Jérôme Mendoza-Trussart — “Transforming Without Causing Harm”: When Performance Depends on Those We Do Not See

Jérôme Mendoza-Trussart

During an assignment, I observed an organisation undergoing a profound transformation. A new director had arrived with a mandate to carry it out. Convinced that a fresh start was necessary, he had replaced the entire leadership team, believing that changing the people would transform the organisation.

Yet the difficulties did not disappear. Employees had to compensate for the dysfunctions to keep operations going. The organisation continued to report results, but it was becoming more fragile: staff turnover was rising, the change was failing to take root, and the human damage was mounting. Eventually, the organisation itself was at risk.

What interests me in this situation is the paradox it reveals: an organisation can continue to deliver results even as the way it operates deteriorates. The results are real, but they depend in part on people compensating for what the system can no longer handle properly. Performance indicators therefore tell only part of the story.

This question informed my analysis and the writing of my book, Transforming Without Causing Harm: what should we look at to understand what actually enables an organisation to transform? Beyond the decisions made, the objectives announced and the results achieved, we need to examine what the work really demands, and what some people have to add, work around or redo to get the job done.

I use the term ‘borne gaps’ for the gaps between the intended way of operating and what the work actually requires, when those gaps are absorbed by individuals rather than addressed by the organisation. They may allow operations to continue for a time. But when they become permanent, they exhaust available resources and weaken the whole system.

The conclusion is simple: a transformation cannot be judged solely on its immediate results. We must also understand what makes those results possible, what they cost, and whether the new operating model can genuinely function without depending on invisible compensations.

Jérôme Mendoza-Trussart

In Transforming Without Causing Harm, I use the image of a building undergoing renovation. Some residents have returned to an almost normal life in refurbished surroundings. Others are still living among dust, protective sheeting and ongoing work. Yet they inhabit the same building and are going through the same renovation.

Within an organisation, some employees are already working under the new rules while others still have to keep the old system running. Both ways of operating may remain necessary at the same time.

This is what I call the pivotal phase: the point at which the old system must still maintain operations because the new one is not yet stable enough to take over completely.

The risk arises when the project timetable is mistaken for the reality of the work. The transformation is declared complete, new practices are rolled out and the expected results are demanded, even though essential conditions are not yet in place.

Teams then have to make things work together that do not yet work together. They keep what works, find ways round what does not and compensate for what is missing. Part of the work needed to maintain operations remains outside the project's formal scope.

To decide whether to slow down, I recommend asking three questions: what does the new operating model not yet cover? What must nevertheless continue to function? Who is carrying what is missing, and for how long have they been doing so?

A gap can be both necessary and temporary. Every transformation involves adaptation. But when these adjustments become permanent, essential to operations and repeatedly fall on the same people, they point to a problem in design or implementation.

Slowing down does not mean abandoning transformation. It means acknowledging that the transition is not complete and intervening where the new arrangements are not yet working. The right pace is not the one that follows the timetable at all costs; it is the one that allows the new system to take over effectively.

Jérôme Mendoza-Trussart

Invisible work leaves traces. We need to learn to recognise them: files being reworked, errors corrected before they are detected, duplicate data entry, interruptions, informal decisions, information sought from colleagues, or essential tasks for which no one is officially responsible.

In the book, for example, Paul is required to switch from an old software system to a new tool. The roll-out is considered complete and adoption indicators are rising. Yet he keeps two windows open and continues moving between the systems to complete his work. The dashboard shows adoption; observing the work reveals a transition that is still unfinished.

Simply adding more indicators is therefore not enough. We need to compare the available data — timeframes, quality, costs and incidents — with what people actually do to produce the results. Where is work being redone? What gets corrected before it is reported? Which tasks are carried out without having been planned? Which difficulties are absorbed locally?

An activity may appear unnecessary while performing an essential function in prevention, coordination or regulation. Removing it may improve an indicator in the short term and weaken the system in the medium term.

I am not trying to attach artificial numerical values to every human contribution. I want to make visible the mechanisms that enable results to be achieved, but also what those mechanisms cost the people producing them.

When a reorganisation overlooks the people who actually keep operations running, they often have to compensate for shortcomings in the new arrangements, manage contradictions, anticipate problems and find solutions on their own. This additional burden is not merely operational: it demands their attention, increases their mental workload and can affect their psychological health.

A transformation genuinely improves an organisation when the quality and continuity of its operations improve without relying indefinitely on these invisible compensations, or imposing on people a human cost that performance indicators fail to measure.

Jérôme Mendoza-Trussart

Disagreement is useful when it allows a decision to be tested before its consequences emerge. But allowing people to speak is not enough: in a hierarchical organisation, challenging a direction can carry a professional cost. Silence therefore means neither agreement nor the absence of a problem.

In Transforming Without Causing Harm, Fabrice knows that a project timetable is unworkable and that the teams are exhausted. Yet in meetings he tones down his reservations. He does not necessarily lack information; he doubts he can express it without exposing himself to consequences. Senior management may then make a decision based on information that has already been filtered.

For an objection to carry real weight, it needs to be addressed in concrete terms. What fact prompted it? What risk does it reveal? Which assumption underlying the decision might it challenge? What do we decide after examining it? If the original decision stands, why?

Constructive challenge is neither systematic opposition nor a right of veto. It means testing assumptions, examining consequences and improving the decision. The leader remains responsible for making the final judgement, but must be capable of hearing information that challenges their analysis.

The quality of governance is therefore not measured by the number of meetings in which everyone is allowed to speak. It is measured by the organisation’s ability to bring difficulties to the surface, examine them without penalising those who raise them, and explain the decisions that follow. During a transformation, hierarchical power must be suspended to allow genuine debate and the identification of the real problems. This is why a transformation process led and organised by an external consultant makes it easier to bring obstacles to light.

Jérôme Mendoza-Trussart

Resistance to change is not always a refusal to make progress. It may indicate that a seemingly secondary practice performs a function the project has failed to identify. A procedure may, for example, compensate for a failure elsewhere in the process. Removing it without addressing that failure merely shifts the problem.

But we must not turn every instance of resistance into proof of some hidden intelligence within the system, either. Certain established balances protect vested interests, power relationships or habits that have become costly.

To tell the difference, I propose examining the facts. What does the contested practice actually make possible? What problem does it solve? Who relies on it? What would happen if it disappeared tomorrow? Conversely, what costs or difficulties does keeping it in place perpetuate?

This analysis helps distinguish what should be preserved, what should be replaced and what should be abandoned. Understanding a practice does not mean legitimising it; it prevents us from abolishing it blindly.

The leader’s role is not to preserve every existing balance. It is to understand what those balances produce, then build a more viable way of operating. A successful transformation does not simply remove the old system: it takes over the useful functions that the previous system performed, sometimes informally.

Jérôme Mendoza-Trussart

Transforming without causing harm does not mean avoiding every rupture or promising that no decision will be painful. Economic constraints may lead to redundancies, the closure of an activity or the disappearance of an occupation. The question is how responsibility is exercised before, during and after the decision.

Before deciding, it is necessary to investigate and observe. Which skills will disappear? Which tasks will still need to be performed? What workload will be transferred to those who remain? Which dependencies currently rest on a handful of people? Can the proposed operating model genuinely ensure continuity of service?

Reducing headcount may improve a ratio while weakening knowledge transfer, quality or the ability to deal with unexpected situations. These consequences do not automatically make the decision illegitimate, but they must be identified and taken into account when making the judgement.

Responsibility does not end with the announcement. It means checking the actual effects of the decision, listening to what the work itself reveals, and correcting anything that jeopardises continuity of operations or places an unsustainable burden on the remaining employees.

Transforming without causing harm is not about eliminating difficulty. It is about refusing to treat human and organisational consequences as secondary damage simply because they are less visible in the initial calculation.

Jérôme Mendoza-Trussart

Artificial intelligence can reduce the time spent on certain tasks and speed up processes. But saving time on a task does not, in itself, prove that the organisation is functioning better.

We need to look at what disappears when a task is automated, what remains necessary and what is transferred elsewhere. The time saved may be absorbed by checking outputs, correcting errors or dealing with situations the tool cannot handle. We must also examine the effects on skills, judgement and the ability to spot anomalies.

In the book, a character demonstrates that AI can process a client file in eleven minutes. The result seems spectacular. But speed alone is not enough to assess the gain: we still need to know what was checked, what was omitted, what needs to be redone and which human activity previously made the outcome reliable.

I use the expression ‘disappearance of useful work’ to describe the loss of a useful activity whose function had not been identified before the transformation. This is not about defending every existing task, but understanding what those tasks make possible before removing or automating them.

To assess a gain, we therefore need to consider time and costs alongside quality, errors, rework, risks and the burden of verification. Automating a task does not necessarily eliminate the work that the task previously made possible. Failing to take the time to examine this exposes the transformation to failure.

Jérôme Mendoza-Trussart

I would ask: ‘How far are you prepared to push your organisation’s performance if it comes at the expense of your teams’ mental health?’

An organisation’s performance is not measured solely by the results it achieves. It is also measured by its ability to produce them sustainably, without exhausting the people who make them possible.

When a transformation leaves problems unresolved, the work does not disappear. It moves elsewhere. Teams compensate for shortcomings in the new arrangements, manage contradictions, anticipate difficulties and find solutions to keep operations running. Some of this work remains invisible, but it comes at a cost: in time, attention, mental workload and, sometimes, psychological health.

The risk is to regard a transformation as successful because the indicators look good, even though the way the organisation functions depends on additional efforts that no one has properly assessed.

The question, then, is not whether to choose between performance and the health of teams. It is whether an organisation can genuinely improve its performance by weakening the people who produce it.

Transforming without causing harm means refusing to confuse achieved results with a way of operating that has become sustainable. It means making the ability of teams to continue working in good conditions a fully fledged criterion of success. It is possible to transform in order to maintain performance in this constantly changing world while also protecting teams’ mental health. It can be done by changing gear.