indicators worse. Those who work poorly receive help that improves their figures without making the cost of that help transparent. Over time, an organisation built this way may discourage its most dependable people.
The crisis hero and the invisibility of those who work well
The third episode concerns a large project. One part is completed on schedule. Other parts accumulate enormous delays because of malfunctions, require additional acceptance testing and travel, and are delivered around a year and a half late. At the closing meeting, however, the prevailing mood is one of satisfaction: the problems have finally been overcome, the project is complete, and those who fought through to the end receive words of praise.
It is understandable to recognise the effort needed to emerge from a difficult situation. Solving a real problem may require skill, perseverance and willingness to help. The difficulty arises when celebrating the solution wipes out the assessment of the preceding performance. Those who finished on time seem merely to have done their duty; those who arrive eighteen months late, after a long succession of problems, may appear to be the heroes who saved the project.
This phenomenon can be called a culture of crisis. Organisations remember what makes a noise. A night spent restoring a failed system is visible; months of preventive maintenance that stop the system from failing are not. A team that makes a dramatic recovery from a delay becomes a story; a team that never falls behind rarely does. Yet, from an economic and organisational perspective, it is often the second approach that is more valuable.
The paradox is simple: those who prevent a problem have done their duty; those who tackle it once it has erupted become heroes. If this logic is repeated, the organisation ends up rewarding the ability to put out fires more than the ability to avoid starting them.
Why this distortion arises so easily
The first reason is psychological: a crisis is visible. It has a beginning, tension, protagonists and an ending. Well-managed normality, by contrast, produces no narrative. The success of prevention lies precisely in the fact that nothing happens.
The second reason concerns accounting: it is easier to measure what has been recorded in the actual figures than to attribute causes, responsibility and indirect costs correctly. A delay may generate costs that do not appear under the same accounting heading: additional staff hours, the postponement of other projects, lost opportunities, deteriorating customer relations, the need to keep temporary infrastructure in place for longer. If we look only at the total for a contract, some of these effects disappear.
The third reason concerns management: evening out the results makes it easier to present the situation to senior management. Two departments with similar figures are easier to explain than one highly efficient department and one highly inefficient department. But a good control system should not make reality more convenient to present: it should make it easier to understand.
Finally, there is a question of time horizon. The manager may be tempted to fix the quarterly figure, whereas the organisation needs to address the cause that will produce the same problem over the next three years.
Some similar examples
The same mechanism appears in many settings. In an IT department, one team keeps the servers stable for months through testing, monitoring and scheduled updates. Another neglects maintenance, suffers a major outage and works for two days and two nights to restore the service. In the end, the second team is applauded for its dedication during the crisis. The first, which prevented that crisis, remains invisible.
In a factory, one production line consistently operates within tolerances and with little waste. Another line accumulates defects; when the situation becomes critical, technicians and managers organise an extraordinary intervention and manage to deliver the order at the last moment. Here too, the ability to recover is positive, but it must not replace an analysis of the cause that made recovery necessary.
In a sales department, one manager makes prudent, realistic forecasts and comes almost exactly on target. Another sets very low targets, easily exceeds them and is celebrated for achieving 120% of target. Without examining how the target was set, the percentage rewards whoever negotiated the initial number more successfully, not necessarily whoever delivered the better result.
Something similar can also happen in public administration or major infrastructure projects: a scheme completed on time is considered normal; a scheme that accumulates years of delay may attract great attention at its inauguration. In the public mind, the final ceremony risks erasing the cost of the path that led to that day.
The principle is recognisable even in everyday life. If someone always pays their bills on time, nobody notices. If someone else accumulates problems and then, through an enormous effort, manages to avert the consequences at the last moment, the final act may seem more deserving than the first person's consistency. The emotion of recovery tends to outweigh the value of continuity.
Measuring well means separating the questions
Many errors of judgement arise because a single indicator is expected to answer different questions. In reality, a project should be examined along distinct dimensions. The first question concerns the forecast: how accurate was the initial budget? The second concerns the economic result: what did the project actually cost? The third concerns time: was it completed when planned? The fourth concerns quality: does the final product meet the requirements? The fifth concerns causes: were the variances due to unforeseeable events or avoidable mistakes? The sixth concerns recovery: once the problem emerged, was it managed well?
These questions may produce different judgements about the same person. One manager may have made a weak initial estimate but managed execution very well. Another may have inherited an accurate forecast and then created operational problems. A third may have inherited a crisis they did not cause and resolved it very effectively. Only by separating these dimensions can we recognise merit and responsibility without confusing them.
Savings also need to be interpreted. Spending less is not automatically an achievement: it might mean having cut necessary activities, reduced quality or deferred costs to the future. Equally, however, spending less cannot automatically be treated as a mistake equivalent to an overrun. We need to ask what was achieved with that expenditure and under what conditions.
An incentive system can create the very behaviour it later deplores
Every organisation teaches people what matters through what it measures and rewards. If spectacular recoveries receive the greatest rewards, crises will become more visible. If those who meet deadlines and budgets are taken for granted, consistency will lose its symbolic value. If costs can be moved from one centre to another to make results more uniform, managers will learn to manage indicators instead of processes.
This does not mean that someone deliberately creates a problem in order to solve it afterwards. The mechanism is subtler. People naturally direct their attention towards what receives recognition. If an organisation continually celebrates those who intervene in a crisis but not those who build robust systems, it will end up investing more energy in recovery than in prevention.
Managers should therefore also reward the absence of adverse events when that absence is the result of good management. The difficulty lies in making visible what normally goes unseen: risks eliminated, errors prevented, accurate estimates, deadlines met, problems solved before becoming crises.
What a good manager should do
A good manager does not have to choose between numbers and reality: they must use numbers to understand reality. They should keep resource costs traceable, allocating them to the work actually performed; distinguish variances in forecasting from variances in execution; record the costs of delay alongside direct costs; recognise recovery from a crisis without erasing its causes; and, above all, compare results against comparable objectives.
When an efficient department lends staff to one in difficulty, that support may be the right decision. But the transfer must remain visible. The management message should be: “B needed ten person-days from A”, not “A and B have similar costs”. In the first case, the information helps decisions; in the second, it hides the problem.
Similarly, when a project is finally completed after a prolonged crisis, it is right to thank those who worked hard to finish it. But the closing meeting should also include a second question: why have we arrived here eighteen months late, and how do we prevent it from happening again? Only then can recognition of effort avoid becoming absolution for inefficiency.
Conclusion: do not adjust the thermometer
The three episodes lead to the same lesson. Good management is not about making results look balanced, but about understanding why they differ. It is not about bringing every figure close to the average, but about making the causes that produced those figures clear. And it is not about celebrating only those who solve the most visible problems, but also about recognising those who work in ways that prevent those problems from arising.
The manager who moves costs from one department to another, who treats a project completed under budget as equivalent to one still open and over budget, or who celebrates recovery from an enormous delay without making distinctions, risks performing the same operation every time: correcting the representation instead of reality.
The simplest metaphor is the thermometer. If the temperature is too high, we can address the cause or tamper with the instrument. In the short term, the second method produces a more reassuring number. In the long term, however, the fever remains.
In a healthy organisation, real merit does not lie solely in knowing how to put out fires. It also lies in designing buildings in which fires are less likely. And those who deliver well, on time and without fanfare have not simply “done their duty”: they have produced precisely the kind of reliability that every good organisation should learn to see and value.
What I Would Like to Leave My Children
My children, I would like all three of you to learn to look behind the numbers, especially when they seem too convenient. Ask what they measure, how they were obtained and who is paying a cost that does not appear in the table.
Do not let yourselves be persuaded that quiet work is worth less than a spectacular rescue. Be grateful to those who solve a problem for you, but also to those whose care and consistency spared you from having it. And if one day it falls to you to judge other people's work, do not confuse those who caused a difficulty with those who inherited it.
I hope you will be able to speak honestly about what you do: the results, the mistakes and the help you received. Do not look for a number that makes you appear better; strive to do something well that the number can bear witness to.
Papy